pennsylvania economic associationaux.edinboro.edu/pea/pub/proceed2009/pea proceedings 2009.pdf ·...

Pennsylvania Economic Association

AAnnnnuuaall CCoonnffeerreennccee JJuunnee 44 –– 66,, 22000099

West Chester University of Pennsylvania West Chester, PA

PROCEEDINGS

OF THE

PENNSYLVANIA

ECONOMIC

ASSOCIATION

2009 CONFERENCE

June 4 – 6, 2009

West Chester University of Pennsylvania West Chester, Pennsylvania

James J. Jozefowicz, Editor Indiana University of Pennsylvania

Proceedings of the Pennsylvania Economic Association 2009 Conference i

PPeennnnssyyllvvaanniiaa EEccoonnoommiicc AAssssoocciiaattiioonn:: 22000088--22000099

Executive Board

• President: James Jozefowicz, Indiana University of Pennsylvania • President-Designate: Gerald Baumgardner, Penn College • Vice President, Program: John Misner, Slippery Rock University • Vice President, Publicity: Orhan Kara, West Chester University • Vice President, Membership: Natalie Reaves, Rowan University • Secretary: Stephanie Brewer, Indiana University of Pennsylvania • Treasurer: James Dunn, Edinboro University of Pennsylvania • Editor, Pennsylvania Economic Review: John Misner, Slippery Rock University • Webmaster: Michael Hannan, Edinboro University of Pennsylvania • Immediate Past President: Kenneth Smith, Millersville University of Pennsylvania

Board of Directors

• Steven Andelin, Penn State - Schuylkill • Deborah Gougeon, University of Scranton • Johnnie Linn, Concord University • Brian O'Roark, Robert Morris University • Abdul Pathan, Pennsylvania College of Technology • Brian Sloboda, US Postal Service

Ex-Officio Directors

• David Altig, Federal Reserve Bank of Atlanta • Thomas O. Armstrong, Pennsylvania Department of Community & Economic Development • Yaw A. Asamoah, Indiana University of Pennsylvania • William Bellinger, Dickinson College • David Culp, Slippery Rock University • Donald Dale, Muhlenberg College • Robert D'Intino, Rowan University • Andrew Economopoulos, Ursinus College • Mark Eschenfelder, Robert Morris University • Joseph Eisenhauer, Wright State University • Donna Kish-Goodling, Muhlenberg College • Andrew Hill, Federal Reserve Bank of Philadelphia • Elizabeth Hill, Penn State-Mont Alto • Mehdi Hojjat, Neuman College • Tahereh Hojjat, DeSales University • Ioannis N. Kallianiotis, University of Scranton • Rick Lang, Federal Reserve Bank of Philadelphia • Daniel Y. Lee, Shippensburg University • Robert Liebler, King's College • Patrick Litzinger, Robert Morris College • Stanley G. Long, University of Pittsburgh/Johnstown

mailto:[email protected]





































Proceedings of the Pennsylvania Economic Association 2009 Conference ii

• Jacquelynne McLellan, Frostburg State University • Tracy C. Miller, Grove City College • Lawrence Moore, Potomac State College of West Virginia University • Gayle Morris, Edinboro University • Heather O'Neill, Ursinus College • William F. Railing, Gettysburg College • Margarita M. Rose, King's College • William Sanders, Clarion University • John A. Sinisi, Penn State University-Schuylkill • Lynn Smith, Clarion University • Osman Suliman, Millersville University • Thomas W. Tolin, West Chester University • Paul Woodburne, Clarion University • Bijou Yang-Lester, Drexel University • David Yerger, Indiana University of Pennsylvania
















Proceedings of the Pennsylvania Economic Association 2009 Conference iii

Editor’s Introduction and Acknowledgements

James J. Jozefowicz Editor of Proceedings

2009 Annual Conference of the Pennsylvania Economic Association The papers published in this volume were presented at the 2009 Annual Conference of the Pennsylvania Economic Association (PEA) held at West Chester University of Pennsylvania from June 4 to 6, 2009. The program lists all presenters, session chairs and discussants. Only the papers and comments submitted according to manuscript guidelines are included in the Proceedings. I thank Debbie Bacco of the Indiana University of Pennsylvania Department of Economics for her editorial assistance with the Proceedings manuscript. The 2009 conference was a great success. Participants from across Pennsylvania and several other states, as well as, other countries gathered to share ideas. University faculty, research professionals, graduate students and undergraduate students presented papers and participated in discussions. In addition to the presentations in the concurrent sessions, the conference featured an excellent talk at Friday’s lunch and a panel discussion arranged by the Federal Reserve Bank of Philadelphia. After Friday’s lunch, Dr. Mark Zandi, Chief Economist of Moody’s Economy.com, presented a talk discussing the current economic crisis and its possible paths of recovery, as well as, the impact of potential health care reform on the U.S. economy. The Friday afternoon Federal Reserve Panel included a presentation on teaching monetary policy during an economic crisis and provided an economic outlook with an emphasis on trends and cycles, as well as, a discussion of monetary policy in the months ahead. The success of any conference depends on many individuals. The PEA extends special thanks to Orhan Kara for his time and energy coordinating all local arrangements. We also thank Dr. Christopher Fiorentino, Dean of the College of Business and Public Affairs at West Chester University of Pennsylvania, for his support and willingness to host the conference. The PEA is grateful to John Misner for assembling the conference program. The efforts of Tom Tolin, who hosted the 2003 PEA Conference at West Chester University, also are greatly appreciated. The PEA also thanks the Federal Reserve Bank of Philadelphia for the Friday afternoon Panel and Discussion and for hosting the Friday afternoon reception. Additional thanks go to the entire PEA board for their work making the conference a success. Lastly, thanks to all of the participants who made the conference an interesting, stimulating and friendly place to share ideas.

Proceedings of the Pennsylvania Economic Association 2009 Conference iv

Table of Contents Conference Agenda ............................................................................................................................................. page 1 Correlations Between Exchange Rates and Trade and Investment Variables Beverly Frickel, Vani V. Kotcherlakota, and Frank Tenkorang, University of Nebraska at Kearney ................ page 19 Technology Transfer and the Keystone Innovation Grant Initiative Thomas Armstrong, Pennsylvania Department of Community & Economic Development ................................ page 25 Examining the Existence and Extent of Anti-Intellectual Attitudes among University Students Robert Balough, Clarion University of Pennsylvania ........................................................................................ page 46 The Study of Vietnamese Small Business Owners in America: Their Patterns of Strategic Decisions Hung M. Chu and Lei Zhu, West Chester University of Pennsylvania ............................................................... page 55 The Effect of Gender on Learning and Success in Economics Classes Orhan Kara, West Chester University of Pennsylvania, and I-Ming Chiu, Rutgers University-Camden .......... page 63 Some Major Campaign Issues of 2008: An Economic Perspective; A Student-Faculty Project William Sanders et al., Clarion University of Pennsylvania .............................................................................. page 71 Katyusha Computations or Rocket Science: Using Economic Order Quantity to Assess Sustainability of Hostile Rocket Offensives Johnnie Linn, Concord University ...................................................................................................................... page 75 Drug Approvals and Drug Safety: Preliminary Results Natalie Reaves, Rowan University...................................................................................................................... page 79 Factors Impacting the Throwaway Society: A Sustainable Consumption Issue John McCollough, Pennsylvania State University-Lehigh Valley ...................................................................... page 86 Discussant Comments: Wage Growth and Employment Growth: Evidence from Wage Records Lynn Smith, Clarion University of Pennsylvania ................................................................................................ page 92 Optimum Currency Areas and Synchronization of Business Cycles in Sub-Saharan Africa Yaya Sissoko, Indiana University of Pennsylvania ............................................................................................. page 93 Does Consumer Sentiment Affect Household Saving? I-Ming Chiu, Rutgers University-Camden ........................................................................................................ page 105 The Effects of Urbanization on Development: Analysis of Latin America and Sub-Saharan Africa Chelsea Kaufman, Clarion University of Pennsylvania ................................................................................... page 116 Possible Relationships between the Theory of Global Warming and Stock Market Declines David Nugent, Slippery Rock University of Pennsylvania ................................................................................ page 127 Economical Valuation of Forests: Iranian Case Study Sadegh Bafandeh Imandoust, Payame Noor University ................................................................................... page 130 Reclaiming Institutions as a Form of Capital Bénédique Paul, University of Montpellier ...................................................................................................... page 137 Author Index ................................................................................................................................................... page 149

Proceedings of the Pennsylvania Economic Association 2009 Conference 1

Pennsylvania Economic Association

22000099 CCOONNFFEERREENNCCEE AAGGEENNDDAA

THURSDAY, June 4 (Location: WCU Graduate Business Center)

• 4:00 pm - 9:00 pm Registration (Third Floor Lobby) • 5:00 pm - 7:00 pm Board of Director's Dinner/Meeting (Room 126) • 7:00 pm - 10:00 pm Reception (Room 126)

FRIDAY, June 5 (Location: WCU Graduate Business Center)

• 8:00 am - 4:30 pm Registration (Third Floor Lobby) • 8:30 am - 10:30am Coffee Break • 9:00 am - 10:15 am Concurrent Sessions • 10:15 am - 10:30 am Break • 10:30 am - 11:45 pm Concurrent Sessions • 12 noon - 1:15 pm Luncheon (Room 126) • 1:15 pm - 2:00 pm Lunch Speaker: Dr. Mark Zandi, Chief Economist,

Economy.com (Room 126) • 2:15 pm - 3:30 pm Concurrent Sessions • 3:30 pm - 3:45 pm Coffee Break • 3:45 pm - 4:45 pm Fed Panel Discussion (Rooms 325 & 326) • 4:45 pm - 5:45 pm Reception (hosted by the Federal Reserve: Room 126)

SATURDAY, June 6 (Location: WCU Graduate Business Center)

• 8:00 am - 9:00 am Registration & Coffee (Third Floor Lobby) • 9:00 am - 10:15 am Concurrent Sessions • 10:30 am -11:00 am General Membership Meeting (Room 126) • 11:15 am Closing (Pre-ordered boxed lunches may be picked up at the Registration

Desk)


FRIDAY, June 5, 2009 8:00 a.m. – 4:30 p.m.

Conference Registration WCU Graduate Business Center (Third Floor Lobby)

8:30 a.m. – 10:30 a.m. Coffee Break: (Third Floor Lobby)

FRIDAY, June 5, 2009 9:00 a.m. – 10:15 a.m.

Session F1A: International Economics Room: 305

Chair: Kenneth Smith, Millersville University Correlations Between Exchange Rates and Trade and Financial Flows Beverly Frickel & Vani V. Kotcherlakota, University of Nebraska at Kearney South Africa’s Evolving Competitive Exposure in U.S. Import Markets Yaya Sissoko & David Yerger, Indiana University of Pennsylvania Does Cultural Distance Influence US Intra-Industry Trade? Roger White, Franklin & Marshall College Discussants: Yaya Sissoko, Indiana University of Pennsylvania Roger White, Franklin & Marshall College Vani V. Kotcherlakota, University of Nebraska at Kearney

Session F1B: Student Session Room: 306

Chair: James Jozefowicz, Indiana University of Pennsylvania

A Cross-Sectional Analysis of Variation in Violent Crime Rates in the Counties of Pennsylvania Antonio Ayllon, Indiana University of Pennsylvania Income Inequality and Population Growth Rates: An Analysis of Post-Industrial Europe Stephanie Bearjar, Indiana University of Pennsylvania

In Poor Health: An Analysis of Income and Women’s Reproductive Health in Pennsylvania Counties

Aleta Haflett, Indiana University of Pennsylvania Discussants: Arpita Biswas, Clemson University Jialu Liu, Indiana University-Bloomington Benedique Paul, University of Montpellier


Session F1C: Economic Development Room: 312

Chair: Michael Hannan, Edinboro University of Pennsylvania

Technology Transfer and the Keystone Innovation Grant Initiative Thomas Armstrong, Pennsylvania Department of Community & Economic Development

Economic Impact of Tolling Interstate 80 on Adjacent Counties

Tracy Miller, Grove City College Discussants: Tracy Miller, Grove City College

Thomas Armstrong, Pennsylvania Department of Community & Economic Development

Session F1D: Health Education and Welfare Room: 320

Chair: Thomas Tolin, West Chester University The Effects of Gender Separate Education on Middle School Students Heather O’Neill, Ursinus College Assessing the Impact of Family Aid on College Persistence Sarah Jackson, Indiana University of Pennsylvania

Educational Achievement of Children of Immigrants: Evidence from the National Longitudinal Survey of Youth Daniel Lee, Shippensburg University Discussants: Sarah Jackson, Indiana University of Pennsylvania Pavani Tallapally, Slippery Rock University Lei Zhu, West Chester University

Session F1E: General Economics and Teaching Room: 204

Chair: David Culp, Slippery Rock University

Examining the Existence and Extent of Anti-Intellectual Attitudes among University Students Robert Balough, Clarion University of Pennsylvania Recent Federal Reserve Monetary and Credit Policy James Dunn, Edinboro University of Pennsylvania

Simulating Money Creation in the Principles Classroom Sharon May, Maryville College Discussants: Sharon May, Maryville College Robert Balough, Clarion University of Pennsylvania James Dunn, Edinboro University of Pennsylvania


Session F1F: Other Special Topics Room: 205

Chair: Steven Andelin, Pennsylvania State University Do Casinos Cannibalize Lottery Revenues? Early Evidence from Pennsylvania Andrew Economopoulos, Ursinus College Gambling: Prevalence, Intensity and Comorbidity Simon Condliffe, West Chester University

Fund Assortments and 401(k) Plan Participation: The Moderating Effect of Gender Maureen Morrin, Rutgers University Discussants: Andrew Hill, Federal Reserve Bank of Philadelphia Bonnie Meszaros, University of Delaware Andrew Economopoulos, Ursinus College

FRIDAY, June 5, 2009 10:15 a.m. – 10:30 a.m.

Coffee Break: Third Floor Lobby

FRIDAY, June 5, 2009 10:30 a.m. – 11:45 a.m.

Session F2A: Economic Development Room: 305

Chair: Thomas Armstrong, Pennsylvania Department of Community & Economic Development Urban-Rural Inequality in China and Mongolia: Economic vs. Subjective Divides Kenneth Smith, Millersville University Human Capital, Returning Migration, and Rural Entrepreneurship in China Jialu Liu, Indiana University

The Effects of Industrial Development on Migration: Evidence from Mexico in the Post-NAFTA Period

Peter Schnabl, University of Delaware Discussants: Peter Schnabl, University of Delaware Kenneth Smith, Millersville University Jialu Liu, Indiana University


Session F2B: General Economics and Teaching Room: 306

Chair: David Nugent, Slippery Rock University

Demand Elasticity Simplified without the Twist Thomas Andrews, West Chester University Exercises in Economics

Gandhi Veluri, Andhra University & Vani V. Kotcherlakota, University of Nebraska at Kearney

Metaphor and Economic Science II

Robert H. Renshaw, Renshaw Enterprises Discussants: Gandhi Veluri, Andhra University Thomas Andrews, West Chester University David Nugent, Slippery Rock University

Session F2C: Miscellaneous Topics

Room: 312 Chair: Mark Eschenfelder, Robert Morris University

Key Sector Analysis of Industrial Clusters: A Northeast Ohio Application Jolien Helsel, Youngstown State University

The Study of Vietnamese Small Business Owners in America: Their Patterns of Strategic Decisions

Lei Zhu, West Chester University

Does the Immigrant-Trade Link Vary Across Migration Corridors? Roger White, Franklin & Marshall College Discussants: Roger White, Franklin & Marshall College Jolien Helsel, Youngstown State University Lei Zhu, West Chester University


Session F2D: General Economics and Teaching Room: 320

Chair: Frederick Tannery, Slippery Rock University Evidence of Student Achievement in a High School Personal Finance Course

Andrew Hill, Federal Reserve Bank of Philadelphia & Bonnie Meszaros, University of Delaware

The Effect of Gender on Learning and Success in Economics Classes Orhan Kara, West Chester University & I-Ming Chiu, Rutgers University-Camden

Some Major Campaign Issues of 2008: An Economic Perspective; A Student-Faculty Project William Sanders, Clarion University Discussants: I-Ming Chiu, Rutgers University-Camden Bonnie Meszaros, University of Delaware Frederick Tannery, Slippery Rock University

Session F2E: Mathematical and Quantitative Models Room: 204

Chair: Stephanie Brewer, Indiana University of Pennsylvania Volatility, Its Measurement, and the Great Moderation Steven Andelin, Pennsylvania State University

Katyusha Computations or Quantitative Management: Using Economic Order Quantity to Assess Sustainability of Hostile Rocket Offensives

Johnnie Linn, Concord University

Discussants: Johnnie Linn, Concord University Steven Andelin, Pennsylvania State University


Session F2F: Public Economics Room: 205

Chair: Lynn Smith, Clarion University How Equal is Equal? A Study of New Jersey Public School Resources Thomas Tolin, West Chester University Gasoline Policy and Pricing Versus Sustainable Development Sadegh Bafandeh Imandoust, Payame Noor University

The Voluntary Provision of a Pure Public Good with Matching Funds: Further Experimental Results Ronald Baker, Millersville University & James M. Walker, Indiana University Discussants: Ronald Baker, Millersville University Lynn Smith, Clarion University Thomas Tolin, West Chester University

LUNCHEON AND SPEAKER

12:00 Noon – 2:00 p.m.

WCU Graduate Business Center: Room 126

“Searching for a Bottom…and Considering the Recovery”

Friday Luncheon Speaker: Mark Zandi, Ph.D.

Chief Economist Moody's Economy.com

Dr. Zandi received his Ph.D. at the University of Pennsylvania, where he did his research with Gerard Adams and Nobel Laureate Lawrence Klein, and he received his B.S. from the Wharton School at the University of Pennsylvania. Dr. Zandi is Chief Economist and co-founder of Moody's Economy.com, where he directs the company's research and consulting activities. Moody's Economy.com, a division of Moody's Analytics, provides economic research and consulting services to businesses, governments and other institutions. Dr. Zandi's research interests include macroeconomic, financial, and regional economics. Recent areas of research include studying the determinants of mortgage foreclosure and personal bankruptcy, an analysis of the economic impact of various tax and government spending policies, and an assessment of the appropriate policy response to bubbles in asset markets. In addition, Dr. Zandi conducts regular briefings on the economy and is frequently quoted in national and global news outlets.

http://www.economy.com/mark-zandi/default.asp?src=economy_homepage

http://www.economy.com/mark-zandi/default.asp?src=economy_homepage


FRIDAY, June 5, 2009 2:15 p.m. – 3:30 p.m.

Session F3A: Financial Economics Room: 305

Chair: Steven Andelin, Pennsylvania State University Asset Prices and Monetary Policy: A Taylor Rule Approach Michael Hannan & James Dunn, Edinboro University of PA Valuation of a Portfolio of Mortgages in Continuous Time Riaz Hussain & Stephen Mansour, University of Scranton

Bond Rating Models: Do Data Sources Matter? Pavani Tallapally, Slippery Rock University Discussants: Riaz Hussain, University of Scranton Pavani Tallapally, Slippery Rock University Michael Hannan, Edinboro University of PA

Session F3B: Law and Economics

Room: 306 Chair: Tracy Miller, Grove City College

Drug Approvals and Drug Safety Natalie Reaves, Rowan University

Effect of Income Taxes on Charitable Contributions: Evidence from Tax Law Change, EGTRRA Arpita Biswas, Clemson University Discussants: Tracy Miller, Grove City College David Nugent, Slippery Rock University


Session F3C: Urban, Rural and Regional Economics Room: 312

Chair: James Jozefowicz, Indiana University of Pennsylvania

Consumption Risk-Sharing: Evidence from German Household Data D. Teja Flotho, IEWE: University of Freiburg

Are Smaller Inner Cities Systematically Underserved by Retail? William Bellinger, Dickinson College

Factors Impacting the Throwaway Society: A Sustainable Consumption Issue John McCollough, Pennsylvania State University-Lehigh Valley Discussants:

James Jozefowicz, Indiana University of Pennsylvania John McCollough, Pennsylvania State University-Lehigh Valley William Bellinger, Dickinson College

Session F3D: Labor and Demographic Economics

Room: 320 Chair: Lynn Smith, Clarion University

Where Have all the Young Men and Women Gone? Labor Force Participation of 16-19 Year-olds

Christopher Maida & Frederick Tannery, Slippery Rock University Wage Growth and Employment Growth: Evidence from Wage Records Frederick Tannery, Slippery Rock University

Stats or Studs? Does it Pay to be Good Looking? The Economic Impact of Lookism Jennifer VanGilder, Ursinus College Discussants: Jennifer VanGilder, Ursinus College Lynn Smith, Clarion University Frederick Tannery, Slippery Rock University


Session F3E: Macroeconomics and Monetary Economics Room: 204

Chair: Thomas Tolin, West Chester University Optimum Currency Areas and Synchronization of Business Cycles in Sub-Saharan Africa Yaya Sissoko, Indiana University of PA

Once Bitten, Twice Shy: The Effect of a Banking Crisis on Expectations of Future Crises Shannon Mudd, Ursinus College

Does Consumer Sentiment Affect Savings Behavior? I-Ming Chiu, Rutgers University-Camden Discussants: I-Ming Chiu, Rutgers University-Camden Yaya Sissoko, Indiana University of PA Shannon Mudd, Ursinus College

Session F3F: Miscellaneous Topics

Room: 205 Chair: William Sanders, Clarion University

What Do College and University Faculty Know about Assurance of Learning Techniques?

Mark Eschenfelder, Lois Bryan, & Tanya Lee, Robert Morris University Optimizing University Summer Course Offerings under a Fixed Price Constraint David Culp, Slippery Rock University

Discussants: David Culp, Slippery Rock University Mark Eschenfelder, Robert Morris University

Session F3G: Student Posters

Third Floor Lounge

The Effects of Urbanization on Development: Analysis of Latin America and Sub-Saharan Africa Chelsea Kaufman, Clarion University Analysis of Real GDP, Term Structure, and US Housing Steven Caplan, Philadelphia University

Use of Mathematics in Economics Madhavi Latha Dasari, Sai Adharsha IIT Concept School


FRIDAY, June 5, 2009 3:30 p.m. – 3:45 p.m.

Coffee Break: Third Floor Lobby

FRIDAY, June 5, 2009 3:45 p.m. – 4:45 p.m.

FFeeddeerraall RReesseerrvvee BBaannkk ooff PPhhiillaaddeellpphhiiaa

PPaanneell DDiissccuussssiioonn Rooms 325-326

"Teaching Monetary Policy in Turbulent Times” -- Andrew Hill

"Questions about the Economic Outlook and Monetary Policy” -- Rick Lang

FRIDAY, June 5, 2009 4:45 p.m. – 5:45 p.m.

Reception hosted by the Federal Reserve Bank of Philadelphia

Room 126

SATURDAY, June 6, 2009 8:00 a.m. – 9:00 a.m.

Conference Registration & Coffee WCU Graduate Business Center (Third Floor Lobby)



Session S1A: Miscellaneous Topics Room: 305

Chair: David Culp, Slippery Rock University 21st Century Capitalism through the Lens of Thomas Hobbes and Adam Smith Robert S. D’Intino, Rowan University & John A. Sinisi, Pennsylvania State University

Possible Relationships Between the Theory of Global Warming and Stock Market Declines David Nugent, Slippery Rock University If “We are all Socialists Now,” Hadn’t We Better Figure Out How to do it Right? Roger McCain, Drexel University Discussants: David Nugent, Slippery Rock University Roger McCain, Drexel University Robert S. D’Intino, Rowan University

Session S1B: Macroeconomics / Environmental Economics Room: 306

Chair: Michael Hannan, Edinboro University of Pennsylvania

On the Predictability of Inflation in Ghana Eric Fosu Oteng-Abayie & Joseph Magnus Frimpong, Knust School of Business

The Impact and Length of the Exchange Rate Fluctuations on U.S. Imported Crude Oil Pricing: An Exchange Rate Pass-Through and VAR Model Analysis

Jui-Chi Huang, Penn State Berks & Yaya Sissoko, Indiana University of Pennsylvania

Economic Valuation of Forests: Iranian Case Study Sadegh Bafandeh Imandost, Payame Noor University Discussants: Michael Hannan, Edinboro University of Pennsylvania Sadegh Bafandeh Imandost, Payame Noor University Jui-Chi Huang, Penn State Berks


Session S1C: Financial Economics Room: 312

Chair: James Dunn, Edinboro University of Pennsylvania

Bank’s Strategic Perspective: From Financial Performance to Strategic Performance Walid Khoufi & Jamel Choukir, University of Sfax Convertibles or Stock: A Cash Flow Comparison Model with Credit Risk Luca Del Viva, University of Pisa

Cost of Capital Estimation with Stock Return Outliers: The Case of U.S. Pharmaceutical Companies Uzi Yaari, Alexandra Theodossiou and Panayiotis Theodossiou, Rutgers University-Camden

Discussants: Luca Del Viva, University of Pisa Uzi Yaari, Rutgers University-Camden Walid Khoufi, University of Sfax

Session S1D: Health Education and Welfare

Room: 320 Chair: Stephanie Brewer, Indiana University of Pennsylvania

A Time-Series Analysis of the College Enrollment Rate of White, Black, and Hispanic Youth

Kazuhisa Matsuda, Ohio Northern University Public versus Private: A Dynamic Model of Health Insurance Choice Cristian Pardo, Saint Joseph’s University

Discussants: Stephanie Brewer, Indiana University of Pennsylvania Kazuhisa Matsuda, Ohio Northern University

Session S1E: Miscellaneous Topics Room: 205

Chair: Jolien Helsel, Youngstown State University The Interactional Basis of Transaction Costs: A Common Issue in Outsourcing Jacqueline Zalewski, West Chester University Reclaiming Institutions as a Form of Capital Benedique Paul, University of Montpellier

Intellectual Property, Rents, and Economic Growth Cynthia Cohen Discussants: Benedique Paul, University of Montpellier D. Teja Flotho, IEWE: University of Freiburg Natalie Reaves, Rowan University



GENERAL MEMBERSHIP BUSINESS MEETING

Room 126

This Annual Business Meeting of the General Membership of the Pennsylvania Economic Association is open to the entire membership of the PEA, including all registrants at the conference. Please plan to attend as door prizes are available.

SATURDAY, June 6, 2009, 11:15 a.m. - Closing

There is no formal closing session, but conference participants are welcome, and encouraged, after the last set of paper sessions, to stay and chat as long as you wish and thank you for helping to make this year’s conference a success. Pre-ordered box lunches will be available for pick-up.

Proceedings of the Pennsylvania Economic Association 2009 Conference

15

Program Author/Participant Index

Name Sessions E-mail

Aka, Arsene [email protected]

Andelin, Steven F1F, F2E, F3A [email protected]

Andrews, Thomas F2B [email protected]

Armstrong, Thomas F1C, F2A [email protected]

Ayllon, Antonio F1B [email protected]

Baker, Ronald F2F [email protected]

Balough, Robert F1E [email protected]

Bearjar, Stephanie F1B [email protected]

Bellinger, William F3C [email protected]

Biswas, Arpita F1B, F3B [email protected]

Brewer, Stephanie F2E, S1D [email protected]

Bryan, Lois 3F3 [email protected]

Caplan, Steven F3G [email protected]

Chiu, I-Ming F2D, F3E [email protected]

Choukir, Jamel S1C [email protected]

Cohen, Cynthia S1E [email protected]

Condliffe, Simon F1F [email protected]

Culp, David F1E, F3F, S1A [email protected]

Dasari, Madhavi Latha F3G [email protected]

Del Viva, Luca S1C [email protected]

D’Intino, Robert S1A [email protected]

Dunn, James F1E, F3A, S1C [email protected]
























16

Economopoulos, Andrew F1F [email protected]

Eschenfelder, Mark F2C, F3F [email protected]

Flotho, D. Teja F3C, S1E [email protected]

Frickel, Beverly F1A [email protected]

Frimpong, Joseph Magnus S1B [email protected]

Haflett, Aleta F1B [email protected]

Hannan, Michael F1C, F3A, S1B [email protected]

Helsel, Jolien F2C, S1E [email protected]

Hill, Andrew F1F, F2D [email protected]

Huang, Jui-Chi S1B [email protected]

Hussain, Riaz F3A [email protected]

Imandoust, Sadegh F2F, S1B [email protected]

Jackson, Sarah F1D [email protected]

Jozefowicz, James F1B, F3C [email protected]

Kara, Orhan F2D [email protected]

Kaufman, Chelsea F3G [email protected]

Khoufi, Walid S1C [email protected]

Kotcherlakota, Vani F1A, F2B [email protected]

Lang, Rick [email protected]

Lee, Daniel F1D [email protected]

Lee, Tanya F3F [email protected]

Linn, Johnnie F2E [email protected]

Liu, Jialu F1B, F2A [email protected]

Maida, Christopher F3D [email protected]


























17

Mansour, Stephen F3A [email protected]

Matsuda, Kazuhisa S1D [email protected]

May, Sharon F1E [email protected]

McCain, Roger S1A [email protected]

McCollough, John F3C [email protected]

Meszaros, Bonnie F1F, F2D [email protected]

Miller, Tracy F1C, F3B [email protected]

Morrin, Maureen F1F [email protected]

Mudd, Shannon F3E [email protected]

Nugent, David F2B, F3B, S1A [email protected]

O’Neill, Heather F1D [email protected]

Oteng-Abayie, Eric Fosu S1B [email protected]

Pardo, Cristian S1D [email protected]

Paul, Benedique F1B, S1E [email protected]

Ratkus, Mark [email protected]

Reaves, Natalie F3B, S1E [email protected]

Renshaw, Robert H. F2B [email protected]

Sanders, William F2D, F3F [email protected]

Schnabl, Peter F2A [email protected]

Sinisi, John S1A [email protected]

Sissoko, Yaya F1A, F3E, S1B [email protected]

Smith, Kenneth F1A, F2A [email protected]

Smith, Lynn F2F, F3D [email protected]

Snyder, Kristopher [email protected]


























18

Tallapally, Pavani F1D, F3A [email protected]

Tannery, Frederick F2D, F3D [email protected]

Theodossiou, Alexandra S1C [email protected]

Tolin, Thomas F1D, F2F, F3E [email protected]

VanGilder, Jennifer F3D [email protected]

Veluri, Gandhi F2B [email protected]

White, Roger F1A, F2C [email protected]

Yaari, Uzi S1C [email protected]

Yerger, David F1A [email protected]

Zalewski, Jacqueline S1E [email protected]

Zhu, Lei F1D, F2C [email protected]













19

CORRELATION BETWEEN EXCHANGE RATES AND TRADE AND INVESTMENT VARIABLES

Beverly J. Frickel, Ph.D. Department of Accounting & Finance

WSTC 300C University of Nebraska at Kearney

Kearney, NE 68849

Vani V. Kotcherlakota, Ph.D. Department of Economics


Kearney, NE 68849

Frank A. Tenkorang, Ph.D. Department of Economics


Kearney, NE 68849

ABSTRACT This paper examines the relationships between real effective exchange rates and trade variables and investment variables for the U.S. from 1996 to 2007. The real effective exchange rate is found to have a negative but weak relationship to the trade variables and a negative but weak relationship with respect to direct investment abroad.

INTRODUCTION

The main objective of the paper is to analyze the relationships between real effective exchange rates and 1) trade variables and 2) investment variables for United States data for the period 1996-2007. The paper is outlined in four sections. In the first section the foreign exchange markets, its functions and exchange rate determination are given. A brief review of literature pertaining to the topic is presented in the second section. Definitions of variables, data sources, and methodology are presented in the third section. Analysis of the relationships between the real effective exchange rate and the selected trade and investment variables forms the core of the fourth section.

FOREIGN EXCHANGE MARKET A foreign exchange market (FEM) refers to the organizational setting with which individuals, businesses, governments, and banks buy and sell foreign currencies and other department instruments. The foreign exchange market is merely a part of the money market in the financial centers in which foreign currencies are bought and sold and is not restricted to any given country or a geographical area. Not all currencies, however, are traded in FEMs. Reasons for no

trading in a particular currency range from political instability to economic uncertainty. The basic function of the foreign exchange market is to facilitate the conversion of one currency into another, i.e., to accomplish transfers of purchasing power between two countries. This transfer of purchasing power is affected through a variety of credit instruments such as telegraphic transfers, bank drafts and foreign bills. The foreign exchange market also provides credit to other national and international countries to promote foreign trade. Another function of foreign exchange markets is the offset of risk through hedging. Hedging applies to anyone who is obligated to make a foreign currency payment or who will enjoy foreign currency receipt at a future time. Exporters hedge against the possibility of foreign currency depreciating against the domestic currency and importers hedge against the possibility of foreign currency appreciating against the domestic currency. International investors make use of foreign exchange markets to hedge an offsetting position or counterweight to any exposure to risk. A geographical hedge is simply geographical diversification of currency holdings in such way that offsets occur across the currencies. It is crucial that currencies held in the hedged portfolio are not correlated in their exchange values. Foreign exchange transactions also take place through interbank trading with banks buying and selling foreign currencies for their customers. Retail transactions are defined as those which are less than one million currency units and these are bank purchases from sales to customers. Wholesale transactions are when the amount is more than one


20

million currency units and occurs between banks with large corporate customers. Banks earn profits in foreign exchange transactions. Banks that regularly deal in the interbank market quote both a bid and an offer rate to other banks. The bid rate refers to the rate at which the bank is willing to sell a unit of foreign currency and the offer rate is the rate at which the bank is willing to buy. The spread is the difference between the bid and the offer rate and at any given time a bank’s bid quote is less than its offer quote. The spread is intended to cover the bank’s costs of implementing the exchange of currencies and foreign exchange dealers who simultaneously purchase and sell foreign currency earn the spread as profit. Currency exchange transactions in the cash or spot market have immediate payment and delivery. Rather than effect an immediate trade, a forward contract is an agreement now for a transaction or trade that is to take place at some specified future time at a specified price with the payment occurring at maturity or settlement. The actual payment and delivery can be a few days, months, or even years in the future. Standardizing the details of forward contracts, such as amounts and delivery dates, facilitates trading of the contracts in an organized market. The CME Group lays claim to the largest regulated foreign exchange marketplace and offers both electronic trading and open outcry trading floors. The various CME Group foreign exchange futures contracts cover 20 currencies including the G10 currencies as well as currencies of other countries such as Brazil, China, and South Africa. With a few exceptions, the CME Group foreign exchange futures contracts mature on the third Wednesday of March, June, September and December. The foreign exchange rate refers to the price of one currency in terms of another currency. Assume that the U.S. is the home country and the U.K. is the foreign country. Also suppose that $2 exchanges for ₤1. There are two ways of quoting or measuring the foreign exchange rate: a) The direct exchange rate quote is the price or value

of foreign currency per unit of domestic currency or ₤.50 per dollar.

b) The indirect exchange rate quote is the price or value of domestic currency per unit of foreign currency or $2.00 per pound.

Since the two exchange rate quotes are reciprocals of each other it is inconsequential whether the direct or indirect quote is used as long as the version selected is used consistently whenever a quote is applied. Under the present international monetary system, nations have a great degree of flexibility in choosing the exchange rate regimes that best fit their economic conditions. At one extreme, is a fixed exchange rate system, where the exchange

rate is stable and can only fluctuate within a very limited band. At the other extreme is a freely fluctuating exchange rate system where the exchange rate is set by the free market forces of demand and supply. The freely fluctuating exchange rate system is also referred to as a freely floating exchange rate system, a flexible exchange rate system, or a clean-floating exchange rate system.

REVIEW OF LITERATURE Many studies have focused on the relationship among exchange rate, trade variables, and investment variables. In this section a brief review of the literature is given. The review is not intended as an exhaustive list and is limited in scope to studies pertaining to analysis of exchange rates published in the previous decade. Goldberg (2004) constructs three industry specific real exchange rate indexes for the U.S. and looks at the extent to which each index co-moves or diverges from aggregate economy wide measures. The article’s intent is to provide tools for analyzing the real and financial effects of exchange rate movements. Belts and Kehoe (2006) examine the relationship between the bilateral real exchange rate of the U.S. and the associated bilateral relative price of non traded goods for five of its trade relationships. They find that this relationship depends crucially on the choice of the price series used to measure relative prices and the choice of the trade partner. Additionally, the relationship is stronger when producer prices are used as a measure of relative prices and when the trade relationship between U.S. and trade partner is more important. Liew, Lim, and Hussain (2003) examine the impact of exchange rate changes on trade balances between Japan and the ASEAN-5 countries of Indonesia, Malaysia, the Philippines, Singapore, and Thailand. They conclude that the role of exchange rates on trade balances has been exaggerated. Their study indicates that real money, more so than the nominal exchange rate, affected trade balances during the sample period of 1986-1999. Kakkar and Ogaki (1999) examine the long run movements of real exchange rates and relative prices of non-tradable's and tradable's. They conclude that real exchange rates may not move in the direction predicted by theoretical models when the producers of these tradable goods experience changes in productivity. A study by Tille (2003) emphasizes the impact of exchange rate movements on U.S. foreign debt. The conclusion from the study is that the depreciation of the dollar would improve the U.S. position by making exports more competitive and

thereby reduce U.S. reliance on foreign financial flows. A valuation effect is examined in the article that provides an additional mechanism through which the current depreciation of the dollars might improve the U.S. net position. Wilamoski and Tinkler (1999) examined the effect of U.S. foreign direct investment in Mexico on U.S. trade with Mexico. Their analysis shows that foreign direct investment has positive impact on exports to and imports from Mexico. Hejazi and Safarian (2003) use a gravity model for 51 countries and confirm the Wilamoski and Tinkler (1999) results. Hejazi and Safarian also found that the impact of U.S. foreign direct investment is stronger on exports than imports. Hence, foreign direct investment may increase trade surplus or decrease traded deficit. Goldberg (2009) reviews several studies on the relationship between exchange rates and foreign direct investment. She points out that research supports the view that the variability of exchange rates contributes to foreign investment. She further summarizes explanations given as to why a country with depreciating currency should expect to experience an increase in foreign investment in that country. From above, the relationship among exchange rate, trade, and investment abroad can be summarized as follows: a. There is a positive relationship between exchange

rate and exports and imports. However, the impact on balance of trade depends on which relationship is stronger. This is usually negative.

b. There is a positive relationship between balance of trade and foreign investment abroad.

The implication of a. and b. is that appreciation of the US dollar will likely have a negative impact on foreign investment abroad.

DATA AND METHODOLOGY In this section the data and methodology used for the study are given. The main data source is the International Financial Statistics Yearbook (2008) published by the International Monetary Fund. The trade variables selected are Real Exports (RX), Real Imports (RM), Real Balance of Trade (BOT), and Real Effective Exchange Rate (RER), while the investment related variables are Direct Investment Abroad (DI) and Portfolio Investment (PI). The choice of the variables is based on the literature and the availability of continuous data for the sample years. To examine the relationship between RER and the trade related variables, a double logarithmic model as given in Equation 1 was fitted for each of the trade variables:

11-t2t10t μ R Logα RER Logαα R Log +++= (1)

Where: Rt = real exports, real imports, or real balance of trade.

Rt-1 = previous year’s value of Rt µ = error term αi = coefficients t = time: 1996 to 2007 However, taken logs of negative BOT values will yield complex numbers; hence the BOT model was estimated using the data in its level form. A model similar to Equation 1 is used to examine the relationship between RER and the investment related variables, Direct Investment Abroad (DI) and Portfolio Investment (PI). The dependent variable is the investment variable It, which is either direct investment abroad (DI) or portfolio investment (PI).

21-t2t10t I Log RER Log I Log μβββ +++= (2) Where: β1 = coefficients

Each model is estimated using ordinary least squares (OLS) method.

Elasticity

The relationship between exchange rates and international trade is affected by the “J-curve” phenomenon. The J-curve shows the initial decline in export value and increase in import value due to exchange rate depreciation but, in the long run, the trade balance increases because the depreciation stimulates exports and discourages imports. To capture the difference between the short run and long relationship between RER and the trade related variables, short run and long run elasticities are estimated. The coefficients of Log RER in the models are the short run elasticities of RER. Long run elasticities are obtained using the equations below:

)1( 2

1L1 α

αε

−= (3)

Where: the coefficients are from Equation 1 following Nerlove and Addison (1958) and Koutsoyiannis (1977).

)1( 2

1L2 β

βε

−= (4)

Where: the coefficients are from Equation 2.

For the BOT model, the short run elasticity was calculated using the usual elasticity formula

BOT

RER α * 1 φ=

φ is the coefficient of RER;

(5)

Where: RER is the average of RER BOT is the average of BOT.



22

RESULTS AND ANALYSIS Table 1 shows the results for the regression analysis of the relationship between real effective exchange rates (RER) and the trade related variables Exports (RX), Imports (RM), Real Balance of Trade (BOT). The three models are all generally of good fit with statistically significant F-values and R-squares above 90 percent. The presence of a lagged dependent variable does raise autocorrelation concerns. Only the RX model, however exhibited the autocorrelation problem. Hence, the results shown for the trade variable Real Exports (RX) are Yule-Walker estimates corrected for autocorrelation. The results shown for Real Imports (RM) and Real Balance of Trade (BOT) are OLS estimates. As the results in Tables 1 and 2 indicate, the intercept is significant for Real Exports (RX) and Real Imports (RM) but not for Real Balance of Trade (BOT). The RER coefficients are not statistically significantly different from zero, except that of log RX. The literature has indicated weak relationships between trade variables and exchange rates. The short run elasticities (coefficient of log RER in Table 1) are negative for all three trade variables. This implies that, in the short run, as the U.S. dollar appreciates, real exports and real imports decline but the former declines faster subsequently causing BOT to decline. While the impact of RER on RX and RM is more elastic in the long run than in the short run, the reverse holds true for BOT. Hence, the J-curve effect is not observed in these results. Table 3 presents the results obtained by regressing the investment related variables Direct Investment Abroad (DI) and Portfolio Investment (PI) on RER. Similar to the trade related variable results, the F values and adjusted R-squares are significantly high. The Log DI results are Yule-Walker estimates to correct for autocorrelation. The results show

RER negatively related to both Direct Investment Abroad (DI) and Portfolio Investment (PI) with only the effect of RER on PI being statistically significant. The coefficients of lagged values of both DI and PI are positive and statistically significant. In the case of DI the short run elasticity is less than one and in the case of PI it is greater than one. The long run elasticities for both DI and PI are significantly higher.

CONCLUSIONS This paper examines the relationships between real effective exchange rates and trade variables and investment variables for the U.S. from 1996 to 2007. The trade related variables considered are imports, exports, and the balance of trade. The real effective exchange rate is found to have a negative but weak relationship to the trade variables. Regarding the investment related variables, the relationship is negative but weak with respect to direct investment abroad. The elasticity is less than one for direct investment abroad and greater than one for portfolio investment. In all cases, except BOT, the long run relationship is stronger than the short run as indicated by the higher elasticities. The results of the trade variables and investment variables together confirm the literature summary above. This study has limited scope and is preliminary in nature. As this study is a pilot study, future research with more observations is needed. Additional considerations such as a time trend and other factors noted in the literature such as real money are also considerations for future research.

Table 1. Relationship between Real Effective Exchange Rate and Trade Related Variables

Dependent variable

Intercept Coefficient

of Log RERt

Coefficient of lagged Dep.

Var. Adj. R2 F value Durbin

h Long run elasticity

Log RX (Exports) 1.016* -0.434* 0.933* 0.952 58.47* -1.689* -6.478 (6.19) (-6.3) (11.51) (0.0001) (0.0045) Log RM (Imports) 0.598* -0.207 0.847* 0.967 147* -1.16 -1.353 (2.18) (-1.06) (16.04) (0.0001) (0.123)

* Statistically significant at 5 percent test level Values in parenthesis: -- Coefficients: T values -- F and Durbin h: P values


23

Table 2. BOT Estimation Results

Intercept

RERt

BOTt-1

Adj. R2 F value Durbin h

Short run elasticity A

Long run elasticity

-3.40 0.396 1.038* 0.941 80.19* -0.343 -0.638 -0.225 (-1.66) (1.24) (7.80) (0.0001) (0.366)

* Statistically significant at 5 percent test level A. – computed using equation 5 Values in parenthesis: -- Coefficients: T values -- F and Durbin h: P values

Table 3. Relationship between Real Effective Exchange Rate and Investment Related Variables

Dependent variable

Intercept Coefficient of Log RERt

Coefficient of lagged Dep.

Var. Adj. R2 F value Durbin

h Long run elasticity

Log DI (Direct Investment Abroad) 0.233 -0.067 0.983* 0.987 196.3* -1.76* -3.94 (0.52) (-0.34) (19.26) (0.0001) (0.039) Log PI (Portfolio Investment) 2.94* -1.32* 0.916* 0.880 34.1* -0.091 -15.71 (2.18) (-2.34) (6.59) (0.0002) (0.434)

* Statistically significant at 5 percent test level Values in parenthesis: -- Coefficients: T values -- F and Durbin h: P values

REFERENCES

Betts, Caroline M.; Kehoe. T.J. (March 2004; revised May 2005) “U.S. Real Exchange Rate Fluctuations and Relative Price Fluctuations.” Federal Reserve Bank of Minneapolis Research Department Staff Report 334. CME Group. (2009). “FX Products; 2009 Product Guide and Calendar.” Retrieved June 1, 2009 from http://www.cmegroup.com/fx.html Goldberg, Linda S. (May 2004) “Industry-Specific Exchange Rates for the United States”. Federal Reserve Bank of New York Economic Policy Review. Goldberg, L. (2009) “Exchange Rates and Foreign Direct Investment.” Kenneth A. Reinert and Ramkishen S. Rajan, eds., The Princeton Encyclopedia of the World Economy. Princeton, N.J. Princeton University Press.

Hejazi, W. and A.E. Safarian (2001). “The Complementarity Between U.S. Foreign Direct Investment Stock and Trade.” Atlantic Economic Journal. Dec, 2001. Kakkar, Vikas; Ogaki, M. (1999). “Real exchange rates and nontradables: A relative price approach.”Journal of Empirical Finance Vol. 6, Issue 2, p. 193-215. Koutsoyiannis, A. (1977). Theory of Econometrics. An Introductory Exposition of Econometric Methods, 2nd edition. Macmillan. Liew, V. K-S., K-P. Lim and H. Hussain (2003). “Exchange Rate and Trade Balance Relationship Experience of ASEAN Countries. “ EconPapers: International Trade, JEL-codes: F31. Retrieved June 1, 2009 from http://econpapers.repec.org/paper/wpawuwpit/0307003.html

http://www.cmegroup.com/fx.html

http://econpapers.repec.org/paper/wpawuwpit/0307003.html


24

Nerlove, M. and W. Addison (1958). “Statistical Estimation of Long-run Elasticities of Supply and Demand,” Journal of Farm Economics, XL4, p 861-80. Tille, Cedric (Jan 2003). “The Impact of Exchange Rate Movements on U.S. Foreign Debt.” Federal Reserve Bank of

New York Current Issues in Economics and Finance. Vol. 9, No. 1. Wilamoski, R. and S. Tinkler (1999). “The Trade Balance Effects of U.S. Foreign Direct Investment in Mexico.” Atlantic Economic Journal, Vol. 27, No. 1, p 24-37.


25

TECHNOLOGY TRANSFER AND THE KEYSTONE INNOVATION GRANT INITIATIVE

Thomas O. Armstrong* University Liaison/Nanotechnology Program Manager

Technology Investment Office Pennsylvania Department of Community and Economic Development

Harrisburg, PA 17120

ABSTRACT The Pennsylvania Keystone Innovation Grant (KIG) initiative provides financial resources for the technology transfer process where research ideas from Institutions of Higher Education (IHEs) can increase the likelihood of an intellectual property (IP) transfer to licensing arrangements with existing or new companies leading to improved or new commercialized products in the Commonwealth. Within Pennsylvania’s IHEs, 96% do not have a technology transfer office (TTO), which severely limits technology transfer, commercialization and economic development possibilities; while 4% do have a TTO, but are faced with financial challenges, which again limits technology transfer and economic development. The competitive KIG awards were created to enhance the transfer process at both types of institutions--IHEs with and without a technology transfer office. The technology transfer process among IHEs include providing resources for research; incentivizing the generation of IP; creating IP policies; assessing and providing IP protections; executing technical licenses to either established companies or start-ups; utilizing license income for greater economic development to the region, state and the IHE; and, finally, continuing the technology transfer process of discoveries to commercialization.

INTRODUCTION Technology-based economic development (TBED) strategies have become a greater priority among states over the last decade so as to enhance state and regional economic development (Douglass, 2006; Geiger and Sa’, 2005). Included within a TBED strategy is the augmentation of a technology transfer infrastructure for leveraging Institutions of Higher Education (IHEs) research and intellectual property (IP) to promote commercialization (Armstrong, et al., 2007; CEO Council for Growth, October 2007; Resource Guide, 2006; Varga, August 2002). The results of an effective infrastructure, sometimes formalized through a technology transfer office (TTO), can potentially generate revenue for an academic institution, build relationships with external stakeholders, increase additional sponsored research, hire more graduate students and post-doctoral fellows, and enhance economic growth and development (Phan and Siegel, 2006; Link, Siegel and Bozeman, May 2006).

One of the key federal policy decisions that impacted IHEs, technology transfer, and economic development was the enactment of Bayh-Dole Act (Act) in 1980. This Act instituted a federal agency uniform patent policy, removal of licensing restrictions, and allowed IHEs to own patents arising from federal research grants.1,2 The primary rationale for the passage of the Act by having IHEs owning IP created with federally funded research is for IHEs to be obligated to file patent applications on those inventions and to seek, hence accelerate, commercialization of new technologies. As a result, a significant number of TTOs nationally were established among universities to manage and protect IP. The four functions of TTOs are to: (1) receive invention disclosures from IHE’s faculty, staff or students; (2) assess inventions for possible steps to commercialization; (3) conduct IP protections; and (4) perform licensing phases from negotiation to enforcement. TTOs embody a formal transfer mechanism focused on the allocation of IP rights that result in a legal instrument such as a patent or license agreement. The number of TTOs increased to at least 194 across the United States with eight established TTOs at Pennsylvania Class Type 15 (very high research universities) and Class Type 16 (high research universities) Carnegie Classified Institutions.3 They are:

• Type 15: Carnegie Mellon University, Pennsylvania State University-Main Campus, University of Pennsylvania, University of Pittsburgh-Main Campus

• Type 16: Drexel University, Lehigh University, Temple University, and Duquesne University.

Technology transfer of IP allows businesses to enhance characteristics of existing products or hedonics demanded by consumers.4 This technology transfer process, which can be facilitated with or without a TTO, provides attributes between a given quantity of a product and its enhanced technical characteristics where individuals possess preferences for collections of these characteristics (Eastwood, Brooker, and Terry, December 1986; Armstrong, Spring 2008; Triplett, 2004). The result is product preferences are valued because they provide the characteristics sought by consumers.

One of the primary goals of a TTO is to transition the academic innovation via IP management to consumer end-users through existing or new firm production. IP management provides incentives for research and development, commercialization and product distribution contributing to local economic development.5 As a result of TTOs facilitated technology transfer, firms pursue greater production. It is expected that sustainable economic development of a region and state is enhanced where an IHE TTO facilitated technology transfer, industry research, and firm production occurs to satisfy consumer demand. Section II of this paper provides a summary of an IHE technology transfer infrastructure or process where IHE’s may or may not have a formal TTO to facilitate the transfer of technology. Summary data from the Association of University Technology Managers (AUTM) indicates the success of the Commonwealth’s transfer of technology process—but also certain gaps within the process. These gaps are a function of a lack of capacity along the spectrum of IHE research to licensing and commercialization of IP, including a significant gap in the number of TTOs relative to the large number of IHE’s within Pennsylvania. A program to close the gaps is expected to accelerate technology transfer and economic development. The paper further discusses the Pennsylvania Keystone Innovation Grant (KIG) initiative. KIG provides additional resources for Pennsylvania’s technology transfer process.

The additional resources can increase the likelihood of transferring IP through licensing arrangements with existing or new companies leading to improved or new commercialized products within Pennsylvania. Early evidence from KIG recipients suggests an improved technology transfer process. Lastly, the conclusion links greater resources through the KIG initiative to a greater potential for commercialized products resulting in the greater regional and state economic development.

TECHNOLOGY TRANSFER INFRASTRUCTURE Pennsylvania’s IHEs, academic medical centers and research institutions contribute to TBED through a technology transfer process or infrastructure that includes providing resources for research; assisting with intellectual property (IP) policies to incentivizing the generation of IP; creating IP policies; disclosing, assessing and providing IP protections; and executing technical licenses to either established companies or start-ups. IP are copyrights that protect the artistic or literary works, trademarks that protect unique product or service identifiers, and, a focus of this paper, patents that protect inventions. The potential licensing income helps with continuing the technology transfer process. Furthermore, the IP generated and used by established or new businesses results in greater economic development for the region, the state, and the IHE (see Figure 1 below).6 Notice the technology transfer process may or may not include a formal technology transfer office (shaded in grey in Figure 1).

Figure 1: Transfer of Technology Process

The transfer of technology though commercialization is just one of three mechanisms where IHE’s transfer knowledge to the private sector (CEO Council for Growth, October 2007; Varga, August 2002). The first mechanism is a mixture of formal and informal networks. The formal networks consist of academic and industry professionals transferring knowledge through collaboratives such as research partnerships, workforce development initiatives, and faculty consulting. The informal networks transfer technical knowledge through communication by faculty through seminars, student internships, professional associations, and continuing education (Link, Siegel, and Bozeman, May 2006).

The second mechanism is the use of an IHE’s physical facilities where knowledge transfers occur through the presence and use of laboratories, libraries, computers, incubators and research parks. The third mechanism, which is the focus of this paper and already mentioned in Figure 1, is the formalized knowledge transfer via a TTO. This section discusses the transfer of technology process moving from left to right within the figure.7 To begin in Figure 1, research faculty and staff at IHEs apply for billions in research support annually from government, industry (sponsored research) and private funding agencies. Of interest is the economic development implication of university research. Mansfield (1991, 1995) surveys of industrial researchers found that spatial proximity

Source: Thomas Armstrong

Invention Disclosures

Faculty, Staff,

Student IP Policies,

Education

Research Intellectual Property

Protections

Invention Assessments Established

Companies

Start-Up Companies

Patents, Copyrights, Trademarks

Technical Licenses Executed

License Income

New Products,

More Jobs, Greater Wealth

Technology Transfer Office, TTO, FunctionsInstitution of Higher Education,

IHECapital, Services


26

between universities and innovating firms is more important for applied R&D than for basic research. For particular industries, he found that knowledge flows are geographically local in the information processing and drugs industry. Firms tend to collaborate with local IHEs for applied research; while for basic research, firms can collaborate with IHEs over longer distances (Acs, et al., 2002). For scientific knowledge to translate into productive applied research, researchers apply for grants to fund investigative processes. Table 1 below summarizes the Association of University Technology Managers (AUTM) annual survey of Total Research Expenditures (TREs) by State from 1996 to 2007. It is important to note that in Tables 1-7 contain data from institutions that responded to the annual AUTM survey, where some data may be understated due to some institutions not reporting. The Pennsylvania respondents to the FY2007 survey are Allegheny-Singer Research Institute, Carnegie Mellon University, Children’s Hospital of Philadelphia, Drexel University, Duquesne University, Fox Chase Cancer Center, Lehigh University, Penn State University, Temple University, Thomas Jefferson University, University of Pennsylvania, University of Pittsburgh, and Wistar Institute. It should be noted that in Tables 1-6, the drop in Pennsylvania figures from 2000 to 2001 can be mainly attributed to having Thomas Jefferson and the University of Pennsylvania not reporting for 2001.

TREs are expenditures made by the institution in support of its research activities that are funded by all sources including the federal government, local government, industry, foundations, voluntary health organizations, and other nonprofit organizations. The numbers of TREs for each state, as well as figures for subsequent AUTM tables, are primarily from U.S. IHEs as well as U.S. Hospitals and Research Institutes that voluntarily reported to AUTM. The bullet highlights are the following:8

• From 1996 to 2007, the percentage change in TREs is 106.5% for Pennsylvania relative to the US average percentage change of 123.70%.

• From 2003 to 2007, the percentage change in TRE is 30.6% for Pennsylvania relative to the US average percentage change of 31.8%. In 2007, Pennsylvania’s ranked 5th in the nation relative to the states that reported TREs.

• Of importance, notice from 1996 to 2002 that the percentage change in TREs was 44.5%--but for the Rendell Administration years of 2003 to 2007, the percentage change in TREs was lower at 30.6% for Pennsylvania. The lower figure reflects the slower increase in the growth in funding from 2003 to 2007 relative to the higher growth in funding available to all states from the previous six years.

Table 1: Association of University Technology Managers (AUTM) Annual Survey:

1996-2007 Total Research Expenditures Reported by State's Institutions of Higher Education, Hospitals, and Research Institutes

Source: Association of University Technology Managers (AUTM), U.S. Licensing Activity Survey

At the academic research stage from Figure 1, constraints upon commercialization of research includes: (1) non-tenured professors tend to focus their efforts upon publications relative to commercialization given the general recognition of publications within the tenure process; (2) grant writing time is more devoted to covering laboratory costs leaving less time for commercialization; and (3) grants typically expire before a commercialized product is developed where a funding gap exists for proof-of-concept research (Gieger and Sa’, January 15, 2009). Now TTOs rarely engage in the grant funding pre-award process. TTOs do provide education, outreach and networking about the transfer of technology process to

induce disclosure and potential for institutional resources to assist in innovation success. Continuing along the transfer of technology process in Figure 1, the academic institution, typically, has already established IP policies up-front. IP policies should include IP ownership definition, outline the IHE’s patent policy, the handling of confidential information, IP licensing and marketing approaches, IP income distribution, and inventors and IHEs rights and obligations (Krattiger et al., 2007). The rationale is to provide certainty and a set of incentives for invention disclosures while balancing the trade-off of providing revenue needs for the following: (1) the IHE for bearing some or all the cost towards the invention, (2) the TTO’s costs, and (3) the additional revenue needs to continue

1996 TREs Rank 1997 TREs Rank 1998 TREs Rank 1999 TREs Rank 2000 TREs Rank 2001 TREs Rank% Change 1996-2007

% Change 2006-2007

Pennsylvania $1,372,479,296 4 $1,459,137,606 3 $1,532,146,447 4 $1,616,758,501 4 $1,838,534,999 4 $1,372,066,027 6 106.5% 11.5%US Average $475,809,929 $480,667,831 $528,382,321 $570,589,456 $578,121,791 $662,387,623 123.7% 8.2%

2002 TREs Rank 2003 TREs Rank 2004 TREs Rank 2005 TREs Rank 2006 TREs Rank 2007 TREs Rank% Change 1996-2002

% Change 2003-2007

Pennsylvania $1,982,781,337 6 $2,170,414,157 5 $2,259,893,075 5 $2,459,635,349 5 $2,541,731,002 5 $2,834,701,790 5 44.5% 30.6%US Average $732,771,468 $807,639,589 $870,627,984 $902,262,423 $983,549,217 $1,064,309,066 54.0% 31.8%


27

the academic institution’s research agenda. IHEs without established IP policies create uncertainty where non-optimal inventions disclosures result at the institution. The process of technology transfer usually continues with the invention disclosure of an IHE’s research innovation to the institution’s TTO. The innovation traditionally comes from the academic research scientist but can also come from staff or students--especially graduate students--of the academic institution, or combination of all three.9 Of importance, there must be a significant research base, either in broad areas or a niche focus, to promote translational research. Translational research is normally supported by external financial sources, especially industry sponsored research. Invention disclosures describe the technology’s basic elements, including how the idea was conceived, how the product works, and what the possible market is. Table 2 summarizes the AUTM annual survey of Invention Disclosures Received (IDRs) by State from 1996 to 2007. IDRs include the number of disclosures, no matter how comprehensive, that are made in the year requested and are

counted by the institution. The bullet highlights are the following:

• From 1996 to 2007, the percentage change in IDRs is 66.4% for Pennsylvania relative to the US average percentage change of 91.3%.

• From 2003 to 2007, the percentage change in IDRs is 36.3% for Pennsylvania relative to the US average percentage change of 32.7%. In 2007, Pennsylvania’s ranked 4th (23rd per $100 million in research expenditures) in the nation relative to the states that reported IDRs.

• Of importance, notice from 1996 to 2002 that the percentage change in IDRs was only 9.5%--but for the Rendell Administration years of 2003 to 2007, the percentage change in IDRs was significantly higher at 36.3% for Pennsylvania. The change in IDRs have increased in Pennsylvania since 2003 relative to the previous six years of change that will potentially increase the number of new patent applications to be used for commercialization purposes.

Table 2: Association of University Technology Managers Annual Survey:

1996-2007 Invention Disclosures Received Reported by State’s Institutions of Higher Education, Hospitals, and Research Institutes*

*Rank is provided comparing state’s actual values and, in parenthesis, actual values per $100 million in total research expenditures. Source: Association of University Technology Managers (AUTM), U.S. Licensing Activity Survey

Of interest are the categories of invention disclosures. From AUTM (2008), therapeutic/medical devices represented 25% of the disclosures which is not surprising given the federal governments significant funding for biomedical research. Computer/electronics represent 9% of disclosures, research tools at 8%, finance/education/art/music at 2%, and plant at 1%. Another 15% of disclosures were characterized as “others” given these disclosures were hard to categorize for early-stage research applications that were difficult to predict. Continuing along the Technology Transfer Process in Figure 1, the TTO places the invention disclosure into an intellectual property management process. The process tracks the number of disclosures received, patent applications managed, license agreements made, contracts existing, and financial remuneration from agreements. This process is recorded, reported, and managed, typically from an IP management electronic database. An IP management database is a critical

tool for a TTO to reduce cost, complexity and recording requirements of managing a disclosed innovation.10 The TTO makes a series of invention assessments from the disclosure to determine whether the technology has commercialization possibilities.11 First, the TTO evaluates the technology for intellectual property (IP) protections for invention tests of novel, non-obvious, useful conditions (criteria for utility patents). If the evaluation is favorable, the TTO then decides whether to patent, copyright or trademark the IP, which can be a time consuming and costly process. This process often involves the TTO or its legal representatives to negotiate with patent, copyright and trademark offices to have the legal protection granted. Because of budgetary limitations and high legal costs, a TTO is selective in pursuing a patent for the most promising technologies. Less promising technologies may be reverted back to the inventor. Nevertheless, the possibility of these

1996 IDR Rank 1997 IDR Rank 1998 IDR Rank 1999 IDR Rank 2000 IDR Rank 2001 IDR Rank% Change 1996-2007

% Change 2006-2007

Pennsylvania 645.0 5 (18) 729.0 4 (14) 769.0 3 (16) 811.0 4 (12) 806.0 3 (16) 507.0 7 (26) 66.4% 13.8%US Average 223.6 235.8 249.7 257.9 248.6 280.1 91.3% 4.4%

2002 IDR Rank 2003 IDR Rank 2004 IDR Rank 2005 IDR Rank 2006 IDR Rank 2007 IDR Rank% Change 1996-2002

% Change 2003-2007



28

technologies realizing commercialization becomes less likely without a TTO processing the technology. Patents are the most common form of statutory intellectual property protection sought by U.S. IHEs (AUTM, 2007).12,13 The significant majority of patents are utility patents which is a “…grant by the (U.S). government to an inventor for any invention that is a new and useful process, machine, article, manufacturer, or composition of matter or any new and useful improvement thereof” (Krattiger et al., 2007). The holder of a patent has the right to exclude others from making, using, selling, or importing the invention in the country for normally a period of 20 years from the date of the patent application (Krattiger et al., 2007). Table 3 summarizes the AUTM data on New Patent Applications from 1996 to 2007. New Patent Applications Filed (NPAFs) is the first filing of the patentable subject matter, and does not include continuations, divisionals, or reissues.14 Gieger and Sa’ (January 15, 2009) mention that two-thirds of new patent applications are for provisional patents which last for a year; the other are full patent applications.15 A U.S. patent application filing is lagged after

an invention disclosure where the filing is often based on the previous year’s disclosures (AUTM, 2007). The bullet highlights are the following:

• From 1996 to 2007, the percentage change in NPAFs is 222.0% for Pennsylvania relative to the US average percentage change of 244.5%.

• From 2003 to 2007, the percentage change in NPAFs is 85.5% for Pennsylvania relative to the US average percentage change of 49.2%.

• Notice that from 1996 to 2002 and 2003-2007, the percentage change in NPAFs was 87.4% and 85.5%--but for the United States within the same years, the percentage change in NPAFs was 111.8% and a significant drop to 49.2%. Given that the percentage change in NPAFs have been relatively constant from 1996 to 2002 and 2003-2007 for Pennsylvania relative to the fall in United States figures, there will be a potential increase in the number of protected IP to be used for licensing or commercialization start-up purposes for Pennsylvania relative to the United States.

Table 3: Association of University Technology Managers (AUTM) Annual Survey: 1996-2007 New Patent Applications Filed Reported by State's Institutions of Higher Education,

Hospitals, and Research Institutes*


During patent protection or once a patent has been received for the technology, the TTO markets the technology to existing interested firms or to a university start-up. Concurrently, the TTO will attempt to secure financial resources to convert the innovation to a product or process. There is evidence of a local economic development dimension with regard to patents. Inventions that cite IHE patents tend to originate in the same region including patents citing published IHE research (Geiger and Sa’, January 15, 2009). From AUTM (2007), licenses are the most common form of technology transfer from a university to a firm. For interested firms, licensing negotiations occur; where, once successful, the firm gains access to the technology. A TTOs most frequent technology transfer exchange is licensing arrangements where they tend to have the sole responsibility for this transaction. One of the critical links between the academic and private industry is the licensing and licensing

option negotiations with mutually beneficial results once executed.16 The relationship between an IHE and private firm to commercialize a technology is determined by signed agreement of a license. Table 4 summarizes the AUTM’s annual survey of Licenses and Options Executed from 1996 to 2007. Licenses & Options Executed (L&OE) are the number of licenses/options that were executed in the year surveyed. The bullet highlights are the following:

• From 1996 to 2007, the percentage change in L&OE is 66.7% for Pennsylvania relative to the US average percentage change of 83.8%.

• From 2003 to 2007, the percentage change in L&OE fell 13.6% for Pennsylvania relative to the US average percentage change increase of 17.8%. Pennsylvania is currently ranked eighth (33rd per $100 million in research expenditures) among all states in 2007.

1996 NPAF Rank 1997 NPAF Rank 1998 NPAF Rank 1999 NPAF Rank 2000 NPAF Rank 2001 NPAF Rank% Change 1996-2007

% Change 2006-2007

Pennsylvania 246.0 3 (15) 317.0 3 (12) 357.0 3 (9) 428.0 3 (9) 479.0 3 (10) 329.0 5 (11) 222.0% -5.6%US Average 72.3 90.6 104.7 118.6 126.0 141.4 244.5% 0.1%

2002 NPAF Rank 2003 NPAF Rank 2004 NPAF Rank 2005 NPAF Rank 2006 NPAF Rank 2007 NPAF Rank% Change 1996-2002

% Change 2003-2007



29

Table 4: Association of University Technology Managers Annual Survey: 1996-2007 Licenses & Options Executed Reported by State’s Institutions of Higher Education,



While mutually beneficial results are possible, execution of licenses or options may not occur as shown by the percentage decline in Pennsylvania for 2003 to 2007. TTO licensing officers may be more risk averse than their industrial counterparts where negotiations may not be culminated. Another example is that IHE’s IP policies may be prohibitive for industry to accept. In addition, monetary incentives for TTO officers may be lacking to pursue aggressive licensing execution deals. Also, TTO staffing may be too limited to pursue numerous licensing deals (Varga, August 2002; Gieger and Sa’, January 15, 2009).17 TTOs license to large companies, small companies, and startups. AUTM 2007 reports from the FY 2006 annual survey that 50.7% of the 4,192 total licenses executed by U.S. IHEs are to small companies, 31.7% to large companies, and only 16.7% to startups.18 Small companies tend to receive the dominant share of IHE licenses. From U.S. hospitals and research institutions, there is a difference where large companies receive the dominant share of 755 executed licenses with 42.5%, small companies with 48.3%, and startups with 8.7%. While no figures are reported for Pennsylvania IHEs or hospitals and research institutions, the data for Table 5 on startups, discussed below, suggest a smaller percentage change in startups within Pennsylvania relative to the U.S. average. Licensing is either nonexclusive or exclusive. Nonexclusive is not radical technology but technology that will become a standard and used in conjunction with existing technology, or developed by a company that needs freedom to operate. Exclusive is where a company must invest substantial resources to commercialize radical technology.19 Out of 4,947 licenses executed by U.S. IHEs, hospitals, and research institutions, 63% are nonexclusive and 37% exclusive licenses for FY 2006 (AUTM, 2007). Licenses of technology that tends to be radical or a significant departure from current technology in the market will tend to be marketed by a new IHE, hospital or research institution start-up as an exclusive license relative to established small or large companies. Out of 4,947 licenses

executed by U.S. IHEs, hospitals, and research institutions, 764 licenses are to startups where 91% exclusive and only 9% nonexclusive licenses for FY 2006 (AUTM, 2007).20 An IHE start-up tends to be based on early-stage innovation or at the proof-of-concept stages. From Geiger and Sa’ (January 15, 2009), start-ups that are faculty lead are likely to occur because (1) research fields offers strong patent protection, (2) decentralized industries allowing niches for distinctive products, (3) technologies that transcend the usage by any one corporation, (4) or radical technologies to existing product or processes. The number of new company formations from an IHE-developed technology tends to be restrictive, though, due to insufficient capital funds required for new start-ups.21 In addition, inadequate incubator and laboratory space relative to localized centers of innovation reduce the likelihood of successful university start-ups. Finally, the inability to recruit highly qualified executive managers limits the number of new start-ups. AUTM (2008) reports funding sources for 555 new start-up companies in 2007 based on university technologies from 191 respondents, which is equivalent to the 554 formed in 2006 by 183 respondents. Internal funding from friends and family at 24.3% was the single largest source of capital investment followed by angel investment at a total of 20.5% and VC investment at 15.9%. State funding is a significant source at 15.9%. Corporate partners and other sources of funding for new start-ups are 5.9% and 8.5% respectively. Their own institution is a source of funding at 9.2% of respondents and SBIR/STTR source of funding is 7.6%. It is also reported that 15.5% of respondents indicate no external funding sources. Note: The percentages do not add up to 100% given a start-up may have reported more than one source of funding. Table 5 summarizes the AUTM’s annual survey of startups reported from 1996 to 2007 as one method to commercialize technology from licensing to established companies or startups. Startups are companies that were dependent upon

1996 L&OE Rank 1997 L&OE Rank 1998 L&OE Rank 1999 L&OE Rank 2000 L&OE Rank 2001 L&OE Rank% Change 1996-2007

% Change 2006-2007


2002 L&OE Rank 2003 L&OE Rank 2004 L&OE Rank 2005 L&OE Rank 2006 L&OE Rank 2007 L&OE Rank% Change 1996-2002

% Change 2003-2007



30

licensing an institution's technology for initiation. The bullet highlights are the following:

• From 1996 to 2007, the percentage change in startups is 52.6% for Pennsylvania relative to the US average percentage change of 155.4%

• From 2003 to 2007, the percentage change in startups is 11.5% for Pennsylvania relative to the US average percentage change of 57.3%.

Table 5: Association of University Technology Managers Annual Survey: 1996-2007 Startups Reported by State’s Institutions of Higher Education,



Generally, the IHE or institution’s TTO receives an initial payment for the license, continues to collect royalties for use of the technology, or receives an equity position where income occurs from the equity or the selling of the equity position. AUTM (2008) reports that out of the 555 new start-up companies in 2007 based on university technologies, 54% involved equity deals with research institutions. The various types of payments include royalties, fixed payments, common stock, R&D funding, lab equipment, consulting services, grant-backs, options, or access to other proprietary resources (Krattiger et al., 2007). Generally, a combination of royalties and equity stakes can disperse risk between the two parties. These funds provide an income stream for additional resources to the TTO office or IHE; thereby permitting financial resources to spur further research and development at the campus level, creation of infrastructure to support students and capital projects, and direct support of researchers. This result is the continuation of the transfer of technology process.

Table 6 summarizes the AUTM data on Licensing Income Received from 1996 to 2007. Licensing Income Received (LIR) includes license issue fees, payments under options, annual minimums, running royalties, termination payments, the amount of equity received when cashed-in, and software and biological material end-user license fees equal to $1,000 or more. The bullet highlights are the following:

• From 1996 to 2007, the percentage change in LIR is 157.4% for Pennsylvania relative to the US average percentage change of 342.4%.

• From 2003 to 2007, the percentage change in LIR is 62.6% for Pennsylvania relative to the US average percentage change of 126.9%.

• Notice that from 1996 to 2002, the percentage change in LIR was only 9.5%--but for the Rendell Administration years of 2003 to 2007, the percentage change in LIR was significantly higher at 62.6% for Pennsylvania.

Table 6: Association of University Technology Managers Annual Survey:

1996-2007 Licensing Income Received Reported by State's Institutions of Higher Education,


The outcomes of TTO services are an expansion of economic development and economic impacts to the region and the state (Technology Transfer Tactics, August 2008). Economic development is enhanced due to close contact university-industrial relationships which leverage scientific breakthroughs that would less likely occur without the close contact relationships. The result is commercialization of many innovations will tend to be local, especially in early

*There is large increase in LIR’s from 2005 to 2006. Part of the reason of the large percentage change is the high 2006 LIR value difference of almost $50 million relative to the 2005 value where over $43 million increase was reported by Wistar Institute only.

By deleting values reported by Wistar in 2003 and 2006, the percentage increase is close to the 62.6% increase from 2003-2007. Rank is provided comparing state’s actual values and, in parenthesis, actual values per $100 million in total research expenditures. Source: Association of University Technology Managers (AUTM), U.S. Licensing Activity Survey

1996 Startups Rank

1997 Startups Rank

1998 Startups Rank

1999 Startups Rank

2000 Startups Rank

2001 Startups Rank

% Change 1996-2007

% Change 2006-2007

Pennsylvania 19.0 3 (12) 27.0 3 (8) 23.0 3 (16) 19.0 3 (17) 20.0 5 (26) 14.0 11 (29) 52.6% -12.1%US Average 4.7 6.1 7.0 6.5 8.0 9.5 155.4% -0.6%

2002 Startups Rank

2003 Startups Rank

2004 Startups Rank

2005 Startups Rank

2006 Startups Rank

2007 Startups Rank

% Change 1996-2002

% Change 2003-2007


1996 LIR Rank 1997 LIR Rank 1998 LIR Rank 1999 LIR RankPennsylvania $14,894,614 9 (23) $19,617,823 7 (21) $46,936,768 5 (10) $16,620,088 12 (25)US Average $13,479,155 $15,270,226 $18,107,662 $20,797,550

2002 LIR Rank 2003 LIR Rank 2004 LIR Rank 2005 LIR RankPennsylvania $16,306,601 14 (28) $23,587,566 13 (27) $22,783,813 14 (30) $25,144,524 15 (29)US Average $26,011,856 $26,284,534 $29,266,112 $42,991,183


31

stage production of innovations because of the necessary relationships needed for implementing the technology with either the faculty developers of the founding company or research team. AUTM (2008) reports that out of the 555 new start-up companies in 2007 based on university technologies, 402 or 72% of new start-ups was in the licensing institutions’ home state. Varga (August 2002) in the review of literature suggests a local academic technology transfer impact on non-routine functions such as R&D, prototype manufacturing, or small volume production relative to insignificant impacts upon routine production activities. Small firm impacts as researched by Acs et al. (1994a,c) show that knowledge spillovers as measured by product innovations from IHEs have a more decisive role in innovation activity of small firms relative to large firms. In addition, economic development is enhanced by the hiring of skilled undergraduates and graduates with tacit knowledge (Armstrong, et al., 2007). New knowledge is transferred through training of students with tacit knowledge to be used by employers for product and process development. This method is the traditional role of IHEs impact on economic development. The establishment of a TTO and the services it provides support an expansion of economic impacts including new products, new company formation, jobs supported by the new technology, greater incomes and value-added expenditures (Woodward and Guimarães, May 2008). Additional benefits as the result of an established TTO and income received from patenting and licensing activity are the following: dissemination of new knowledge through new products and processes; rewards and incentivize inventors; supports directly or indirectly the operations of the TTO to further innovation; and advances IHE research activities. Established companies and start-up companies directly provide greater expenditures and jobs due to the introduction of new technologies with IHEs having greater revenues for expanded spending purposes.

A fully functioning TTO takes time to establish a robust technology transfer program and impact the local economy. Krattiger et al. (2007) report that a technology transfer program may take eight to ten years to build an IP portfolio, establish contacts, and develop skills. Furthermore, an IHE technology transfer program may take up to two decades to substantially impact the local economy. Part of the reason for the delayed time to impact the economy is the technology transfer process which typically takes on average from six to ten years from the moment of invention disclosure to when significant income is generated from a license. Table 7 below provides data from the AUTM 2007 survey for technology transfer activities per $100 million in research spending. The data follows previously discussed Figure 1: Transfer of Technology Process. The Process is an “hour glass” type of shape where there is a “funnel” down of research to license income received from established and start-up companies to a “funnel” up from established and start-up companies producing greater income and job impacts due to economic multiplier impacts (Armstrong, 2006). The $100 million figure was used as the low end of estimates ranging from $100 to $500 million in research expenditures annually to justify the costs of a fully functioning TTO at an IHE or consortium of IHEs (Krattiger et al., 2007). The bullet highlights are the following:22

• The invention disclosure rate is 40 disclosures per $100 million in research expenditures for the U.S. while it is 38 for Pennsylvania.

• As expected, there is a 42% decline from invention disclosures to new patent applications files in the US while only 26% decline for Pennsylvania.

• The data report a 75% U.S decline and 83% Pennsylvania decline from invention disclosures to license and options executed.23

• There are 1.13 U.S. and 1.02 Pennsylvania start-ups per $100 million in research spending.

• Finally, 5.6% of research spending results in licensing income received in the U.S. (4.3% in 2006) and 1.4% for Pennsylvania for 2007 (3.0%. in 2006).

Table 7: 2007 Technology Transfer Activities per $100 million in Research Spending*

*Total U.S. research spending for 2007 was $47,893,907,971 and for Pennsylvania was $2,834,701,790. Source: 2007 Association of University Technology Managers Survey

Technology Transfer Activity (TTA) United States % Change from Invention

Disclosures Received Pennsylvania % Change from Invention Disclosures Received

Invention Disclosures Received 40.19 37.85New Patent Applications Filed 23.41 -41.8% 27.94 -26.2%Licenses and Options Executed 10.15 -74.7% 6.53 -82.7%Startups 1.13 -97.2% 1.02 -97.3%Licensing Income Received $5,603,484 (5.6% of research) $1,352,625 (1.4% of research)


32


33

Currently, Pennsylvania is suffering from a formal technology transfer infrastructure gap. About 96% of the 158 Pennsylvania IHEs (excluding branch campuses) have no formal TTO.24 This figure only includes Pennsylvania IHEs that the Department of Community and Economic Development’s (DCED’s) Technology Investment Office can provide partnership interaction in 2008. While some of the IHE’s within the gap may have some or informal technology transfer capacity such as a branch campus connected to the TTO of the main campus or a signed agreement with a TTO of another IHE, the compelling fact is that without a significant technology transfer infrastructure having a formal TTO relationship, promising IP may not be disclosed, patented, licensed, or commercialized. Pennsylvania’s technology transfer infrastructure gap leads to unrealized technology development within the Commonwealth and inhibits the Commonwealth’s economic development. This gap becomes apparent in the IHE’s voluntarily reporting in the AUTM annual survey of Licenses and Options Executed and Startups and, of course, the significant majority of Pennsylvania IHEs are not reporting. The challenges facing Pennsylvania’s technology transfer process is to provide appropriate resources to accelerate the research to commercialization stages of IP by creating the technology transfer infrastructure where it does not exist or reducing the gaps within the technology transfer process at Pennsylvania’s IHEs. The consequences are that relatively fewer IHE technologies get licensed for commercialized products due to IP not getting protected, transferred, and commercialized. Keystone Innovation Grants were first introduced in FY2005-06 with two additional fiscal years of funding providing additional financial resources to enhance Pennsylvania’s technology transfer infrastructure for the goal of greater regional and state economic development.

PENNSYLVANIA’S KEYSTONE INNOVATION GRANTS INITIATIVE

The Keystone Innovation Grant (KIG) program is a companion program to the Keystone Innovation Zone (KIZ) program (Armstrong and Yazdi, 2004) statutorily authorized under Act 2004-12. In April 2004 the Commonwealth set out to catalyze IHEs and industry collaborations by creating the KIZ program. The KIZ program focuses Pennsylvania’s commitment to creating new technologies and entrepreneurs – using Pennsylvania’s IHEs to deliver economic development opportunities throughout the Commonwealth. KIGs, as a companion program, provide funding to encourage the technology transfer and commercialization of intellectual property between Pennsylvania’s technology-oriented businesses and entrepreneurs and KIZ-participating Institutions of Higher Education (IHEs) and to spur the development of new businesses in the Commonwealth.

The Technology Investment Office within the Pennsylvania Department of Community & Economic Development (DCED) oversees the KIG program which aligns with the other entrepreneurial, workforce and technology development programs administered within DCED, including the Ben Franklin Technology Development Authority, and other state government programs. The mission of DCED’s Technology Investment Office is to serve as a catalyst for growth and competitiveness for Pennsylvania companies and universities through technology-based economic development (TBED) initiatives including funding, partnerships and support services. KIG funding provides financial resources, especially in the critical stages of technology transfer, where revenue flow is negative, so as to increase the likelihood of commercialization success and economic development. The primary purposes for KIG are focused on the following (Pennsylvania Department of Community and Economic Development, February 2008):

• To provide seed capital in the form of grants or loans for faculty and students to perform proof of concept efforts including business plan analysis, marketing analysis, prototyping, patent research and filing, intellectual property, licensing and royalty agreements and other uses to be approved by the Department of Community and Economic Development upon request.

• To provide seed capital in the form of grants or loans for KIZ companies, which must be less than eight years old that are licensing/transferring technology from a Pennsylvania IHEs participating in a KIZ.

• To provide capital to support additional incubator services dedicated to companies created from the transfer of technology from IHEs or for companies, especially KIZ companies, collaborating directly with KIZ-participating IHEs.

• To hire staff in a technology transfer office or to create a shared resource to provide technology transfer assistance.

The Appendix provides a list and description of 58 KIG awards totaling $9.9 million for the past three fiscal years.25 Each KIG may not exceed $250,000 per award and must be matched by the KIZ-participating IHE applicant, with a total lifetime award of $750,000 (Pennsylvania Department of Community and Economic Development, February 2008). A KIG proposal must address all of the following on (Pennsylvania Department of Community and Economic Development, February 2008):

• A description of the existing technology transfer approach being taken at the IHE. This should

include where appropriate: (1) the number of dedicated full-time equivalent employees dedicated to this function and (2) the level of success in patent filings, awards, licensing agreements, royalties received, products commercialized, new companies created, in the most recently recorded calendar or academic year. Any other measure that the Pennsylvania IHE is currently collecting related to technology transfer deemed appropriate to provide.


34

• The technology transfer activities to be undertaken. The activities may include the addition of personnel who are directly related in transferring technology to the local businesses.

• The quantifiable goals, objectives and milestones to be achieved.

• How the activities, goals, objectives and milestones will integrate with the strategic plan adopted by the KIZ. It is desired that a review team be created with KIZ Partnership members to provide guidance and oversight regarding the award of grants to specific projects.

• The role of the applicant and other members of the KIZ Partnership.

• Identification of a dollar-to-dollar match, which may be in-kind if DCED determines that the proposed match can be readily identified and tracked, and

which is directly related to the stated goals, objectives and milestones. As the eligible activities defined under this program are intended to be new activities to enhance technology transfer, it is expected that match funds will be a commitment of new resources by the IHE.

• The KIZ Coordinator’s review must include: verification of the KIZ Partnership status in the form of an approval/support letter, which indicates the application’s alignment with the KIZ goals and objectives, and stipulation that the application meets the KIZ’s targeted industry sectors.

Pennsylvania launched the KIG initiative with the first awards in FY2005-06 followed by two additional fiscal years of funding to enhance Pennsylvania’s technology transfer infrastructure. The overall goal of KIGs is to apply financial resources in the critical stages of technology transfer from discovery to prototype where revenue flow is negative, and there is a strong likelihood of a lack of commercialization success (see New Company/New Product Life Cycle graph below). This area is termed the “Valley of Death” that refers to the critical stage of new company formation and/or new product development which is most risky and where most companies/products fail. Notice that the KIG area in the Life Cycle graph enhances the Figure 1: Transfer of Technology Process shaded grey and extending to licenses executed to existing or start-up companies, which may or may not be overseen by a TTO.

KIG PIN

Source: Alan Brown, Pauline Yankes and Thomas Armstrong, 2008.

For firms to take advantage of scientific breakthroughs at IHEs, Zucker and Darby (March 2005) report that access to knowledge, primarily retained by discovering scientists or excludable knowledge, is needed. Top scientists become the human capital resource where new firms are built or existing firms are transformed. In addition, as scientists increase their collaboration with industry researchers, there is a strong positive effect on a company’s success. KIG funding provides Pennsylvania IHEs with additional resources to foster commercialization of university IP transferred to existing or new firms for enhanced products.

A robust technology transfer program requires financial support from the earliest stages of developing ideas at the research stage from research bench to prototyping to full scale demonstration. AUTM (2004) reports that IHE start-ups within the “Valley of Death” are supported by venture capitalists (18.6%), angel (16.4%), or corporate investments (5.5%). One third of the rest has personal financial resources or no external funding while the remaining obtained financial resources through continued research. KIG funding was created to assist with additional financial resource needs.


35

KIG awards are helping one of the biggest obstacles for enhancing technology transfer at institutions with and without a technology transfer office (TTO)--insufficient resources as reported by Gieger and Sa’ (January 15, 2009). Within the Commonwealth, about 4% of all Pennsylvania IHEs do have a TTO. KIG awards enhance technology transfer at institutions with a TTO: Carnegie Mellon University, Drexel University, Lehigh University, Pennsylvania State University, Temple University, Thomas Jefferson University, University of Pennsylvania, University of Pittsburgh, and Wistar Institute. An example of a KIG enhancing TTO services is illustrated by the University of Pittsburgh (see Appendix for summary KIG descriptions of all KIG recipients).

• University of Pittsburgh’s Office of Technology Management (OTM) seeks to better manage the growth of its technology commercialization activities and facilitate the development and implementation of more-effective business opportunity development and commercialization awareness/education strategies. KIG resources will assist in the following initiatives: (1) help the University’s OTM develop stronger systems and refine already-existing systems for more-effectively “triaging” new University technologies for market; that is, to add value and prepare them for licensing to existing companies and start-up companies; (2) provide a more formalized funding mechanism for innovators challenged with pre-commercial proof-of-concept and other value proposition development issues; (3) improve both the quantity and quality of interactions with potential industry partners, investors and regional research partners; (4) help the OTM build on its existing educational initiatives and develop a long-term educational strategy to support and foster future innovation commercialization activity at the University of Pittsburgh and in the region; and (5) help the OTM continue to promote and foster a more “academically entrepreneurial” environment on campus that results in engaging more faculty, staff and students actively in substantive and meaningful innovation development and commercialization.

Within the Commonwealth, about 96% of Pennsylvania’s IHEs do not have a TTO severely limiting technology transfer, commercialization and the growth of economic development possibilities. About half of the competitive KIG grant recipients are without a formal TTO. KIG funding helps IHEs without a TTO to accelerate technology transfer in the following areas illustrated by a KIG recipient (see Appendix for summary KIG descriptions of all KIG recipients):

• To assist an IHE in establishing a TTO with cooperation from an IHE already having a TTO. An

example is the establishment of the Office of Technology Management (OTM), a TTO, at the University of the Sciences in Philadelphia (USP) mentored by the TTO at Thomas Jefferson University (TJU) by the KIG award. TJU provided (1) an educational outreach program at USP to increase faculty awareness of technology commercialization where the OTM has taken the lead to outline services, (2) training of a technology transfer coordinator to handle day-to-day technology transfer-related activities at USP, (3) consultation and problem-solving service to the Director of the newly-established OTM, and (4) services to license USP’s research results and generate license revenue that benefits USP’s research community. Also, the two offices work together to connect their faculty to potential commercial partners through research collaborations and technology packaging opportunities.

• To create an IHE relationship with an existing TTO. KIG funding will support the development of the Technology Transfer and Commercialization Resource Network (TTCRN) to serve the following fourteen Pennsylvania State System of Higher Education (PASSHE) universities: Bloomsburg, California, Cheyney, Clarion, East Stroudsburg, Edinboro, Indiana. Kutztown, Lock Haven, Mansfield, Millersville, Shippensburg, Slippery Rock and West Chester. PASSHE TTCRN will (1) develop its technology transfer and commercialization activity through a contract with the Penn State Research Foundation, which will serve as the technology transfer agent for PASSHE; (2) develop, define and publish IP Guidelines and Procedures for technology transfer and commercialization activity for students, staff, and faculty; (3) provide educational forums on technology transfer and commercialization to encourage and support innovation, scholarly activity and sponsored research for faculty, students and staff; and (4) engage a professional training/intellectual property firm to create the online, narrated tutorial for faculty, students and employees.

• To support technology transfer functions within a group of IHEs. KIG funding beginning in 2005-06 helped to create the Innovation Transfer Network (ITN). ITN connects faculty and students with private sector companies in an effort to drive technology transfer results. Forging new relationships between faculty and business, increasing opportunities for project collaboration, providing seed grant funding, offering intellectual property support and services, and raising the visibility through media and outreach, all serve to


36

build the necessary infrastructure in South Central Pennsylvania to increase the level of technology transfer to the private sector from 13 of the region’s institutions of higher education: Dickinson College, Elizabethtown College, Franklin & Marshall College, Harrisburg Area Community College, Harrisburg University of Science & Technology, Kutztown University, Lancaster General Hospital College of Nursing and Health Sciences, Lebanon Valley College, Messiah College, Millersville University, Penn State Harrisburg, Penn State Milton S. Hershey College of Medicine, and Shippensburg University.

• To support technology transfer functions with a single IHE. Susquehanna University is a liberal arts college that currently does not have capacities in technology transfer. KIG funding will help to providing resources and support networks necessary in to promote an innovation economy with the following steps: (1) launch the Greater Susquehanna Business Idea Challenge with student and professional categories and generate 10 or more qualified entries in the life sciences/health care, information technology, wood products, and diversified manufacturing; (2) inaugurate an entrepreneurship symposium focusing on a topic or theme of interest to entrepreneurs and small businesses in Central Pennsylvania; (3) provide a minimum of 3 student interns to business start-ups or small businesses; (4) conduct campus-based needs assessment and asset inventory to assist Susquehanna in understanding their potential, targeting specific areas for support, and connecting campus resources to community needs; (5) conduct a campus-based entrepreneurship competition to encourage innovation, creativity and new thinking around research and scholarship as having commercial viability; and (6) integrate intellectual property/technology transfer protocols into a newly created policies and procedures manual.

As depicted in Figure 1: Transfer of Technology Process and the New Company/New Product Life Cycle graph,

accountability measures are required so that the KIG initiatives are tracking positive technology-based economic outcomes (Measuring Up, 2007). From the technology idea to prototype to production via licensing to an existing company or new university/industry spinout, metrics are generated to determine impact of the KIG initiatives. Table 8 summarizes the impact of the KIG initiatives since 2006 including more detailed breakout impact data since July 1, 2007. Table 8 shows evidence that $9.9 million were awarded to 58 Keystone Innovation Grant (KIG) recipients where 42 of the KIG recipients received multiple KIG awards for a total of $7.7 million and 16 recipients received a single award for a total of $2.2 million as an integral component for regional and state technology-based economic development. Since 2006 to December 31, 2008, KIG funding not only resulted in $9.9 million in match but resulted in leveraging these funds for a total of $107.1 million. The total leverage funding includes venture capital, private equity, debt financing, federal, local, foundation funding, or other funding sources and excludes other Pennsylvania state sources. The result of state KIG funding is a 10.9 to 1 leverage ratio. About $3.4 million is devoted to technology research, development, testing and evaluation as the result of KIG funding from July 1, 2007 to December 31, 2008. While no figures are reported for invention disclosures, 1,091 patents were filed since 2006 with 268 patents awarded. The result of patenting intellectual property is the starting up of new companies or the licensing to existing companies. From the protected IP, 133 licenses were granted earning $21.5 million since June 1, 2007. There were 71 newly created start-ups where 57 were from Pennsylvania IHEs. Seven new products were developed. In addition, 598 businesses were assisted. The human capital impact upon Pennsylvania due to KIG-funded projects has been positive. Since 2006, there have been 64 combined graduate and undergraduate internships. In addition since July 1, 2007, 34 undergraduates and 28 graduates enrolled at an IHE due to technology transfer activities encouraged by KIG funding.

Table 8: KIG Impact Metrics

A primary rationale for the KIG initiative beginning in FY2005-06 is to provide resources to advance the technology transfer process of IP from IHE research to a greater likelihood of profitable commercialization of enhanced characteristics within existing or new products. The AUTM Tables 1-6 report data from 1996 to 2007 while the KIG impact Table 8 report data from 2006 and on. It can be expected that the additional KIG resources will positively enhance Pennsylvania’s technology transfer process as surveyed in future annual AUTM Tables 1-6 with all else in the economy remaining the same in the following ways:

• KIG funds and dollars leveraged totaled about $117 million that have become available where some of the funds are used for research expenditures including research, development, testing and evaluation of $3.4 million that is expected to advance research expenditures by state institutions in Table 1.

• KIG awards are expected to have a positive impact on invention disclosures received as shown in Table 2 and patents filed in Table 3 shown in Table 8 with 1,091 patents filed from 2006 to the end of calendar year 2008.

• KIG initiative is expected to have a positive impact on licenses and options executed in Table 4 where

133 licenses have already been granted due to KIG from July 1, 2007 to December 31, 2008 in Table 8.

• KIG impacts report 57 university spin-outs from 2006 to 2008 in Table 8 that are expected to have a positive impact on AUTM’s Table 5 start-ups.

• KIG licensing revenue earned, $21.5 million, is expected to have a positive impact on AUTM’s Table 6 licensing income received.

CONCLUSIONS

An IHE’s potential for generating IP depends upon four key factors: (1) the amount of research occurring in patent-rich fields, (2) the quality of academic scientists and engineers, (3) the degree which faculty encourages entrepreneurial activities, and (4) the size and effectiveness of the technology transfer process without or better with a TTO (Geiger, and Sa’, January 15, 2009). As to the last factor, the challenges facing Pennsylvania’s technology transfer process is to provide appropriate resources to accelerate the research to commercialization stages of IP by creating the technology transfer infrastructure where it does not exist or reducing the gaps within the technology transfer process at Pennsylvania’s IHEs.

Dollars Leveraged* $107,114,675 KIG Funds** $9,864,427Jobs Created 243 Jobs Retained 278Newly Created Startups 71 Industry Spin-Outs 29Institutions of Higher Education Startups 57 Businesses Assisted 598

Patents Awarded 268 Patents Filed 1,091Seed Capital Awarded $978,700 New Technologies Developed 648

Licenses Granted 133 Undergraduate and Graduate Internships 64

Licensing Revenue Earned $21,537,441 Participating Undergraduates Enrolled 34

Research, Development, Testing, and Evaluation**** $3,356,666 Participating Graduates Enrolled 28

7

Keystone Innovation Grant Impacts: 2006 to 12/31/2008

Keystone Innovation Grant Impacts: 7/1/2007 to 12/31/2008***

Source: Semi-annual report submissions. Category definitions reported in Measuring Up (2007).

****Research, Development, Testing, and Evaluation (RDT&E) expenditures include monies spent on labor, services, materials and equipment used to conduct research, development, testing and evaluation as recognized by your company or institution of higher education, regardless of where the RDT&E was physically undertaken. RDT&E excludes in-kind contributions provided by your institution of higher education, company and/or third-parties.

*Leveraged dollars is the sum of venture capital, private equity, grants, and debt financing dollars.

***Collection of additional data began 12/31/2007 as the result of reporting requirements from Measuring Up (2007).**KIG funds for FY05-06, FY06-07, & FY08-09.

Development of New Products


37


38

With this in mind, the Technology Investment Office of DCED is in the process of updating KIG, to be called Innovation Grants, with its public and private partners, contingent upon the Ben Franklin Technology Development Authority approval and appropriation from the General Assembly. As the Commonwealth continues to advance university developed IP, along with academic medical and non-profit research institutions, as a key technology-based economic development initiative partnering with higher education institutions and industry, KIG and Innovation Grants will provide a key component for sustainable economic development.

ENDNOTES *The author would like to thank Jeannine Marttila, Mark Cresswell, Seth Maset, and Kyle Yurick and discussant for their assistance and comments. The conclusions do not necessarily reflect the positions of the Pennsylvania Department of Community and Economic Development. All possible errors are the author’s. 1. Article I, Section 8, Clause 8, of the U.S. Constitution states “The Congress shall have Power…To promote the Progress of Science and useful arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries,” that provides the basis for the protection of intellectual property in the United States. Now intellectual property using the Department of Community and Economic Development funds is owned by the grantee as long as the grant funded project accomplishes their deliverables and close out their contract/audit/close-out report. 2. The Bayh-Dole Act (P.L. 96-517) on December 12, 1980 was signed into law to become effective in July 1981. Some of the important provisions are (1) establishes a uniform patent policy for the first time; (2) IHEs are encouraged to collaborate with commercial concerns to promote utilization of inventions arising from federal funding; (3) IHEs may elect to retain title of inventions conceived utilizing federal funding; (4) IHEs must file applications for patents on inventions the elect to own; (5) federal government retains a nonexclusive license to practice the invention throughout the world for governmental purposes; (6) share royalty income with inventors and remaining income for research and education; (7) preference in licensing must be given to small businesses; (8) and preference for U.S. industry (Bremer, January 2006). 3. The number of U.S. respondents to the annual AUTM survey (IHEs, hospitals, and research institutions) has increased from 120 in 1991 to 189 in 2006 and 194 in 2007, where the institutions responding have a high likelihood of a formal TTO. The figures maybe understated by the degree of

institutions having a formal TTO but not responding to the annual AUTM survey. Fox Chase Cancer Center, Thomas Jefferson University, and Wistar Institute are Pennsylvania hospitals or medical research institution that have formal TTOs but are not listed as a Type 15 or 16 Carnegie Classified Institution. The Carnegie Classifications are based on data through 2004. 4. Transfer of technology can be considered a basic extension of an IHE’s mission to teach, to research by generating new knowledge, and to provide service for society. Transfer of university knowledge to private businesses has a long tradition through agricultural extension serviced at US land grant IHEs (Rosenberg and Nelson, 1994). 5. The other goals of a TTO is to provide service to faculty and generate income for the IHE and inventors (Geiger and Sa’, January 15, 2009). Gieger and Sa’ (January 15, 2009) report that for most AUTM IHE’s with an established TTO, revenues from IP do not cover the cost of operating a TTO. 6. A vigorous technology transfer process consists of the following three implemented policies: (1) government support of science education, research and related infrastructure including technology transfer infrastructure, (2) rule-of-law protections including IP, and (3) market reliance on determining which technologies and products should be developed (Krattiger et al., 2007). Information for the technology transfer process is from CEO Council for Growth (October 2007). 7. While the focus of this paper is on transfer of technology process and patent IP protections, copyrights and trademarks are important protected IP too. Copyrights grant exclusive rights to authors or in some cases the author’s employer. Formal copyright registration is not required for protection, but registration does result in enhanced rights including the creating the presumption of ownership within five years of publication. Copyright protection applies to any original works of authorship fixed in a tangible medium for a wide range of works including books, plays, software, music and lyrics, artwork (pictorial, graphic, or sculptural), motion pictures, and architectural works. Works by a single author or joint authorship is protected by the life of the author or last surviving author plus 70 years. Works for hire copyright duration is for 95 years from the last date of publication or 120 years from the date of creation, whichever expires first. IHEs generally grant faculty ownership of copyright for scholarly works, works such as computer software and databases may not qualify for the scholarly works exception (Harris and Smith, January 2006). A trademark or brand name is a word, name, or symbol which is adopted and used in commerce by a business to identify the products, distinguishable from products sold and indicates the source of the goods. A service mark is similar


39

to a trademark but is used in the sale or advertising of services rather than products. A trademark is governed under federal Lanham Act, Title 15 of the U.S. code) and state law. Federal registration provides protection throughout the United States while state registration is only enforceable within that state. Generally, the name of a business cannot be registered unless it is used as a trademark or service mark. Registration is not required to establish rights in a mark—actual use in commerce is all that is required. However, there are advantages to register including preventing registration of similar marks, secure injunctive relief and damages in a federal court, and have the mark treated as incontestable after five years of use. A federal application can be filed with the U.S. Patent and Trademark Office for a good used in interstate commerce. A federal trademark/serve mark registration is valid for ten years and is renewable for ten year periods as long as the mark remains in use exception (Needle, January 2006). 8. Table 1 AUTM figures are compared to the National Science Foundation’s Survey of Research and Development Expenditures at Universities and Colleges from FY 2001-2006. The total for 2001-2006 show that the NSF figures for Pennsylvania are understated by 1.3% and for the United States, the NSF figures are 8.6% higher than AUTM’s survey figures. Of course, individual fiscal years will be different. 9. Financial support for graduate students is often a defining factor in the speed of early-stage technology development (AUTM, 2007). 10. The IP management data bases are used by the some of the following Pennsylvania IHEs: Access-based, customized management database (University of the Sciences in Philadelphia), Inteum (Lehigh University and Penn State University), SMIRP (Drexel University), and Wellspring (Carnegie Mellon University). 11. While there is no best method for evaluating technology, Krattiger et al. (2007) discusses five approaches and pros and cons: (1) costs approach—price of a product is based on the cost of developing a product; (2) income approach—technology evaluation is based on anticipated revenues discounted; (3) market approach—the value of the technology is based on comparable technology; (4) hybrid approach—technology evaluation is a combination of income and market approach; and (5) royalties approach—technology evaluation is based on royalty rates applied to similar technologies. 12. A U.S. patent filing generally corresponds to seek patent protection of a single invention disclosure. However, there may be two or more invention disclosures combined with a single patent application, and a single invention disclosure may generate more than one U.S. patent application (AUTM,

2007). Other forms of IP protections are (1) trademarks conferring an exclusive right to use the mark in commerce, (2) geographical indications indicating the name of the place of origin of a particular good, (3) copyrights providing statutory protection for the original works of authors, and (4) trade secrets protecting confidential know-how information (Krattiger et al., 2007). 13. From Gieger and Sa’, chapter 4 (January 15, 2009), IHEs contributions tend to be concentrated in a few areas among hundreds of patent classes. For example, 19.0 percent of all biotechnology, 18.8 percent of drug, and 17.6 percent of chemistry or chemistry engineering patents were awarded by IHEs from 2001 to 2003. Other areas where patents tend to be significantly represented and awarded to IHEs from 2001 to 2003 were electrical engineering/computer science (7.1 percent), materials (6.1 percent), surgery (5.8 percent), and optics/physics (5.3 percent). 14. As an example from FY 2004, almost 1/3 of new patent applications filed are issued (Gieger and Sa’, January 15, 2009). It generally takes about three to five years from a time of patent application is filed until it issues in the U.S. If patent protection is pursued outside the U.S., protection should be pursued no later than 12 months form the filing date of the original U.S. patent application (Hermanns and Wagner, January 2006). 15. Provisional patents were offered to inventors beginning in 1995 by the US Patent and Trademark Office. Given that a provisional patent requires less information and at about one-tenth the cost of a full patent, while protecting the inventor for 12 months upon filing the provisional patent, patent filings have increased. Furthermore, provisional patents can provide an extra year of IP protection from 20 to 21 years if a utility patent is awarded after the provisional patent. From Geiger and Sa’ (January 15, 2009), filings were more than 60 percent of disclosures in 2006 while only 30 percent in 1995. 16. A licensing option is an agreement where a potential licensee is granted a certain time period to evaluate the technology and negotiate the terms of a licensee agreement. 17. Geiger and Sa’ (January 15, 2009) report that for major research IHE in 2004, the average licensing officer at a TTO would handle 24 disclosures, 20 patent applications, and less than seven licenses. 18. There were 161 IHE respondents. The survey does not provide a definition between small and large companies. 19. Licenses can grant exclusive rights to specific “fields of use”. 20. University licensed inventions tend not to be ready for commercialization and require further development. Jensen,


40

Thursby and Thursby (2001) from a mid-1990s survey of major universities report that 48 percent of licensed inventions were simply a proof-of-concept and 29 percent were only a lab scale prototype. In addition, Jensen and Thursby (August 1998) report only 12 percent of inventions were at the stage for close-to-practical-use. 21. Generally, a standard benchmark for investor’s geographical proximity is “two-hour rule” (CEO Council for Growth, October 2007). 22. From Krattiger et al. (2007) as a rule of thumb, ten invention disclosures may lead to one patent, and one license might come from ten patents. In other words, only 10% of patents provide royalties. In the United States, the cost of a trademark (including attorneys’ fees) is approximately $1,200–$2,000. A provisional patent application costs $2,500–$8,000, and a non-provisional application costs $6,000–$30,000. The cost of filing and maintaining a patent globally is approximately $500,000. 23. AUTM (2008) reports that 7% of the disclosures submitted for FY2007 were licensed in the same year that they were received. The high percentage reflects the TTO’s ability to find compatible commercialization partners. 24. As of July 2008, the existing IHE TTOs within Pennsylvania are at Carnegie Mellon University, Drexel University, Duquesne University, Lehigh University, Pennsylvania State University, Temple University, University of Pennsylvania, University of Pittsburgh; and non-IHE TTOs are Fox Chase Cancer Center, Thomas Jefferson University, and Wistar Institute. 25. A total of $10 million was provided in the legislation for the three fiscal years of funding. Due to mandated budget reductions for FY 08-09, KIG awards were reduced from $3,500,000 to $3,364,427 resulting in about $9.9 million awarded. KIG funds cannot be used for the following: (a) to support general administrative overhead or indirect costs at KIZ participating IHEs; (b) fund travel outside of the country; (c) perform building construction or renovation; (d) planned long-term (beyond the grant year) support of staff salaries; and (d) support non-KIZ companies with either financial assistance or to support internships at non-KIZ companies (Pennsylvania Department of Community and Economic Development, February 2008).

REFERENCES Acs, Z.; Audretsch, D., and A. Varga. 2002. Patents and innovation counts as measures of regional production of new knowledge,” Research Policy, 31, 1069-1085.

Acs, Z.; Audretsch, D., and M. Feldman. 1994a. “R&D spillovers and recipient firm size,” The Review of Economics and Statistics, 76, 336-340. IBID. 1994c. “R&D spillovers and innovative activity,” Managerial and Decision Economics, 15, 131-138. Armstrong, T. Spring 2008. “Nanomics: The economics of nanotechnology and the Pennsylvania initiative for nanotechnology,” Pennsylvania Economic Review, 1-19. IBID; Bohl-Fabian, L., Smith-Aumen, A,, and Khalil Y. 2007. Economic development initiatives of the Pennsylvania State System of Higher Education, Pennsylvania Economic Association Proceedings, 83-91. IBID; Bernotsky, R., Lorraine; Bohl-Fabian, Lou; Loedel, H., Peter; and Khalil Yazdi. 2006. Economic impact study of the Pennsylvania State System of Higher Education, Pennsylvania Economic Association Proceedings, 20-29. IBID and Khalil Yazdi. 2004. “Keystone innovation zones: Technological transfer entrepreneurial zones for Pennsylvania.” Pennsylvania Economic Association Proceedings, 191-198. Association of University Technology Manager. 2004. “AUTM U.S. Licensing Activity Survey: FY 2004 Survey Summary,” 1-49. IBID. 2007. “AUTM U.S. Licensing Activity Survey: FY 2006 Survey Summary,” 1-44. IBID. 2008. “AUTM U.S. Licensing Activity Survey: FY 2007 Survey Summary,” 1-51. Bremer, H. January 2006. History of laws and regulations affecting the transfer of intellectual property, In: AUTM Technology Transfer Practice Manual, 3rd Edition. CEO Council for Growth. October 2007. “Accelerating Technology Transfer in Greater Philadelphia,” Economy League of Greater Philadelphia, 1-36. Douglass, J. A. 2006. Universities and the Entrepreneurial State: Politics and Policy and a Wave of State-Based Economic Initiatives. Research and Occasional Paper Series, Center for Studies in Higher Education: University of California, Berkley, CHSE. 14.06. Eastwood, B. D., Brooker, R. J., and D.E. Terry. December 1986. Household Nutrient Demand: Use of Characteristics Theory and a Common Attribute Model, Southern Journal of Agricultural Economics, 235-246.


41

Geiger, R.L. and C.M. Sa’. January 15, 2009. Tapping the Riches of Science: Universities and the Promise of Economic Growth, Harvard University Press. IBID. 2005. Beyond Technology Transfer: US State Policies to Harness University Research for Economic Development, Minerva 43(1): 1-21. Harris, R. K. and S.K. Smith. January 2006. Copyright Protection, In: AUTM Technology Transfer Practice Manual, 3rd Edition. Hermanns, K.R. and E.W. Wagner. January 2006. The Patent Application Process Outside the United States, In: AUTM Technology Transfer Practice Manual, 3rd Edition. Jensen, R. and M. Thursby. August 1998. Proofs and Prototypes for Sale: The Tale of University Licensing, NBER Working Paper 6698. IBID and J. G. Thursby. 2001. Objectives, Characteristics and Outcomes of University Licensing: A Survey of Major U.S. Universities, Journal of Technology Transfer, 26: 9-21. Krattifer, A., Mahonye, R.T., Thomson, J.A,, Bennett, A.B., Satyanarayana, K., Graff, G.D., Fernandex, C., and S.P. Kowalsik. 2007. An Executive Guide to Intellectual Property Management in Health and Agricultural Innovation: A Handbook of Best Practices. MIHR (Oxford, UK), PIPRA (Davis, US), Oswaldo Cruz Foundation (Fiocruz, Rio de Janeiro, Brazil), and bioDevlopments-International Institute (Ithaca, US). Available at www.ipHandbook.org. Link, A. N., Siegel, D.S., and B. Bozeman. May 2006. An Empirical Analysis ,f the Propensity of Academics to Engage in Informal University Technology Transfer, Rensselaer Working Papers in Economics, Rensselaer Polytechnic Institute, New York, (0610): 1-24. Mansfield E. 1991. Academic Research and Industrial Innovation, Research Policy, 20: 1-12. IBID. 1995. Academic Research Underlying Industrial Innovations: Sources, Characteristics and Financing, The Review of Economics and Statistics, 77: 55-65. Needle, W. January 2006. Trademark Primer, In: AUTM Technology Transfer Practice Manual, 3rd Edition.

Pennsylvania Department of Community and Economic Development. 2005. Pennsylvania TechFormation: A Status Report and Growth Strategies for Technology-Based Economic Development, Harrisburg, PA. IBID. 2007. Measuring Up Enhanced Metrics for a New Economy, Harrisburg, PA. IBID. February 2008. Keystone Innovation Grants Program Guidelines, Harrisburg, PA. Phan, P. H. and D. S. Siegel. 2006.The Effectiveness of University Technology Transfer, Foundations and Trends in Entrepreneurship. 2(2): 77-144. Resource Guide for Technology-Based Economic Development: Positioning Universities as Drivers, Fostering Entrepreneurship, Increasing Access to Capital. August 2006. Economic Development Administration, U.S. Department of Commerce. Rosenberg, N. and R. Nelson. 1994. American Universities and Technical Advance in Industry, Research Policy. 23: 323-348. Technology Transfer Tactics. August 2008. Economic Impact Study Offers Proof of Tech Transfer’s True Value, 2(8): 113-128. Triplett, J. 2004. Handbook of Hedonic Indexes and Quality Adjustments in Price Indexes: Special Application to Information Technology Products, Statistical Analysis of Science, Technology and Industry Working Paper, OECD. Varga, A. August 2002. Knowledge Transfers for Universities and the Regional Economy: A Review of the Literature, Available at: http://ephd.ktk.pte.hu/SURVEY12.pdf. Woodward, D. P. and P. Guimarães. May 2008. Economic Impact of the South Carolina Research Authority, Moore School of Business, University of South Carolina. Zucker, L. G. and M.R. Darby. March 2005. Soci-Economic Impact of Nanoscale Science: Initital Results and NanoBank, National Bureau of Economic Research Working Paper Series 11181.

APPENDIX: KIG RECIPIENTS, FUNDED, AND KIG DESCRIPTIONS

Beaver County Community College

FY 06-07$150,000

KIG funds will be used to create the KIZ Entrepreneurship and Technology Development program.

Bloomsburg University FY 05-06$200,000

The KIG initiative will be used to develop an on-line educational knowledge transfer and IP tool kit; to fund faculty and entrepreneurial staff development; and to hire staff in the TTO at

Geisinger Ventures.

Bucknell University

FY 06-07$150,000

FY 08-09$71,643

The Keystone Innovation Grant is to facilitate technology transfer activities associated with biomedical innovation, led by Bucknell University and Geisinger Health System, and

supported by the Bucknell Small Business Development Center that focuses on device prototyping.

Bucks County Community College

FY 08-09$191,539

The KIG support allowed for the development of a technology transfer entity (TTE), Launch Innovations PA. In this proposal, we build on that success working with Bucks County Community College. In addition, Bucks County Community College will implement an

internship program that meets the needs of the students and entrepreneurs at the Pennsylvania Biotechnology Center and the Bucks County KIZ. Students and faculty will be

engaged in technology creation projects. Secondly, the College will establish a Seed/Assistance Fund to provide product development dollars for research ideas, support

market feasibility studies of commercialization prospects; inject seed money for new business ventures; fund competitions to highlight research and create jobs.

Carnegie Mellon University

FY 05-06$250,000

FY 06-07$200,000

FY 08-09$191,539

Carnegie Mellon’s Center for Technology Transfer and Enterprise Creation (CTTEC) has a multi-year pilot project to further increase entrepreneurial activity through the addition of

enhanced services and the creation of CTTEC. Over the next year, CTTEC plan to enhance the Enterprise Creation process by (a) increasing the number of potential start-ups to bring through the process; (b) continuing and enhancing the outreach activities through additional

Partner Connection Conferences and increased engagement with Internal and External advisors and entrepreneurially minded groups; and (c) exploring and increasing research, commercial and potential spin-off synergies with the University of Pittsburgh and other KIZ

partners.

Delaware Valley College FY 06-07$200,000

KIG funding will defray the cost of a technology transfer professional to work at Delaware Valley College (DVC), focusing on identifying and stimulating technology transfer activities on

campus and within the KIZ.

Drexel University

FY 06-07$175,000

FY 08-09$191,539

During the second phase of the project, the success in matching Drexel University's technologies with the guiding hands of business development experts and industrial partners

will continue in four new or expanded areas. The focus of the current application is to continue funding Technical/Clinical and Commercial proof of concept work in coordination with the Coulter Foundation's Translational Partnership program and industrial partners,

pushing promising technologies towards commercialization as well as continuing the development of on-line and on-site public access to our Biomedical Product Development

Curriculum. Additionally, Drexel will expand the "commercialization system" to include translational research mentoring and community outreach.

East Stroudsburg University

FY 05-06$250,000

FY 06-07$150,000

The KIG will provide resoucres for continuation of seed funds to support technology transfer, commercialization and collaboration with the Pocono Mountains, Lackawanna/Luzerene

Zones.

Franklin & Marshall College

FY 05-06$250,000

FY 06-07$150,000

FY 08-09$143,638

The KIG submitted by Franklin & Marshall College will result in:- The continuation of a successful Central PA regional collaboration involving 13 higher

education institutions, two Keystone Innovation Zones, Ben Franklin Technology Partners and the Venture Investment Forum,

- Funding additional seed grants which encourage faculty to develop and expand on innovative projects, and

- Linking research and product development activity on higher education campuses with private sector partners.

KIG Recipient KIG DescriptionFunded


42

Lehigh University

FY 05-06$250,000

FY 06-07$200,000

FY 08-09$191,539

Lehigh University’s Office of Technology Transfer (OTT) continues to build a diverse portfolio of technologies available for commercial exploitation due to a significant degree through

financial support provided by the Commonwealth of Pennsylvania’s Department of Community and Economic Development. The emphasis in the third year of the KIG program

will be to expand the marketing and commercialization efforts of Lehigh’s OTT with the goal of establishing industrial partnerships that will drive further research and development as well as real life applications of university created intellectual property. The attainment of this goal will

ultimately result in jobs creation and economic development.

Lycoming College FY 08-09$46,615

Lycoming College's Keystone Innovation Grant will support the creation of the Center for the Study of the Community and the Economy (CSCE), which will provide direct client services, education outreach activities and student internships for Williamsport-Lycoming Keystone

Innovation Zone (WLKIZ) companies and clients. Each component will supplement existing WLKIZ services, bridge the gap between idea development and business start-up, strengthen

the informal collaborations between Lycoming faculty and the local region, and provide Lycoming students with enhanced learning opportunities. The initiatives in the KIG

application and the establishment of the CSCE at Lycoming College will help move the number of WLKIZ clients from idea development toward the creation of sustainable businesses, which will in turn strengthen the economic development of this region.

Marywood University FY 08-09$143,638

The purpose of the Northeast Commercialization Collaborative is to enhance technology transfer and commercialization in Northeastern Pennsylvania. This goal will be accomplished

by promoting collaborative scholarly activities among colleges and universities that will ultimately lead to the successful transfer of technology from institutions of higher education to

the marketplace. KIG funds will be distributed by the Lackawanna, Luzerne, and Pocono Mountains Keystone Innovation Zones in coordination with the Northeastern Pennsylvania Technology Institute. Funds will be used to continue the commercialization efforts funded

during FY05-06 and FY06-07 and to support new technology transfer projects underway at regional universities.

Mount Aloysius College FY 06-07$100,000

The funding through the KIG program is to further the development of a collaborative technology transfer process among the four GJKIZ schools, which also includes Pennsylvania

Highlands Community College.

PA College of Optometry (Salus University)

FY 06-07$150,000

The Keystone Innovation Grant will fund the establishment of the Lankenau Institute for Medical Research (LIMR) Chemical Genomics Center that will serve as a core facility for the

611 KIZ and beyond.

Pennsylvania College of Technology

FY 05-06$250,000

FY 06-07$200,000

The KIG funding will create a "Product Innovation Center", participate in four business plan contests, and continue support of "Plastics SourceNet".

Pennsylvania State System of Higher

Education (through East Stroudsburg University)

FY 08-09$191,5390

The KIG project will enable the Pennsylvania State System of Higher Education, PASSHE, to establish a Technology Transfer and Commercialization Resource Network that will initiate

technology transfer services for faculty, students and staff inventors at all 14 Universities. The Network will provide training, professional resources and systematic procedures to facilitate invention disclosure and marketing at PASSHE Universities. It will also raise awareness of

commercialization opportunities.

Pennsylvania State University - Main

Campus (State College)

FY 05-06$250,000

FY 06-07$200,000

FY 08-09$191,539

The goals of this KIG are to leverage the research, education, and technology commercialization capacity of the Penn State University Park Campus for innovation at two

selected Keystone Innovation Zones (KIZs): the I-99 Corridor KIZ and the Navy Yard KIZ, as well as for innovation in all KIZs across the Commonwealth. The KIG will support efforts to partner KIZ companies with Penn State faculty and students across academic disciplines,

launch new KIZ companies to commercialize Penn State technology, link KIZ companies to Penn State’s global industrial research partners, expand engineering design, innovation, and entrepreneurship education at all levels, and develop sustainable energy education, research,

and commercialization programs in partnership with KIZ companies.

KIG Recipient Funded KIG Description


43

Pennsylvania State University - DuBois

FY 08-09$191,539

As the primary Institution of Higher Education for the Tri-County KIZ, the KIG grant will enable the campus to: enable the Center for Technology Transfer and associated Group for

Entrepreneurial Development and Group for Academic Practicum by employing a Technology Transfer Coordinator; provide seed capital for business start-up, research and development,

proof of concept, licensing and commercialization; implement a strategic marketing campaign; provide internships to assist in technology transfer projects; and provide educational

opportunities relative to the programming needs associated with innovation, entrepreneurship, and technology transfer.

Pennsylvania State University - The Behrend

College

FY 06-07$150,000

FY 08-09$143,638

The four universities making up the Erie Keystone Innovation Grant Team, Edinboro, Gannon, Mercyhurst and Penn State Behrend, are growing technology transfer services and programs from the ground up. Such activities are filling the gap between technology development and

commercialization by harnessing the resources of faculty, students, alumni and entrepreneurship curricular development. As the Erie universities make the shift between

traditional education and research objectives to entrepreneurial culture and business development, the Keystone Innovation Grant is providing funding to accelerate this transition, and will foster the growth of new technology companies employing Pennsylvania-graduated

students.

Pennsylvania State University - Harrisburg

FY 05-06$250,000

FY 06-07$200,000

FY 08-09$191,539

They Keystone Innovation Grant project submitted by Penn State Harrisburg for the Innovation Transfer Network will: drive technology transfer activity among 13 universities and colleges in South Central Pennsylvania; connect private sector business leaders with faculty at institutes of higher education in the region to accelerate commercialization; and provide a

networking exchange of information to support entrepreneurial activity between business, institutes of higher education and economic development.

Robert Morris University

FY 06-07$75,000

FY 08-09$143,639

This Keystone Innovation Grant project involves acquisition of the FARO ScanArm, the Creaform HandyScan 3-Dimensional laser scanner, and the Mimics Medical Imaging software that will allow RMU-CARES to expand it's state-of-the-art laboratories. The Center will provide a wide range of prototyping/reverse engineering, technology transfer, and education/training services to Pennsylvania companies, particularly in the ever growing bio-medical industry.

Saint Francis University

FY 05-06$250,000

FY 08-09$95,770

Saint Francis University and Mount Aloysius College Keystone Innovation Grant initiative funds will be used to in relation to the following initiatives: 1) The Entrepreneurship

Collaborative will promote collaboration and cooperation between students of the two schools to increase entrepreneurial activity in the region; 2) Mount Aloysius will expand

entrepreneurship activities and continue their Abandoned Mine Drainage research; 3) Saint Francis will implement the work already done for a Center for Technology Entrepreneurship;

and 4) CERMUSA at Saint Francis University will continue commercialization work on several of its most promising projects.

Saint Vincent University FY 08-09$95,746

The Keystone Innovation Grant will help to establish the Saint Vincent College Innovation Center, a project focused on the introduction and development of systemic produce and

process innovation applications. The Innovation Center will foster the growth and development of new and existing businesses by providing them access to the technology, business knowledge, and operational excellence capability housed within Saint Vincent

College and the resources at other Westmoreland County KIZ institutions, including Pennsylvania State University New Kensington, Seton Hill University, University of Pittsburgh

at Greensburg, and Westmoreland County Community College.

Susquehanna University FY 08-09$95,770

Susquehanna University’s KIG will promote entrepreneurship both on campus among faculty and students from a variety of disciplines, and in the region, with particular emphasis on start-ups and small businesses located within the Greater Susquehanna KIZ. The grant will support

outreach, education and programming for entrepreneurship and the development of a campus-based infrastructure, including asset/needs assessments, to help stimulate the

movement of ideas and innovation to commercially viable products or services.

Temple University

FY 06-07$150,000

FY 08-09$181,664

The non-profit Center for Human Antibody Therapeutics (CHAT) being established as a core facility at the Lankenau Institute for Medical Research (LIMR) will use a human monoclonal antibody production technology to facilitate the validation of intellectual property contributed

by the BioLaunch611+ KIZ partner institutions, so that technologies using antibodies for diagnostics and therapeutics for the treatment of cancer can be more efficiently transferred to

the healthcare marketplace.

Funded KIG DescriptionKIG Recipient


44

Thomas Jefferson University

FY 08-09$95,770

The KIG will enable TJU and USP to build on the successful buddy-mentor program and transition it into a partner system in which TJU's Office of Technology Transfer will work as a

partner of the Office of Technology Management (OTM) at USP to evaluate, seek patent protection, and market and license new USP technologies. In addition, USP's OTM will begin

to take the lead in coordinating educational outreach events outlining OTM's services. The two offices will also work together to connect their faculty to potential commercial partners

through research collaborations and technology packaging opportunities.

Thomas Jefferson University

(TJU)/University of Sciences in Philadelphia

(USIP)

FY 06-07$100,000

The KIG will fund the USP-TJU buddy-mentor system to establish a technology transfer office at the USP to commercialize USP faculty research results. Utilizing TJUs existing transfer

team's experience the mentor team will establish a tech transfer office at USIP.

University of Pennsylvania

FY 05-06$250,000

FY 06-07$150,000

FY 08-09$95,770

The University of Pennsylvania will use the new Keystone Innovation Grant to launch the Commercialization Acceleration Program (CAP). Building off of a FY2008 pilot, CAP will include students from across the University and across disciplines. Students will obtain

entrepreneurship experience by engaging in experiential education projects that are based on Penn investigator discoveries. CAP will introduce a formal oversight board and steering

committee as well as new types of research commercialization projects in FY2009.

University of Pittsburgh

FY 05-06$250,000

FY 06-07$200,000

FY 08-09$191,539

This grant funding supports efforts to develop: more effective “triaging” strategies and tools for assessing the commercial potential of Pitt technologies, as well as provide pre-

commercialization gap funding; new business development strategies and programs that encourage more effective interaction between faculty innovators and potential industry partners; and the Pitt Innovator Initiative to educate and engage more faculty, staff and

students in the commercialization process, and create a proactive community of innovators.

University of the Sciences

FY 05-06$50,000

KIG funding will be used for development of universal IP policy for life sciences to be adopted by the University City KIZ partners and for workshops/training for University of Sciences

faculty and students.

University of the Sciences in Philadelphia/

Wistar Institute

FY 06-07$150,000

KIG funds will be used to strengthen the region's drug discovery and development capabilities through acquiring a cutting-edge High Throughput Screening and Development (HTSD)

system to be located and operated at the Wistar Institute.

Villanova University

FY 06-07$150,000

FY 08-09$143,638

The second year KIG effort will develop Villanova and Widener University projects seed-funded in the first year effort to the next stage of commercialization. Business development,

including securing financing, intellectual property protection, prototype technology development, and market studies will be achieved through a technology transfer function to

be established at Villanova University. Additional new projects will also be seed-funded through the grant funding.

Wilkes University

FY 05-06$250,000

FY 06-07$150,000

FY 08-09$143,638

KIZ Northeast Commercialization Collaborative, NEC2, aims to support projects at various stages along the path to commercialization and technology transfer associated with targeted

KIZ industry sectors. These sectors will play an important role in ongoing economic development in NEPA and include life sciences/biotechnology, homeland security,

nanotechnology/advanced manufacturing/plastics, healthcare, financial services/back office finance, and information technology/new media.

KIG Recipient Funded KIG Description


45


46

EXAMINING THE EXISTENCE AND EXTENT OF ANTI-INTELLECTUAL ATTITUDES AMONG UNIVERSITY STUDENTS

Robert S. Balough

Department of Economics Clarion University of Pennsylvania

Clarion, PA 16214

ABSTRACT

The existence of social and peer pressure on students to perform below one’s ability academically in order to be socially accepted has been recognized for some time. Previous studies have found attitudes of anti-intellectualism to exist to varying extents among students of all racial/ethnic backgrounds but to be particularly problematic for minority students attempting to advance themselves. This behavioral phenomenon is known among minorities as “acting white” and, in certain circumstances, can have a significant impact on behavior because of social/external costs of performing well academically. Unlike previous studies that have focused on this problem in the behavior of secondary students, this paper examines the extent to which anti-intellectual attitudes exist among university students.

The extent of the phenomenon is measured over various strata of the university student population using survey data. Differences in identified anti-intellectual attitudes are estimated across ethnic and racial categories as well as such characteristics as sex, family income, fraternity/sorority membership, economic background, and sports participation. These findings are largely at odds with previous studies of high school students. While racial differences in these attitudes are not as evident at the university level compared to previous studies at the secondary level, there are significant differences present among other factors examined.

LITERATURE REVIEW

Sociologists have long known that behavior of individuals is often shaped by the norms of the groups with which they associate. Similarly economists view behavior and changes in behavior as economic situational responses, that is, as a response to costs and benefits of that behavior. These costs and benefits are often determined to greater or lesser extent by the groups with which individuals associate. Individuals without experience or knowledge in an area will often seek advice prior making decisions. Households accumulating saving dollars, for example, will seek advice on how to invest those funds in order to earn the highest return possible given their preference toward risk. These decision makers will generally have some ability to correctly assess and information given to them prior to making decisions. Younger individuals learning social skill also seek advice.

They seek advice, however, from peers who often are no more knowledgeable than they. They are also generally not able to assess information about behavioral costs and benefits due to lack of experience and knowledge and do not, therefore, discount this information at an appropriate rate.

Anti-intellectualism can have many connotations in American society. At an aggregate level, the concept was well defined by Richard Hofstadter (1963) in his Pulitzer Prize-winning book entitled Anti-Intellectualism in American Life, written in part as a response to the McCarthy era of American politics. During the past eight years, however, anti-intellectualism has enjoyed a resurgence under President George W. Bush. Statements attributed to the President about his not having read a book since graduating college, his statement at the Yale 2001 commencement regarding his success in spite of being a “C” student, and his many misstatements over the years has made, one could argue, an anti-intellectual climate acceptable.

Anti-intellectualism in education is considered common and prevalent. Limited research, however, has been conducted at the higher education level. Trout (2007) and Spann and Davison (2004) provide interesting summaries and a great deal of insight into anti-intellectualism among college students, but do not offer an empirical investigation.

Anti-intellectualism among minority students, on the other hand, is well documented. The phenomenon is often referred to as “acting white.” John Ogbu and Signithia Fordham (1986) discovered, in a Washington D.C. high school that many black students shunned high academic performance and earning good grades because their peers considered such behavior as “selling out” or “acting white.” Being a high achiever, therefore, was costly, namely not being accepted socially by one’s peers. Horvat and Lewis (2003) suggested that an ‘oppositional culture’ exists among minority students that defines social interactions and influences the manner in which individuals assimilate into the dominant culture. According to this theory anything viewed as being a characteristic of the dominant culture is disavowed in order to maintain cultural identity. Intellectualism and academic achievement somehow became associated with the white dominant culture and was therefore shunned by the minority culture in favor of pursuits perceived to be more likely to

yield reward, such as sports. Anti-intellectualism among black students is, in this theory, a result of the white culture’s historical domination in education and intellectualism in general. In contrast to this oppositional culture theory, John McWhorter (2000) argues that black students have sabotaged themselves with what he calls “victimology.” Being a victim of past discrimination in educational opportunity combined with the attitude that education and intellectualism is for the white dominant culture, has left black students unwilling to compete academically. This, he argues, has resulted in the persistence of an achievement gap between black and white students even in the face of greatly improved educational opportunities for black students.

Various other studies have examined the differences between the academic achievements of blacks and whites. Ronald Ferguson (2001) examined students in Shaker Heights, Ohio and identified a strong anti-intellectual attitude among black students even in this integrated upper-class Cleveland suburb. Instead of exploiting the advantage black students may have had in such a situation, they shunned learning and achievement. This Shaker Heights phenomenon has become virtually synonymous with “Acting White.”

Studies by Fryer and Torelli (2005) and Austen-Smith and Fryer (2003) have made significant contributions to this research taking a more economic view of behavior. Their research attempts to measure the cost to high school students in terms of social acceptance and the number of friends they would have based upon academic achievement. Their research indicated that for white students earning higher grades more often resulted in greater popularity and for black students higher grades led to only modestly higher popularity. They not only identified a diminishing return in terms of popularity to increased academic achievement, they also identified a saturation point, in that, black student popularity increased up to about a GPA of 3.5 but then began to decrease as the GPA approached the 4.0 mark. Hispanic student saturation point was even lower, at about a GPA of 2.5. This popularity saturation point did not exist for white students in the study. Another interesting finding of Fryer and Torelli was that this “acting white” anti-popularity phenomenon was virtually non-existent for black student who attended predominately black high schools.

RESEARCH METHODOLOGY

The purpose of the present research effort is to attempt to measure the extent to which anti-intellectual attitudes exist among university students. Using survey data, the extent of the phenomenon is measured over various strata of the university student population. Differences in identified anti-intellectual attitudes are estimated across ethnic and racial categories as well as such characteristics as sex, family income, fraternity/sorority membership, economic

background, and sports participation. The survey instrument used is presented in the Appendix A. The questionnaire asked 14 questions regarding demographics and thirty six questions answered on a five point Likert agreement scale. Appendix B presents the responses to the demographic questions. For this analysis Sex, Race, and Sports Participation will be examined. For ease of analysis, the response scale for the non-demographic questions was reduced to three levels; agree, neutral or no opinion, and disagree. The instrument was administered during the spring 2009 semester at Clarion University of Pennsylvania. Sampling was conducted using a stratified cluster technique. Several highly populated courses at the university were selected to represent students from various colleges and class levels. The instrument was administered to randomly-selected individual sections of each of these courses. Student participation was voluntary and dependent upon attendance on the day the instrument was administered. Administration of the instrument was conducted by one of two faculty members, both reading the same script to student participants.

While the instrument included 36 experimental opinion questions, only fourteen of these questions are examined here as an indication of the respondent’s anti-intellectual attitudes and two are used as an indication of whether the student has observed anti-intellectualism on the part of others. Agreement with the statements in questions 18, 24, 27, 28, 33, 35, 43, and 44 were taken to indicate that the respondent displayed an anti-intellectual attitude of some degree. Disagreement with question 21 indicated the same. Questions 19, 45, 46, 47, and 50 ask the students if they have observed others being harassed, have been the perpetrators of harassment, or are sensitive to harassment by others.

Table One below reports the proportionate responses to the questions examined after reduction from a five to a three point scale. Questions where agreement would indicate an anti-intellectual attitude have been bolded.

As can be seen in the table, the level of agreement with the bolded questions varies and ranges from 4.9% to 17.1%. Roughly 5 to 20% of students indicated some degree of an anti-intellectual attitude in the application of the survey. While these numbers are significantly greater than zero, they do not indicate widespread or pervasive problem of anti-intellectualism at this university. If student attitudes at Clarion University are representative of university students in general, these results indicate the existence of, but not a pervasive problem with, anti-intellectual attitudes among students at the higher education level as was found with many studies at the secondary level.


47

RESULTS

The specific hypotheses to be tested are as follows: Student anti-intellectual attitudes as measured by the individual question examined are independent of the sex of the respondent. Stated another way, the proportion of respondents expressing agreement, neutrality, and disagreement are the same for males and females. The test used is the Chi-square test of concordance. Table Two below gives the Chi-square test statistic and the observed significance level (p-value) for each opinion question considered relative to sex. For those relationships found to be significant at the 5% level or better an agreement ratio, the ratio of males to females that agree with the question, is also reported. Males in the sample, for example, were 3.5 times more likely to agree with question 18 than females. As before, those questions where agreement is associated with anti-intellectualism are bolded.

The results here show that proportionate responses to several questions related to anti-intellectual attitudes are significantly different based upon sex of the respondent. As the agreement ratios indicate males tend to show stronger anti-intellectual attitudes than females. Some of these results are striking such as questions 18, 34, and 35 where men are 3.5 to 4.7 times more likely to be in agreement with the anti-intellectual statement. Interestingly, however, males are 1.7 times more likely to agree with question 50; indicating males are more compelled to mask their academic effort than females. Males are significantly more likely to be harassers (question 42) than females and females are significantly more likely to have been harassed (question 45) than males.

In addition to sex of the respondent, race of the respondent is also examined here. As seen in Appendix A, the variable race has five levels of responses. Of the 350 total respondents, 302 were white, 28 black, and the remaining 20 spread over the other three categories. The lack of responses for these other categories made a five-level comparison impractical. Combining white with Hispanic and black with the others was considered, but the analysis provided here examines only the 330 observations of black and white students. Reducing the variable race to two levels improves the reliability of the test statistics reported. The results for the three levels of agreement against the two levels of race are given in Table Three.

Unlike sex, race of the respondent led to significantly different responses in only three questions, only one of which is used to measure anti-intellectualism. That question, number 27, states, “Earning higher grades will not make it easier to find a job after graduation.” Blacks were 1.31 times more likely to agree with this statement than white students. This could be a reflection of black student perceptions of the

impact of race in the job market as much as it is a reflection of an attitude toward anti-intellectualism.

Lastly, sports participation is examined. There is a common perception that athletes shun academic pursuits and attend college primarily for athletic opportunities. This perception is perhaps more common in NCAA Division I schools with large athletic programs, but is not non-existent in smaller schools. The results of the test of the hypotheses regarding sports participation are given below in Table Four. While the results here are not as pronounced as for the variable sex, there are several questions that are significant. Athletes are significantly more likely to respond in agreement with 28, 33, 35, 43, and 44. All of these questions were among those used to indicate an attitude of anti-intellectualism. Athletes were 3.5 times more likely to disapprove of class participation than non-athletes (question 33), were 1.5 times more likely to believe high-achieving students hurt the grades of other students (question 28), 2.5 times more likely to disapprove of friends that work hard academically (question 35), and 2.8 times more likely to agree that having friends and looking good is more important than working hard (question 44).

DISCUSSION AND CONCLUSION

This research sought to determine the extent to which anti-intellectual attitudes exists among college students and whether these attitudes are related to characteristics such as sex, race, and athletic participation. Strong evidence of anti-intellectualism found in many studies of high school students is not generally supported for students participating in this study.

The “acting white” phenomenon, strongly evident in public high schools in previous studies, is not found to be supported at this public university as race was generally not found to lead to a significant difference in attitudes toward anti-intellectualism. Several reasons could explain this result. One possibility, indicated in Appendix B, is that approximately half of the black students participating in this study attended predominantly black high schools where, according to previous studies, a social stigma toward black student achievement is typically nonexistent. Another possibility is the logical explanation that black students attend college for the same reasons white and other students do, to graduate and improve their chances of leading productive lives. In other words, finding that black students are generally no more likely to display anti-intellectual attitudes while in college than their whiter counter parts, is what one should hope to find in a study such as this and not expect it to be an anomoly needs to be explained.

This study did find, however, that to the extent that anti-intellectual attitudes do exist, they are not equally evident in males and females. Males were found to be more likely to


48


49

display anti-intellectual attitudes than females. The same was found for students who participated in college athletics. Weak evidence is found for differences in anti-intellectual

attitudes based upon family and no evidence is found for different class levels and student employment status.


50


Appendix A – Survey Instrument

1. Sex A. Female B. Male 2. Class Standing A. Freshman C. Junior E. Grad B. Sophomore D. Senior 3. Race A. American Indian B. Asian C. Black D. Hispanic E. White 4. Current overall grade point average A. Below 2.5 C. 3.00-3.49 E. 4.00 B. 2.50-2.99 D. 3.49-3.99 5. How would you describe the high school you attended? A. Predominantly White B. Predominantly Black C. Racially mixed D. Other / Unknown 6. How would you describe the high school you attended? A. Urban B. Suburban C. Rural D. Other / Unknown 7. How would you describe the high school you attended? A. Public B. Private 8. Approximately how large is the high school you attended? A. < 100 in graduation class B. 100 to 500 in graduation class C. > 500 in graduation class 9. Which of the following best describes your family’s A. < $25,000 income? B. $25,000 to 49,999 C. $50,000 to 74,999 D. $75,000 to 99,999 E. > $100,000 10. How would you describe your socio-economic class A. Upper class growing up? B. Upper-middle class C. Middle class D. Below middle class 11. What is the highest level of education attained by either A. Master’s, professional degree, or higher of your parents? B. Bachelor’s degree C. Associate degree or some college D. High school diploma E. Did not complete high school 12. Do you now or did you ever participate in college sports?

A. Yes B. No

13. Do you work while attending school? A. Full time B. Part time C. Not employed 14. Are you a member of a fraternity or sorority? A. Yes B. No


51


52

Please rate your level of agreement with each of the following statements using the following scale: A – Strongly Agree; B – Agree; C – Neutral/No Opinion; D – Disagree; E – Strongly Disagree.

Part B – College Study Habits15. I believe I do best in classes in which I study alone. A B C D E 16. I prefer to sit in the front of class even if my friends sit in the back. A B C D E 17. The best way to get through college is to put out the least amount of effort possible. A B C D E 18. I do what I need to get by in my classes. Graduating is important, grades are not. A B C D E 19. I find that taking notes in class helps me do well in my classes. A B C D E 20. I like to take classes with my friends because we help each other succeed. A B C D E 21. A higher grade average helps you get a better paying job after graduation. A B C D E 22. I try very hard to be on time for the start of each of my classes each semester. A B C D E 23. I always try to find group study partners in the classes I take. A B C D E 24. I could get better grades in some of my classes but it doesn’t really pay to do the extra work. A B C D E 25. Graduating from college is the best way I can advance myself and to have a better life than my

parents. A B C D E

26. Regular attendance in a class usually doesn’t help improve my grade in the class. A B C D E 27. Earning higher grades will not make it easier to find a job after graduation. A B C D E 28. No one likes the students who study too much and hurt the grades of the rest of the class. A B C D E 29. Most other students in my classes with whom I study are of the same race as me. A B C D E 30. My racial/ethnic status will keep me from getting a good job regardless of how well I do in college. A B C D E 31. I usually put extra effort into a class only if there is a chance I can raise my grade. A B C D E 32. My social background and status will keep me from getting a good job regardless of how well I do in

college. A B C D E

33. Students who participate in class cause problems by making the other students look dumb. A B C D E 34. Most of my friends seem to study less for their classes than I prefer to do. A B C D E 35. I don’t like to have friends that put too much importance on studying and grades. A B C D E

Part C – Friendships and Social Behavior36. Most of my friends are the same race as me. A B C D E 37. I interact with my friends on a daily basis and rarely have a day to myself. A B C D E 38. I have difficulty studying or working in groups when the other students are same-race peers. A B C D E 39. Most people that know me would say that I am somewhat of a loner. A B C D E 40. I had more friends who were of other races than me in high school than I do in college. A B C D E 41. I am open to making new friends in my classes, in student clubs, or wherever I am. A B C D E 42. On at least one occasion I have harassed a friend for working too hard on school work. A B C D E 43. I have friends who waste a lot of their time studying too much. A B C D E 44. Working hard is not as important as having friends and looking good. A B C D E 45. On at least one occasion I have been harassed by my friends for working too hard. A B C D E 46. I am sensitive to criticism from my peers about trying to do too well. A B C D E 47. In high school students who worked hard to get good grades were often chastised for trying to be

something they are not, that is, trying to get above those with whom they were raised. A B C D E

48. I have lost at least one friend since coming to college because I have been viewed as being above my identity.

A B C D E

49. I tend to try to associate with peers who place the same importance on success as I do. A B C D E 50. I often try to not let my friends know if I am working hard at something for fear they will ridicule my

efforts. A B C D E

Appendix B – Sample Profile

Cumulative Cumulative Q1 SEX Frequency Percent Frequency Percent ---------------------------------------------------- Female 195 55.7 195 55.7 Male 155 44.3 350 100.0 Cumulative Cumulative Q2 CLASS Frequency Percent Frequency Percent ------------------------------------------------------- Freshman 162 46.3 162 46.3 Sophomore 76 21.7 238 68.0 Junior 33 9.4 271 77.4 Senior 79 22.6 350 100.0 Cumulative Cumulative Q3 RACE Frequency Percent Frequency Percent ------------------------------------------------------------ Black/Other 37 10.7 37 10.7 White/Hispanic 309 89.3 346 100.0 Frequency Missing = 4 Q3 Cumulative Cumulative RACEBW Frequency Percent Frequency Percent ---------------------------------------------------- Black 28 8.5 28 8.5 White 302 91.5 330 100.0 Frequency Missing = 20 Cumulative Cumulative Q4 QPA Frequency Percent Frequency Percent ------------------------------------------------------- Below 2.5 64 18.3 64 18.3 2.50-2.99 113 32.4 177 50.7 3.00-3.49 103 29.5 280 80.2 3.50-4.00 69 19.8 349 100.0 Frequency Missing = 1 Cumulative Cumulative Q5 HS5 Frequency Percent Frequency Percent ------------------------------------------------------------ Predomin White 277 79.4 277 79.4 Predomin Black 14 4.0 291 83.4 Racially mixed 55 15.8 346 99.1 Other/Unknown 3 0.9 349 100.0

Frequency Missing = 1 Cumulative Cumulative Q6 HS6 Frequency Percent Frequency Percent ----------------------------------------------------------- Urban 46 13.2 46 13.2 Suburban 112 32.2 158 45.4 Rural 172 49.4 330 94.8 Other/Unknown 18 5.2 348 100.0 Frequency Missing = 2 Cumulative Cumulative Q7 HS7 Frequency Percent Frequency Percent ----------------------------------------------------- Public 324 92.6 324 92.6 Private 26 7.4 350 100.0 Cumulative Cumulative Q8 HS8 Frequency Percent Frequency Percent --------------------------------------------------------- < 100 grads 101 28.9 101 28.9 100 to 500s 208 59.6 309 88.5 > 500 grads 40 11.5 349 100.0 Frequency Missing = 1 Cumulative Cumulative Q9 FAMINC Frequency Percent Frequency Percent ------------------------------------------------------------- < $25,000 42 12.1 42 12.1 $25K to $49,999 88 25.3 130 37.4 $50K to $74,999 104 29.9 234 67.2 $75K to $99,999 55 15.8 289 83.0 > $100,000 59 17.0 348 100.0 Frequency Missing = 2 Cumulative Cumulative Q10 SOCCL Frequency Percent Frequency Percent ------------------------------------------------------------------- Upper class 7 2.0 7 2.0 Upper-middle 88 25.2 95 27.2 Working class & below 254 72.8 349 100.0 Frequency Missing = 1


53


54

Cumulative Cumulative Q11 PARED Frequency Percent Frequency Percent -------------------------------------------------------------- Master or higher 60 17.1 60 17.1 Bachelor degree 88 25.1 148 42.3 Assoc/some Coll 82 23.4 230 65.7 HS or less 120 34.3 350 100.0 Q12 Cumulative Cumulative SPORTS Frequency Percent Frequency Percent ---------------------------------------------------- Yes 84 24.0 84 24.0 No 266 76.0 350 100.0 Cumulative Cumulative Q13 WORK Frequency Percent Frequency Percent ---------------------------------------------------------- Full time 25 7.1 25 7.1 Part time 182 52.0 207 59.1 Not employed 143 40.9 350 100.0 Q14 Cumulative Cumulative GREEK Frequency Percent Frequency Percent --------------------------------------------------- Yes 33 9.4 33 9.4 No 317 90.6 350 100.0

REFERENCES

Ainworth-Darnell, J. and D. Downey, 1998. Assessing the Oppositional Culture Explanation for Racial/Ethnic Differences in School Performance, American Sociological Review, LXIII, 536-553. Akerlof, G. and R. Kranton, 2000. Economics and Identity, Quarterly Journal of Economics, CXV, 715-753. Austen-Smith, David, and Roland Fryer, Jr, 2003. The Economics of ‘Acting White’, National Bureau of Economic Research Working Paper No. 9904. Austen-Smith, David and Roland Fryer, 2005. An Economic Analysis of ‘Acting White’, Quarterly Journal of Economics, Vol. 120, No. 2: 551-583.

Cook, Phillip and Jens Ludwig, 1998. The Burden of Acting White: Do Black Adolescents Disparage Academic Achievement?, In Christopher Jenks and Meredith Phillips, eds., The Black-White Test Score Gap, Brookings Press. Cutler, David, and Edward Glaeser, 1997. Are Ghettos Good or Bad?, Quarterly Journal of Economics, CXII, 827-872. Ferguson, Ronald, 2001. A Diagnostic Analysis of Black-White GPA Disparities in Shaker Heights, Ohio, Brooking Papers on Education Policy, 347-414. Fordham, Signithia, 1996. Blacked Out: Dilemmas of Race, Identity, and Success at Capital High, University of Chicago Press. Fordham, Signithia and John Ogbu, 1986. Black Students, School Success: Coping with the Burden of Acting White, The Urban Review, XVIII, 176-206. Fryer, R. and P. Torelli, 2004. Understanding the Prevalence and Impact of ‘Acting White,’ Working paper, Harvard University. Fryer, R., and P. Torelli, 2005. An Empirical Analysis of ‘Acting White’, National Bureau of Economic Research Working Paper No. 1134. Hofstadter, Richard, 1963. Anti-Intellectualism in American Life, Knopf. McWhorter, John, 2000. Losing the Race: Self-Sabotage in Black America, Free Press. Ogbu, John and Astrid Davis, 2003. Black American Students in an Affluent Suburb: A Study of Academic Disengagement, Lawrence Erlbaum Associates, Inc. Spann, G. and B. Davison, 2004. Anti-intellectualism in the New Century, Paper presented at the annual meeting of the American Sociological Association, Hilton San Francisco & Renaissance Parc 55 Hotel, San Francisco, CA. http://www.allacademic.com/meta/p109121_index.html Trout, Paul, 2007. Student Anti-Intellectualism and the Dumbing Down of the University, Adjunct Nation. http://mtprof.msun.edu/spr1997/TROUT-ST.html.

THE STUDY OF VIETNAMESE SMALL BUSINESS OWNERS IN AMERICA: THEIR PATTERNS OF STRATEGIC DECISIONS

Hung M. Chu

Department of Management West Chester University of Pennsylvania

West Chester, PA 19383

Lei Zhu Department of Economics and Finance

West Chester University of Pennsylvania West Chester, PA 19383

ABSTRACT Data were collected from 279 Vietnamese American entrepreneurs in Philadelphia, Pennsylvania; San Jose, and Orange county California and Houston, Texas. Findings showed that the main reasons for business ownership were to be independent, obtaining job security, being able to utilize past experience and training, maintaining personal freedom and achieving personal satisfaction and growth. Regarding the factors contributing to their business success, Vietnamese American entrepreneurs indicated that hard work and charisma, friendliness to customers are the two most crucial variables. Good location, appropriate training, good customer service and family support are also the conditions leading to business success. Among the problems encountered by Vietnamese entrepreneurs, excessive competition, lack of formal management training, unsafe location and unreliable and undependable employees and discrimination from customers were also reported.

INTRODUCTION The collapse of South Vietnam in 1975, created an exodus of approximately 3 million people who fled their homeland and scattered around the world. To date there are 1.6 million (Wikipedia, 2006) Vietnamese refugees resettled in the United States. It has been reported that there were only a few thousand of Vietnamese who lived in America before 1975. They were primarily students who came for their education, members of Vietnamese diplomatic corps, spouses of American civilians and military personnel who served in Vietnam. Data provided by the State Department showed that Vietnamese refugees in America account for 1.5 million in 2005 as compared to 1.2 million in 2000 (Wikipedia, the free encyclopedia 4/9/2009).This figure has placed Vietnamese Americans as the fourth largest minority group in the U.S. behind the Chinese (2,400,000), Filipinos (1,800,000), and Indians (1,600,000). A majority of these refugees live in metropolitan areas in the West and the Southwest, especially in Orange County and San Jose, CA, and Houston, TX. Some

of them have also settled in Philadelphia, PA., Alexandria, VA., Minneapolis, MN, and Seattle, WA. as well. Quickly assimilated into the mainstream of American society, quite a few Vietnamese refugees have opened their own businesses. For instance, in Santa Clara County which includes San Jose, there are more than 5,000 Vietnamese American business owners, according to De Tran publisher of the weekly Viet Mercury. Nguoi Viet Yearbook (2005) reported that there were 4,500 Vietnamese small businesses in Orange County in 1996 but at the end of 2004 this same county was the home of 10,389 Vietnamese owned businesses. Although there are no reliable statistics on Vietnamese-American entrepreneurs in the United States, it is believed that the self-employment rate among this minority group is to be high. This study attempts to discover some critical motives for Vietnamese Americans to become business owners and the problems encountered when opening and running their businesses. Factors leading to their success are investigated as well. The role of family and friends in the foundation and expansion phases of a business is also examined. Furthermore, results of this study may provide policy makers with some helpful information in fostering an environment conducive to the growth of self-employment among this minority group.

LITERATURE REVIEW It is commonly understood that individuals who seek business ownership as their career may do so for various reasons. Some indicated that obtaining high income and providing jobs for family members are the most important motives (Pisturi et al, 2001). Results of a study by Swierczek and Ha (2003) showed that to many Vietnamese entrepreneurs, challenge and achievement are more important motivators than were necessity and security (Swierczek and Ha, 2003). A study of motivation by Benzing, Chu and Callanan (2005) revealed some differences in motivation


55

among regions of Vietnam. Entrepreneurs in Ho Chi Minh City are motivated by personal satisfaction and growth. In Hanoi, small business owners cited job creation as their most important factor leading to business ownership. The census bureau report (Philadelphia Inquirer, 1995) indicated that although the education level attained by Asian American males is usually higher than that of non-Hispanic white males, their earnings are not at par with their counterparts. This may be due in part to the glass ceiling encountered by Asian males in their workplace. Further, the stereotype of Asian males as team players not team leaders would not help their chance of advancement. Discouraged by the lack of opportunity in their career path, many of them strike out on their own and become business owners. Researches on the factors leading to business success revealed that there is a strong link between managerial skills and business success (Chu, Benzing & McGee, 2007; Benzing, Chu & Szabo, 2005; Yusuf, 1995; Gosh, Kim & Meng, 1993). According to Huck and McEween (1991) Jamaican entrepreneurs’ success depend on their understanding of customers’ need, access to capital, support of family, and networking with friends from former schools and colleges. It should be noted that the importance of family to the success of a business in China can never be emphasized enough. Due to an extremely low level of funding available to small and medium- sized enterprises in the country, family members not only are the source of start-up funds, but entrepreneurs’ wives and children are often asked to work when no reliable employees can be found (Liao and Sohmen, 2001). Vietnamese entrepreneurs however believe that friendliness to customers, having a good product at a competitive price, good customer services and a reputation for honesty are especially important factors leading to business success (Chu and Benzing, 2004). Results from a study of more than 300 small enterprises in 69 countries (Kisundo, Brunatti & Wilder, 1999) suggested that the problems encountered by entrepreneurs were quite similar. High taxes and tax regulations are the most serious in South and Southeast Asia. Inadequate infrastructure, inflation, labor regulation, and laws governing the starting and operating of a business were also considered as critical problems facing business owners. Lack of infrastructure, corruption, high tax, tax regulations and financing were said to be critical for micro and small enterprises (MSEs) in the Middle East and North Africa. Central and European entrepreneurs cited high taxes and tax laws, financing, corruption, and inflation as the most important road blocks to business success. Small business owners in Latin America indicated corruption and inadequate infrastructure, crime and theft, financing and tax regulations among the worst problems encountered. The most critical challenges facing SMEs in Sub-Sahara included corruption, complicated tax laws, bad infrastructure, inflation, theft and lack of capital. In

a study of Ethnic minority entrepreneurs in England, Fadahunsi, Smallbone and Supri (2000) found that the most critical problems encountered include the lack of financial sources, regulatory requirements, access to markets, discrimination by finance providers and language. Some of these problems may be considered as challenges faced by Vietnamese American entrepreneurs as well. Further discussion may be seen in the result section.

RESEARCH METHODOLOGY 279 entrepreneurs in San Jose, and Orange County, California, Houston, Texas and Philadelphia, Pennsylvania were randomly selected for this study. Community leaders were enlisted in administering the survey. Yearbooks or business directories of the communities were used. The administrators were instructed to select every other entry in the directories, for interview purposes. Entrepreneurs were asked to participate in the survey after a detailed explanation on the research’s purpose and an estimated time needed for the face-to-face interview was given. Non- profit organizations, companies with more than 50 employees were excluded. The questionnaire used in this study was developed by Hung M. Chu (Chu & Katsioloudes, 2001) and has been used in studies of entrepreneurs in Vietnam, China, Thailand, Romania, Poland, Bulgaria, Turkey, Ghana, Kenya and Nigeria. It was originally written in English but was translated into Vietnamese and checked for inter-translator consistency. The strength of motivation items, success factors and problem variables were measured using a five-point Likert scale. The scale ranges from “1” indicating that the motive is unimportant to “5” which signifies the most important level of the variable. A mean score was computed for each one of the variables included in the three mentioned categories. A higher mean score indicates that the variable was more important to the entrepreneurs. For the total sample, a nonparametric test (the Wilcoxon rank sum test) was used to determine if one factor is significantly more important than the other factors. The Wilcoxon rank sum test was used instead of a t-test because the score was not normally distributed as determined by the Anderson- Darling test. When ordinal data with non-normal distribution were analyzed, the Wilcoxon test is thought to be more powerful test of the difference between two population medians (Hollander and Wolfe, 1999).

RESULTS The general characteristics of the sample are reported in Table 1. Among the entrepreneurs surveyed, 58 percent are female and 42 percent are male. Although traditional Vietnamese culture clearly specifies the role of women as staying home to take care of children and household chores,


56

results of the survey show otherwise. The results show a dramatic increase of women’s participation in business ownership. This shift may be due to the new society where they live, work and have assimilated. With regard to marital status, 92 percent of the respondents are married. Researchers found that immigrant entrepreneurs use family members as unpaid labor to cut the cost of running business. Especially in labor intensive businesses, married entrepreneurs receive assistance from their spouses and children. With respect to education level, 44 percent reported that they completed high school, 17 percent attended high school, 19 percent had some college education and only 2% completed college. A majority of respondents reported that they established their business by themselves, or purchased the entity from others. Inheritance of these businesses seems not to exist. Of the total entrepreneurs surveyed, 91.4% were personal beauty service providers such as nail or hair salon owners, 4.7% involved in retailing and 3.2% were other types of service providers. The average age of Vietnamese American entrepreneurs is reported at 46, which is older than those in other countries. This is likely because most of Vietnamese Americans came to America after the collapse of South Vietnam in 1975. It took time to improve their English skills and learn new trades. The time they devoted to business is 62 hours per week. As compared to other studies, this research shows Vietnamese American entrepreneurs devoted more time to their business than other groups of entrepreneurs. Motivation Vietnamese American entrepreneurs were asked to rank 11 motivation factors for their business ownership on a five point Likert scale, with five (5) being “extremely important” and one (1) being “the least important”. The results can be seen in Table 2. Aspirations like “to become my own boss”, “to have job security” and “to be able to use my past experience and training” were the three most important variables. According to Fernandez and Kim (1998), limited employment opportunity stimulates self-employment interest. Noncollege immigrants are less likely to be employed in professional occupations in the U.S. and as a result, business ownership is seen as the most productive option. As shown in Table 1, 90% respondents indicated that they didn’t receive a college degree since a majority of Vietnamese Americans came to America primarily as refugees. Given the language barrier and their limited skills, they were offered lower-status and lower paying positions. For others who had a strong educational background in their home country, their credentials were often not recognized by

American employers. To achieve the American Dream many Vietnamese refugees started their business to control their destiny, gain personal autonomy and find job security. This is consistent with the theory of Labor Market Discrimination (Le, 2008), which states that many people engaged in entrepreneurship because of unfavorable employment conditions. Vietnamese American entrepreneurs were not only motivated by extrinsic rewards such as job security and more income, but were also motivated by intrinsic rewards. The respondents stated that “maintaining personal freedom” and “achieving personal satisfaction and growth” were also the driving forces behind business ownership. These are related to their Asian cultural traits of self-sufficiency and personal achievement. Success As shown in Table 3, hard work, charisma and friendliness to customers were rated as equally important factors contributing to the business success. Finding a good location, gaining appropriate training, and providing good customer service were found to be three critical variables for a prosperous enterprise. Support of family and friends was also identified as necessary ingredients needed for the success of a business. Vietnamese American entrepreneurs face numerous challenges. Limited English ability and lack of necessary skill sets forced them to work long hours to gain a foothold in America. As reported in Table 1, on average Vietnamese American small business owners work 62.12 hours per week. This is even longer than most of entrepreneurs working in the developing countries. To compete with American businesses, Vietnamese American business owners realized that it was important to find a niche market. The focus on providing personal beauty services and specializing in the nail salon industry is a good example. It seemed to be an easy path to success because little training is needed and lack of English skills did not interfere with providing quality services. This strategy turned out to be very successful. Vietnamese nail salons are currently dominating the industry as a whole. In California, 80 percent of nail technicians are Vietnamese Americans. Nationwide, Vietnamese hold 43% of the licenses (Tran, 2008). However, this profession is very demanding. It is very common to work 10 hours a day and 7 days a week. Besides hard work, Vietnamese American entrepreneurs indicated that friendliness to customers, good customer service and appropriate training are crucial to building a strong customer base.


57


58

Many small businesses, especially supermarkets or restaurants serving ethnic Vietnamese food, locate inside traditional urban Asian ethnic enclaves, such as Little Saigons and Chinatowns. Those locations provide good traffic to the business and protect them from racial hostility. Research shows that immigrant entrepreneurs place “finding a good location” as another important factor for profitable business operations. Vietnam is a Confucianist society, which values family support. When making a strategic decision, the Vietnamese American entrepreneurs usually seek advice from family and friends rather than financial or legal advisors. As shown in Table 4, 43.7% respondents relied on family members’ opinions and 25.8% consulted their friends. In addition, family members provided 78.8% and friends provided 45.8% of the capital for the business. The close relationship with family members and friends is a key to success. Problems Regarding the problems faced by Vietnamese American entrepreneurs, the respondents suggest that fierce competition and lack of management training were the two major problems encountered. One of the competitive advantages of Vietnamese American business is low cost of services. According to Hoang (2006), in a Vietnamese nail salon, customers only pay $10 for a standard manicure and $25 for a spa chair pedicure. It’s not surprising that this low cost leads to excessive competition in the industry. Another hurdle is that the first generation immigrants from Vietnam are not highly educated. In our sample, only 1.8% business owners completed college degree. Lack of systematic management knowledge and training restricts them from sustainable growth. The difficulty of recruiting and retaining good employees is a very common problem in a small business due to the limited growth space and compensation benefits. The situation could be worse in a low skill, labor intensive business. Once the employee gains substantial experience and builds a good customer base, they often leave to open up a new business and take their clientele with them.

Although Vietnamese American entrepreneurs have achieved great success, they still face discrimination because of their Asian ethnicity and lower education levels. In addition, with their locations in less affluent neighborhoods, they may become the target of crimes. “Discrimination from customers” and “unsafe location” were reported as obstacles to develop business.

CONCLUSION Although there is growing research interest in small business development of Asian immigrants, Asian entrepreneurs are often treated as a group. There is very limited literature that focuses on the study of Vietnamese American entrepreneurs. This study begins to shed light on some of the motivation, success factors and problems of Vietnamese American entrepreneurs. As with other refugees who came to America for a better life, the Vietnamese Americans also face a variety of hurdles including cultural differences and ongoing discrimination. They established small businesses to realize their potential and fulfill their needs. Through business ownership they were able to control their destiny and had the opportunity to use their past experience and training. Factors contributing to their success were hard work, friendliness to customers, good location, appropriate training, good customer service and support of family and friends. Although Vietnamese American entrepreneurs have had much success, they still face some problems, such as fierce competition, lack of management training, undependable employees, unsafe business locations and discrimination. The growth of Vietnamese American small businesses contributes to the strength of the U.S. economy. Through diversification and strong work ethic, Vietnamese American entrepreneurs continue to be a niche business group to watch. Since research related to this group is limited, it is important to implement a long term research agenda.

TABLE 1. General Characteristics of Vietnamese Small Business in America

Frequency Percent

Gender

Male

117 41.9

Female

162 58.1

Marital status Married 256 91.8 Single 23 8.2

Education level achieved No formal education

1 0.4 Some grade school 13 4.7 Completed grade school 12

4.3

Some high school 48 17.2 Completed high school 122 43.7 Some college 54

19.4

Completed college 5 1.8 Some graduate work 0 0 Not mentioned 24

8.6

Type of business ownership Established by you 129 46.3

Bought from another owner 137 49.1 Inherited 9 3.2

Independently owned 2 0.7 Franchise business 0 0

Owned in partnership 2 0.7

Type of business Retailing 13 4.7 Wholesaling 1 0.4 Personal Beauty Service 255 91.4 Manufacturing 0 0 Agriculture 0 0 Other Service 9 3.2 Not mentioned 1 0.3

Mean age of entrepreneur 46.28 years

Avg. working hrs per week 62.12


TABLE 2. Mean Score for Motivation (5= extremely important, 4= very important, 3= mildly important, 2= not very important, 1= unimportant)

Motivational factors Mean Std. Dev.

To be my own boss 4.25 0.737 So I will always have job security 4.13 0.813 To be able to use my past experience and training 3.97 0.767 To maintain my personal freedom 3.87 0.804 For my own satisfaction and growth 3.63 0.851 To increase my income 3.49 0.846 To provide jobs for family members 2.24 1.215 To be closer to my family 2.22 1.147 To prove I can do it 2.19 1.183 To build a business to pass on 2.09 1.224 To gain public recognition 1.54 0.846

TABLE 3. Mean Score for Factors Contributing to Business Success (5= extremely important, 4= very important, 3= mildly important, 2= not very important, 1= unimportant)

Success factors Mean Std. Dev.

Hard work 4.09 0.605 Charisma and friendliness to customers 4.09 0.572 Location 3.70 0.788 Appropriate training 3.64 0.819 Good customer service 3.54 0.930 Support of family and friends 3.47 0.722 Marketing factors such as sales promotion 2.98 0.913 Good product at a competitive price 2.91 1.172 Previous business experience 2.90 1.244 Access to capital 2.85 0.852 Community involvement 2.61 0.970 Good general management skills 2.60 1.224 Ability to manage personnel 2.20 1.320 Maintenance of accurate records of sales/expenses 2.14 1.313 Political involvement 1.62 0.935 Satisfactory govt. support 1.59 1.033



61

TABLE 4. Support from Family and Friends

Source of Advice Frequency Percent

Legal Advisor 13 4.6 Financial Advisor or Bank/Lending Institution 4 1.4 Friends 72 25.8 Family 122 43.7 Other Business Owners 50 17.9 Others 40 14.6

Source of Business/Financial Capital Frequency Percent

American bank 44 15.7 Vietnamese bank 1 0.3 SBA 0 0.0 Family 220 78.8 Friends 128 45.8 Personal Savings 44 5.7 Others 1 0.3

TABLE 5. Mean Score for Each Problem Faced by Entrepreneur (5= extremely important, 4= very important, 3= mildly important, 2= not very important, 1= unimportant)

Problem Mean Std. Dev.

Too much competition 3.27 0.646 Lack of management training 3.13 0.670 Unsafe location 2.63 1.073 Unreliable and undependable employees 2.36 1.271 Discrimination from customers 2.16 1.432 Inability to maintain accurate and informative accounting records 2.14 1.428 Weak economy 2.07 1.482 Not having financial capital 1.97 0.974 Any other problem 1.97 1.362 Lack of marketing training 1.95 1.341 Limited parking 1.88 0.969 Language problem 1.65 1.000 Too much govt. regulation or bureaucracy 0 0

REFERENCES Benzing, C., H.M. Chu and B. Szabo. 2005. Hungarian and Romanian Entrepreneurs in Romania- Motivation, Problems and Differences. Journal of Global Business. 16:77-88. Benzing, C., H.M. Chu and G. Callann. 2005. Regional Comparison of the Motivation and Problems of Vietnamese Entrepreneurs. Journal of Developmental Entrepreneurship. 10: 3-27. Chu, H.M. and C. Benzing. 2004. Vietnamese Entrepreneurs: Motivation, Problems, and Success Factors. Journal of Global Business. 15 (28): 25-33. Chu, H. M. and M. I. Katsioloudes. 2001. Cultural Context and the Vietnamese-American Entrepreneurial Experience. Journal of Transnational Management Development. 7 (2): 37-46. Chu, H.M., C. Benzing, and C. McGee. 2007. Ghanaian and Kenyan Entrepreneurs: A Comparative Analysis of Their Motivations, Success Characteristics, and Problems. Journal of Developmental Entrepreneurship.12: 295-322. Fernandez, M. and K.C. Kim. 1998. Self-Employment Rates of Asian Immigrant Groups: An Analysis of Intragroup and Intergroup Differences. International Migration Review. 32 (3): 654-681. Fadahunsi, A., D. Smallbone, and S. Supri. 2000. Networking in Ethnic Minority Enterprises: an Analysis of Small Minority Owned Firms in North London. Journal for Small Business and Enterprise Development. 3: 228-240. Gosh, B.C., T.S. Kim, and L.A. Meng. 1993. Factors Contributing to the Success of Local SMEs: An Insight from Singapore. Journal of Small Business and Entrepreneurship. 10 (3): 33-45.

Hoang, D. 2006. The Back of the Hand: Vietnamese American Nail Salons. Asian American Movement E-zine. December 15, 2006. Hollander, M., and D.A. Wolfe. 1999. Nonparametric Statistical Methods. New York: Wiley. Huck, J.F. and T. McEwen. 1991. Competencies Needed for Small Business Success: Perception of Jamaican Entrepreneurs. Journal of Small Business Management. 3: 90-93. Le, C.N. 2008. Asian Small Businesses. Asian-Nation: The Landscape of Asian America. <http://www.asian-nation.org/small-business.shtml>. Liao, D., and P. Sohmen. 2001. The Development of Modern Entrepreneurship in China. Stanford Journal of East Asian Affairs. 1: 27-33. Nguoi Viet Yearbook. 2005. Nguoi Viet Daily News. Overseas Vietnamese. Wikipedia, the free encyclopedia. Pisturi, D., W. Huang, D. Oksoy, Z. Jing, and H. Welsch. 2001. Entrepreneurship in China: Characteristics, Attributes, and Family Forces Shaping the Emerging Private Sector. Family Business Review. 14 (2): 141-153. Swierczek, F. W., and T.T. Ha. 2003. Motivation, Entrepreneurship and the Performance of SMEs in Vietnam. Journal of Enterprising Culture. 11 (1): 47-68. Tran, D. 2005. The Vietnamese American Community. Asian-Nation. <http://www.asian-nation.org/vietnamese-community.shtml>. Tran, M. 2008. Vietnamese Nail Down the U.S. Manicure Business. Los Angeles Times. May 5, 2008 Vietnamese Americans. 2009. Retrieved April 20, 2009 from <http://www.everyculture.com/multi/Sr-Z/Vietnamese -Americas.html>.


62

http://www.everyculture.com/multi/Sr-Z/Vietnamese%20-Americas.html

http://www.everyculture.com/multi/Sr-Z/Vietnamese%20-Americas.html

THE EFFECT OF GENDER ON LEARNING AND SUCCESS IN ECONOMICS CLASSES

Orhan Kara and I-Ming Chiu West Chester University and Rutgers University-Camden

West Chester, PA 19383 and Camden, NJ 08102

ABSTRACT To investigate the effect of gender on learning and success in economics classes, seven hundred and fourteen principles of microeconomics and macroeconomics students are surveyed in two public universities. Results indicate that gender was a significant factor contributing to learning and success as measured by grades in addition to a number of hours worked, SAT scores, number of missed classes, recommending the course to a friend, instructors, being a junior, number of economics courses taken, course, and interest in the course. While the effect of the number of hours per week spent on studying for the class was negative, GPA, age, staying in university housing, number of mathematics classes taken, instructor’s use of graphs to explain a topic, being a fourth year student, enrolling for a class because of the reputation of an instructor had positive effect on students grades. Female students performed better than male students in our study, indicating a changing gender pattern.

INTRODUCTION

There is an extensive research about the factors contributing to student performance and success in economics classes. Main factors researchers investigated include math skills (Cohn, Cohn, Hult, Balch, and Bradley, 1998; Cohn and Cohn, 2001; Hill and Stegner, 2003; and Ballard and Johnson, 2004), who is the instructor and teaching methods (Watts and Bosshardt, 1991; Vachris, 1999; Colander, 2005; Goffe and Sosin, 2005; Laband and Piette, 1995; and Robb and Robb, 1999), absenteeism (Durden and Ellis, 1995, Chan, Shum, and Wright, 1997; Marburger, 2001; and Cohn and Johnson, 2006), class size (McConnell and Sosin, 1984; and Aries and Walker, 2004), student effort (Borg, Mason, and Shapiro, 1989; Didia and Hasnat, 1998; Krohn and O’Connor, 2005; Lumsden and Scott, 1987; and Park and Kerr, 1990), employment (Paul,1982), seating location (Benedict and Hoag,2004), personality type (Borg and Shapiro, 1996), and gender. Gender is one of the factors influencing student success and learning in economics classes. Although researchers studied this subject heavily, they have not reached to an agreed upon conclusion. Some studies did find that male students generally performed better (Siegfried, 1979; Lumsden and Scott,1985; Lage & Treglia, 1996; Walstad & Robson, 1997; and Borg & Stranahan, 2002) while others did not find any effect of gender in economics classes (Williams, Waldauer,

& Duggal, 1992; Greene, 1997; Saunders& Sounders, 1999; and Ballard & Johnson, 2005). After a detailed analysis of the previous studies’ findings, Siegfried (1979) is not able draw any strong conclusions, but he concludes that gender differences start developing during high school and extend to college years. Following Siegfried’s (1979) study, researchers tried to understand and explain this difference. Lumsden and Scott (1985) argued that gender differentials might be explained by the type of exams students take in economics classes. Observing that male learning rate was higher than that of female, Lumsden and Scott (1985) found that female students did better on essay exanimation with an average of seven points than male students who performed slightly better, an average of four points in multiple choice exams. Lage &Treglia (1996) examined whether gender differences disappear if economics classes are modified by making classes more gender inclusive and found that there was an increase in female performance and that all students performed better. However, they observed significant gender differences. Focusing on the multiple choice test performance, Walstad & Robson (1997) cited four reasons as to why female students did not perform as well as male students: social & cultural influence, cognitive differences, instructional differences, and fixed or constructed-response format of an economics test. By employing differential item functioning method they reported a gender difference in economics classes. Similarly, Borg & Stranahan (2002) investigated the gender difference in economics classes and confirmed the fact that female students performed worse than male students did. However, they concentrated on the personality types based on the Kiersey-Bates temperament types in principles of macroeconomics to control the type of personality of the students. Using an ordered probit model, Borg & Stranahan (2002) concluded that gender affects the performance in economics class although the effect varies according to personality types. On the other hand, there are other studies showing no gender effect in economics classes. For example, Williams et al. (1992) , examining the gender effect in intermediate macroeconomics, intermediate microeconomics, and economics statistics classes, found neither significant nor consistent gender difference by using multivariate regression analysis. Although their conclusions indicated that males outperformed females on the essay section in the


63

macroeconomics and microeconomics, females outperformed males on the essay section of statistics. While females did better in the numerical section of micro exams, males scored better in the numerical section of macroeconomics, indicating no pattern of gender difference. Likewise, Greene (1997) studied the verbal ability of women in order to test the claim that females posses higher verbal abilities, which leads to more success in verbal evaluation in economics classes compared to the written evaluations. Using his introductory macroeconomics class over a four-year period, Greene (1997) concluded that females were not better than males based on the reading comprehension diagnostics. Saunders & Saunders (1999) looked at the influence of the gender of instructor on the gender differences in learning economics. They collected the data from introductory economics classes over a six-year period from 1984 to 1990 and used multivariate analysis to analyze the effect of gender difference. However, they found no support to the idea that the gender of the instructor would explain the difference in male and female success in economics classes. In addition, Ballard & Johnson (2005) investigated the role of expectations of grade and gender to test the hypothesis that women are less likely to expect to earn better grade in economics, which may lead to a lower grade as a result of self-fulfilling property. 1462 students who took introductory microeconomics classes from the same instructor were included in the study. An ordered probit technique was used to analyze the data. Although their findings suggested the positive effect of the expectations and the success on the class in terms of final grade, the gender effect is found to be very small and statistically insignificant. In summary, the gender gap in economics classes is an unsettled issue and the evidence is mixed. Especially early studies illustrated that male students performed better in economics classes than female students whereas relatively new line of research indicated no significant difference in male and female learning and success in economics classes. Therefore we tried to investigate this issue by controlling more variables and using an ordered logit model as the next section explains.

DATA AND METHODOLOGY The data used in this study come from two sets of surveys administered at the University of North Dakota and West Chester University. At the University of North Dakota, the survey was given to students during the final exam in 2003 in order to increase the response rate. All the instructors of principles of economics classes (both micro and macro) agreed to give the survey to their students, and all the students who took the final exam filled out the questionnaires. The same survey was administered at West Chester University during the final exams in spring 2004. All

instructors but one cooperated during the process. After collecting surveys, we also obtained the grades of the students who completed the survey. There are thirty four questions in the survey and 744 responses are recorded. In order to determine the effect of gender on learning economics, an ordered logit model is applied (Greene, 2002). Since the grades, our dependent variable, are ranked in order from the best to the worst (from A to F), we use the ordered logit model. We depart from Park and Kerr (1990) who use a multinomial logit approach. The following model is estimated by using Limdep software: Grades = β0 + βi (class and student attributes) + error (1) i= 1,..,3, 4 Grades are the final grades that students obtain, A, B, C, D, and F. Class and student attributes include GPA, gender, age, course, university housing, number of hours per week worked at a job, number of mathematics courses taken, number of economics courses taken, SAT score, expected grade at the beginning of the semester, expected grade at the end of the semester, number of hours per week spent studying for the class, number of missed classes, textbook rating by student, understanding when the instructor uses graphs to explain a topic, understanding when the instructor uses equations to explain a topic, interest in the course, whether to recommend the course to a friend, university, instructor (eight dummy variables for nine instructors), year of study (three dummy variable for sophomore, junior, and senior), and dummy variables for reasons for registering in the specific class.

RESULTS

Sample characteristics are given in Table 1, summarizing the information about the 714 principles of microeconomics and macroeconomics students who were surveyed. 479 (64.4%.) students identify themselves as male and 265 (35.6%) as female. The average grade is 2.64 out of 4, a little below a B, and the average GPA is also around B (2.98). About 22% of the students received a grade of A, while 35% received B’s, 31% C’s, and the remaining 12% received D’s and F’s. Students are mainly first year students (39%) and sophomores (about 40%). Juniors and seniors constitute 21% of the sample. More than half of the students maintain a job during the semester, working an average of twenty hours each week. The average age of students is twenty; 90% of the students are twenty-one years old or younger. In addition, 60% of the surveyed students are in the principles of microeconomics class and the remaining 40% are in macroeconomics class. 57% of the students surveyed live in university housing during the time the survey was administered.


64

Half of the students indicate that they had at least one math and one economics classes prior to taking the current class. Based on the student self reported SAT scores, the average was 1413. WCU students had lower scores compared to the UND students. As Table 1 shows, the average expected grade changed very little from the beginning to the end of the semester. The adjustment in the expected grade took place in lower and higher end of the scale. While 25% of the students expected a grade of D in the beginning of the semester, 18% of students reported an expected grade of D at the end of the semester. Similarly, 13% of the students indicated an expected grade of A at the start of the semester, whereas 9% expected an A at the end of the course. On average, students spent less than three hours per week studying and missed an average of five classes throughout the semester. 62% of the students reported four or fewer missed classes. On a scale of one to ten, one being very poor and ten being excellent, students rated the textbooks at about a six on a ten-point scale. Students indicate that they better understand when instructors use graphs and equations to explain a topic, with the ranks averaging 7.1 and 7.17 respectively. While the students’ interest was about average in the class (the rank equals 5.71), the majority of the students (about 60%) find the class useful and 73% state that they would recommend the class to a friend. With respect to registering for a specific class, 60% of the students state convenience, 24% percent report a conflict with other classes, 3% note a conflict with work, and 3% note a conflict with personal affairs. 10% of the students cite the reputation of the instructor as a deciding factor when choosing a class. 77% of the students prefer a fifty-minute class to seventy-five minute class, and 58% respond that morning classes are preferred to afternoon classes. Results for correlations between grades and the independent variables are presented in Table 2. Positive, significant correlations exist between grades and the following variables: GPA, number of hours worked, number of economic courses taken previously, SAT scores, expected grade at the beginning of the semester, number of hours spent studying for the class, number of attended classes, instructors' use of graphs and equation, and interest in the course. We estimate a negative correlation between grades and the number of hours per week spent studying. This result is puzzling, but it is consistent with some previous studies (Didia and Hasnat, 1998; Krohn and O’Connor, 2005). The ordered logit estimates for equation (1) are given in Table 3 and Table 4. According to the Table 3, the model is satisfactory as χ2 and log likelihood diagnostics are acceptable. The ordered logit method does not produce the familiar F-test, and a likelihood ratio test is conducted to examine the overall explanatory power of the model. The

value of χ2 test statistic with thirty four degrees of freedom is 104.77, implying that the model fits well and the independent variables are jointly significant. The μ’s , the cut off points for different grades, are all statistically significant. Dividing the estimated coefficient by the standard error does not give the usual t-statistics, and we used the term Z in Table 3 to avoid confusion. Seven variables are statistically significant at the 5% level, including gender, course (macro vs. micro), the number of hours worked, SAT score, the number of missed classes, whether or not to recommend the course, and instructor six. In addition, the estimated coefficients for number of economics courses taken an interest in the course are statistically significant at ten percent level. The interpretation of the estimated coefficients of ordered logit model is not as straightforward as ordinary least square estimates. A coefficient estimate indicates a change in the log of the odds ratio. We first transform the coefficient by using the exponential function to find the antilog (eβ), and then we use the value computed from the transformation to predict the odds ratio. Since we estimate the coefficient for gender as .3305, this yields 1.39 as the odds ratio. This result implies that female students are 1.39 times more likely to get a better grade compared to male students. Likewise, the odds ratio for being a junior, a macroeconomics student, a senior, and registering because of the reputation of the instructor are 1.6, 1.5, 1.4, and 1.04, respectively. Table 4 reports the marginal effects of variables on grades. Gender, being in macroeconomics, instructors, being a junior and senior, reason for registering in a certain class due to conflicts with other courses, work, and personal affairs, and reputation of instructor are important variables in learning and success in economics classes. The remaining variables have smaller effects on grades.

CONCLUSIONS

After investigating the effect of gender on learning in principles of economics classes by controlling several variables, we reach the following conclusion. The principles of microeconomics and macroeconomics classes studied are dominated by male students. Unlike previous studies that found that male students did better in economics classes, our results point to the opposite conclusion. There is a changing pattern of the gender effect in the economics classes where female students seem to be doing better than male students.


65


66

Table 1: Descriptive Statistics

Variables Mean Std. DeviationGrade 2.64 1.011GPA 2.98 .567Age 19.83 2.515Year in School : 1=Freshman 2=Sophomore 3=Junior 4=Senior 1.86 .850# of hours per week worked 19.80 9.237University housing: 1= Yes, 0=No .57 .495Number of mathematics courses taken 1.80 1.332Number of Economics courses taken 1.65 1.978SAT Score 1412.89 339.552Expected Grade at the beginning of the semester 2.32 .994Expected Grade at the end of the semester 2.31 .873# of hours per week spent on studying for the class 2.79 2.143Number of missed classes 4.57 4.247Textbook rating 6.01 2.200Understanding when the instructor uses graphs to explain a topic 7.10 2.275Understanding when the instructor uses equations to explain a topic 7.17 2.202Interest in the course 5.71 2.345Usefulness of the course 6.26 2.273Whether to recommend the course .73 .447Preference: 1=50 minute class 0= 75 minute class .77 .419Preference: 1=Morning class 0= Afternoon class .58 .494

Table 2: Correlations with Grade Correlation Significance GPA 0.60** (0.000) Gender 1=Female, 0=Male 0.04 (0.253) Age 0.03 (0.381) University housing: 1= Yes, 0=No 0.03 (0.498) # of hours per week worked at a job -0.11* (0.021) Number of mathematics courses taken 0.04 (0.232) Number of Economics courses taken 0.10** (0.007) SAT Score 0.23** (0.000) Expected Grade at the beginning of the semester -0.16** (0.000) Expected Grade at the end of the semester 0.03 (0.496) # of hours per week spent on studying for the class -0.09* (0.023) Number of missed classes -0.22** (0.000) Textbook rating 0.03 (0.477) Understanding when the instructor uses graphs to explain a topic 0.29** (0.000) Understanding when the instructor uses equations to explain a topic 0.24** (0.000) Interest in the course 0.18** (0.000)

*. Correlation is significant at the 0.05 level (2-tailed) **. Correlation is significant at the 0.01 level (2-tailed)

Table 3. Ordered Logit Model Estimates Variable Coefficient Std. Error P[|Z|>z] Constant 3.6443 0.2964 0.0000 GPA 0.0004 0.0004 0.2281 Gender 1=Female, 0=Male 0.3305 0.1473 0.0249 Age 0.0002 0.0004 0.6914 Course 1=Macro, 0= Micro 0.3797 0.1898 0.0454 University housing: 1= Yes, 0=No 0.0005 0.0007 0.5012 # of hours per week worked at a job -0.0003 0.0001 0.0310 Number of mathematics courses taken 0.0005 0.0004 0.2022 Number of Economics courses taken 0.0005 0.0003 0.0770 SAT Score 0.0003 0.0001 0.0005 Expected Grade at the beginning of the semester -0.0005 0.0009 0.5868 Expected Grade at the end of the semester 0.0008 0.0006 0.1695 # of hours per week spent on studying for the class -0.0004 0.0004 0.3592 Number of missed classes 0.0019 0.0005 0.0003 Textbook rating -0.0003 0.0011 0.7553 Understanding when the instructor uses graphs to explain a topic 0.0005 0.0014 0.7431 Understanding when the instructor uses equations to explain a topic -0.0007 0.0011 0.5445 Interest in the course -0.0026 0.0014 0.0670 Whether to recommend the course 0.0013 0.0005 0.0066 University 0.0924 0.3922 0.8137 If Instructor 1= 1, 0=Otherwise -0.5125 0.3390 0.1306 If Instructor 2= 1, 0=Otherwise -0.4077 0.6382 0.5229 If Instructor 3= 1, 0=Otherwise -0.6704 0.4881 0.1696 If Instructor 4= 1, 0=Otherwise 0.0800 0.4697 0.8647 If Instructor 5= 1, 0=Otherwise -0.7418 0.5340 0.1648 If Instructor 6= 1, 0=Otherwise -0.8432 0.3224 0.0089 If Instructor 7= 1, 0=Otherwise -1.0306 0.6422 0.1085 If Instructor 8= 1, 0=Otherwise 0.1801 0.5503 0.7434 Sophomore =1, 0=Otherwise -0.0278 0.1673 0.8679 Junior =1, 0=Otherwise 0.4720 0.2212 0.0328 Senior=1, 0=Otherwise 0.3663 0.3570 0.3049 Reason for registration: Conflict w/ course -0.1328 0.1789 0.4577 Reason for registration: Conflict w/ work -0.6161 0.4120 0.1348 Reason for registration: Conflict w/ personal affairs -0.5819 0.4413 0.1873 Reason for registration: reputation of instructor 0.0357 0.2289 0.8759 Mu(1) 1.8699 0.1075 0.0000 Mu(2) 3.6981 0.0809 0.0000 Mu(3) 5.3679 0.0976 0.0000 Dependent Variable: Grades Log likelihood function= -978.1016 Restricted log likelihood= -1030.485 χ2 = 104.7664 Degrees of freedom = 34


67


68

Table 4: Marginal Effects for Ordered Logit Model Variables F D C B A Constant 0.0000 0.0000 0.0000 0.0000 0.0000 GPA 0.0000 0.0000 -0.0001 0.0000 0.0001 Gender 1=Female, 0=Male -0.0051 -0.0230 -0.0504 0.0207 0.0579 Age 0.0000 0.0000 0.0000 0.0000 0.0000 Course 1=Macro, 0= Micro -0.0059 -0.0263 -0.0579 0.0233 0.0668 University housing: 1= Yes, 0=No 0.0000 0.0000 -0.0001 0.0000 0.0001 # of hours per week worked at a job 0.0000 0.0000 0.0000 0.0000 -0.0001 Number of mathematics courses taken 0.0000 0.0000 -0.0001 0.0000 0.0001 Number of Economics courses taken 0.0000 0.0000 -0.0001 0.0000 0.0001 SAT Score 0.0000 0.0000 0.0000 0.0000 0.0001 Expected Grade at the beginning of the semester 0.0000 0.0000 0.0001 0.0000 -0.0001 Expected Grade at the end of the semester 0.0000 -0.0001 -0.0001 0.0001 0.0001 # of hours per week spent on studying for the class 0.0000 0.0000 0.0001 0.0000 -0.0001 Number of missed classes 0.0000 -0.0001 -0.0003 0.0001 0.0003 Textbook rating 0.0000 0.0000 0.0001 0.0000 -0.0001 Understanding when the instructor uses graphs to explain a topic 0.0000 0.0000 -0.0001 0.0000 0.0001 Understanding when the instructor uses equations to explain a topic 0.0000 0.0000 0.0001 0.0000 -0.0001 Interest in the course 0.0000 0.0002 0.0004 -0.0002 -0.0005 Whether to recommend the course 0.0000 -0.0001 -0.0002 0.0001 0.0002 University -0.0015 -0.0067 -0.0141 0.0065 0.0157 If Instructor 1= 1, 0=Otherwise 0.0101 0.0428 0.0733 -0.0483 -0.0779 If Instructor 2= 1, 0=Otherwise 0.0079 0.0336 0.0590 -0.0377 -0.0627 If Instructor 3= 1, 0=Otherwise 0.0135 0.0569 0.0943 -0.0639 -0.1009 If Instructor 4= 1, 0=Otherwise -0.0013 -0.0056 -0.0122 0.0052 0.0139 If Instructor 5= 1, 0=Otherwise 0.0161 0.0661 0.1006 -0.0760 -0.1068 If Instructor 6= 1, 0=Otherwise 0.0190 0.0772 0.1113 -0.0888 -0.1187 If Instructor 7= 1, 0=Otherwise 0.0268 0.1033 0.1218 -0.1191 -0.1327 If Instructor 8= 1, 0=Otherwise -0.0027 -0.0121 -0.0277 0.0102 0.0323 Sophomore =1, 0=Otherwise 0.0004 0.0020 0.0042 -0.0019 -0.0048 Junior =1, 0=Otherwise -0.0066 -0.0300 -0.0723 0.0211 0.0878 Senior=1, 0=Otherwise -0.0050 -0.0231 -0.0563 0.0159 0.0685 Reason for registration: Conflict w/ course 0.0022 0.0098 0.0201 -0.0098 -0.0223 Reason for registration: Conflict w/ work 0.0131 0.0544 0.0848 -0.0629 -0.0893 Reason for registration: Conflict w/ personal affairs 0.0123 0.0512 0.0804 -0.0593 -0.0846 Reason for registration: reputation of instructor -0.0006 -0.0025 -0.0055 0.0024 0.0062

REFERENCES Aries, J. J. & Waller, D. M. 2004. Additional evidence on the relationship between class size and student performance. Journal of Economic Education, pp. 341-329. Ballard, C.L.& Johnson, M.F. 2004. Basic math skills and performance in an introductory economics class. Journal of Economic Education, pp. 3-23. Ballard, C. & Johnson, M. 2005. Gender, expectations, and grades in introductory microseconds at a US university. Feminist Economics, 111, pp. 95-122. Becker, W. E.& Watts, M. 1995. Teaching tools: Teaching methods in undergraduate economics. American Economic Review, 8612, pp. 448-453. Benedict, M. E. & Hoag, J. 2004. Seating location in large lectures: Are seating preferences or location related lo course performance? Journal of Economic Education, pp. 215-231. Borg, M. O., Mason, P. M., & Shapiro, S. L. 1989. The case of effort variables in student performance. Journal of Economic Education, 20, pp. 308-313. Borg, M. & Shapiro, S.L. 1996, Personality type and student performance in principles of economics. Journal of Economic Education, pp. 3-25. Borg, M. O. & Stranahan, H. 2002. The effect of gender and race on student performance in principles of economics: The importance of personality type. Applied Economics, 34, pp. 589-598. Chan, K. C; Shum, C. & Wright, D. J. 1997. Class attendance and student performance in principles of finance. Financial Practice and Education, 7 , pp. 58-65. Cohn, E., & Cohn, S. 1994. Graphs and leaning in principles of economics. AEA Paper and Proceedings, pp. 197- 200. Cohn, E, Cohn, S., Hult, R. E., Balch, D.C., & Bradley, 1998. The effects of mathematics background on student learning in principles of economics. Journal of Education for Business, pp. 18-22. Cohn, E., Cohn, S., Balch, D. C.,& Bradley, J. 2001. Do graphs promote leaning in principles of economics? Journal of Economic Education, pp. 299-310. Cohn, E. Cohn, s, Balch, D.C.& Bradley, J. 2004. The relation between student attitudes towards graphs and performance in economics. The American Economist, 482, pp. 41-52.

Cohn, E. & Johnson, E. 2006. Class attendance and performance in principles of economics. Education Economics, 142, pp. 211-233. Colander, D. 2005. What economists teach and what economists do. Journal of Economic Education, pp. 249-260. Didia D. & Hasnat, B. 1998. The determinants of ptomaine in the university introductory finance course. Financial Practice and Education, 8, pp. 102-107. Durden G. C.& Ellis, L. V. 1995. The effects of attendance on student learning in principles of economics. AEA Papers and Proceedings, pp. 343-346. Greene, B. 1997. Verbal abilities, gender, and the introductory economics course: A new look at an old assumption. Journal of Economic Education, 7, pp. 13-30. Greene, W. H. 2002. Econometric Analysis. 5th Ed. Prentice Hall. Goffe, W. L. & Sosin, K. 2005. Teaching with technology: May you live in interesting times. Journal of Economic Education, pp. 278-291. Hill, C.D. & Stegner. 2003. Which students benefit from graphs in a principles of economics class? The American Economist, 472, pp 69-77. Krohn, G. A. & O’Connor, C. M. 2005. Student effort and performance over the semester. Journal of Economic Education, pp. 3-28. Laband D. N. & Piette, M. J. 1995. Better Learning from better management: How to improve the principles of economics course. Does who teaches principles of economics matter? AEA Papers and Proceedings, pp. 335-338. Lage, M. J. & Treglia, M. 1996. The impact of integrating scholarship of women into introductory economics: Evidence from one institution. Journal of Economic Education, 6, pp. 26-36. Lumsden, K. G. & Scott, A. 1987. The economics student reexamined: Male-female differences in comprehension, Journal of Economic Education, 18, pp. 365-375. Marburger, DR. 2001. Absenteeism and undergraduate exam performance. Journal of Economic Education, 322,pp. 99-109. McConnell, C.R. & Sosin, K. 1984. Some determinants of student attitudes toward large classes. Journal of Economic Education, pp. 181- 190.


69


70

Park, K. H. & Kerr, P. M. 1990. Determinants of academic performance: A multinomial logit approach. Journal of Economic Education, 21, pp. 101-111. Paul, H. 1982. The impact of outside employment on student achievement in macroeconomic principles. Journal of Economic Education , pp. 51-56. Robb, R.E. & Robb, A. L. 1999. Gender and the study of economics: The role of gender of the instructor. Journal of Economic Education, pp. 3-19. Siegfried, J. J. 1979. Male-female differences in economic education: a survey. Journal of Economic Education, 10, pp. 1-11. Vachris, M.A. 1999. Teaching principles of economies without “chalk and talk”: The experience of CNU online. Journal of Economic Education, pp. 292-303.

Saunders, K. T. & Saunders, P. 1999. The influence of instructor gender on learning and instructor ratings. Atlantic Economic Journal, 27, 4, pp. 460-473. Walstad, W. B. & Robson, D. 1997. Differential item functioning and male-female differences on multiple-choice tests in economics , Journal of Economic Education, pp. 155-171. Watts, M. & Bosshardt, W. 1991. How instructors make a difference: Panel data estimates from principles of economics courses. The Review of Economics and Statistics, pp. 336-340. Williams, M. L., Waldauer, C. & Duggal, V. G. 1992. Gender differences in economic knowledge: an extension of the analysis. Journal of Economic Education, pp. 219-231.

SOME MAJOR CAMPAIGN ISSUES OF 2008: AN ECONOMIC PERSPECTIVE A STUDENT-FACULTY RESEARCH PROJECT

William V. Sanders, Tyler Cotherman, David Dunkin, Nicholas Larson,

Christopher Lundgren, Jared Schmaeder and Scott Stegman Economics Department

Clarion University Clarion, PA 16214

ABSTRACT During the 2008 primary campaign and the ensuing Presidential and congressional campaigns, a number of issues were debated by the candidates. A number of these were chosen for scrutiny here: the abnormal profits of integrated oil companies; the loss of U.S. manufacturing to other countries; the “exporting” of jobs overseas though foreign trade; and the “health crisis” facing the country from increasing health prices/costs. The subjects were found to lack substance.

INTRODUCTION

This paper does not attempt to determine which candidate has the best solution to these problems, but to define the problems themselves. Students working on this project are sophomores taking their second course in Economic and Business Statistics at Clarion University, and chose the subjects as the topic of their term project. The scope of the assigned project is a multiple regression with a minimum of three independent variables and thirty observations. After a first run, students meet with the instructor to identify a revision that would improve the estimation. Further revisions are not required. The six students taking part in this exercise worked through more revisions than usual, and on multiple equations. The full texts of their term projects, a syllabus for the course and a text version of the accompanying presentation of this paper may be found at //jupiter.clarion.edu/~sanders/PEA09 for a limited time. The equations below report the students’ initial results.

BIG OIL PROFITS Candidates railed against the huge profits of “big oil” (i.e., integrated) companies, probably tapping into the American distrust of big companies and government. Television news reports and newspaper articles emphasized the record profits of oil companies during the runup in gasoline prices. Baker (2008) in the San Francisco Chronicle, Ellis (2008) on CNNMoney, and Mouwad and Krauss (2009) in the New York Times (11) are just a sample. Profits were described variously as record, incredible, unjustifiable, huge, and soaring. Government committees were formed or convened, and talk of a windfall profits tax was heard.

Student input: Gasoline prices were regressed with crude oil prices, profit rates, GDP, and stocks of gasoline: ^ p=-31.5+3.17oilp–.006oilp2+.01gdp+6458.25q-1

(1)

(24.3) (-4.9) (12.5) (2.2) R2 = 0.9840 Ra2 = 0.9835 F = 1956 where p ≡ quarterly ave. price of unleaded regular gasoline in US dollars

oilp ≡ price in US dollars/bbl of oil gdp ≡ real GDP ($million, 2000=100) q ≡ gasoline stock

The source paper for these results may be found in the web file Larson. These results indicate that gasoline prices can be explained by market forces of supply and demand without the need to find a sinister plot. In fact, a quick look on the Internet tells the same story. Energy Tomorrow (2009), an industry group for example, shows integrated oil earnings vs. investment, returns on investment, environmental expenditures, and returns relative to other industries. In fact, the rates of return are rather average, relative to other sectors. The ownership of integrated is also given by category. Most is owned by pension funds, retirement accounts and mutual funds. In other words, it is owned by many small investors, rather than some sinister elite group. There is also a demonstration that “big oil” does a lot of investment and a lot of environmental spending. While large companies have large profits, the profit rates are unremarkable, as noted by the Everyday Economist (2006). As in times past, no major problem can be identified, let alone solved.

U.S. MANUFACTURING BASE Another popular debate topic was how to best stop the erosion of the U.S. manufacturing base, i.e. how to stop losing our manufacturing base to other countries. The allure of what the Economist (2006) calls “economic nationalism” is that some faceless bureaucrat is siding with evil or otherwise undeserving foreign workers, in order to steal domestic workers’ livelihood.


71

Quite typical of this claim was the Alliance for American Manufacturing (2007) claim that 3.2 million jobs had been lost between 2000 and May of 2007! While jobs are quoted, actual manufacturing output data seldom is. Student input: The relation of U.S. manufacturing value to world manufacturing, and a time trend of a proportion of U.S. to world manufacturing. ^ USMFG=39312805+.193626WMFG (2) (5.487) R2 = .6984 Ra2 = 0.6752 F = 30.1 where USMFG ≡ U.S. Mfg. value added

WMFG ≡ world manufacturing.

As shown in equation (2), a linear relationship exists between U. S. and world manufacturing, which indicates that U.S. manufacturing is growing in proportion to that of the world in value added terms. The ratio of the two is then estimated as a time trend. ^ RATIO=39312805+.193626T (3) (5.487) R2 = .6984 Ra2 = 0.6752 F = 30.1 where RATIO ≡ U.S. Manufacturing to world mfg. ratio

T ≡ year In fact, the ratio of U.S. to world manufacturing is not falling, according to equation (3). Statistical problems might even lead the reader in the opposite direction. The source paper may be found in the web file Lundgren. Again, the student research hits the mark. According to studies cited by the United Nations (2007), for example, the U.S. share of world manufacturing has remained constant for decades. Production has risen fairly steadily, and if anything, with less fluctuation than most other economies. The facts that manufacturing jobs are giving way to service and white-collar jobs, and that labor is being replaced by capital in manufacturing do not indicate a decline in manufacturing.

THE U.S. IS EXPORTING JOBS

Trade in products and outsourcing (trade in services) is often blamed for taking jobs away from American workers in the popular press. China, NAFTA and the World Trade Organization are often vilified by politicians standing in the ruins of Detroit. CNN.COM (2009) even compiled a list of “Exporting America” companies. The AFL-CIO (2009) has also chimed in on the subject, with WAL-MART heading their list of alleged job exporters.

Student input: Employment, unemployment and labor force participation were regressed with exports and imports, income and interest rates. Numbers in parentheses are t-values. ^ N=39232+.002Xt+148.2GDP+1726I (4) (1.25) (1.73) (31.3) R2=.9962 Ra2=.9960 F=3870 ^ N=38625+.0006Mt+125.3GDP+1764.6I (5) (0.62) (1430) (37.67) R2=.9961 Ra2=.9959 F=3770 ^ N=38154+.0004Bt+92.7GDP+1796.4I (6) (0.20) (1.09) (60.59) R2=.9961 Ra2=.9958 F=3741 where N ≡ Civilian Employment (1000) Xt ≡ Exports ($ million) Mt ≡ Imports ($ million) Bt≡ Net Exports ($ million) GDP ≡ GDP ($ million 2000=100) I ≡ Interest Rates (% quarterly) According to equations (4), (5) and (6), employment tends to rise with trade activity, as given by exports or imports. The balance of trade is indeterminate, however. ^ U=-1007.9-.007Xt+2.07GDP+117.2I (7) (-3.198) (4.350) (1.536) R2= .5282 Ra2= .4960 F= 16.419 ^ U=-761.1-.005Mt+1.93GDP+99.71I (8) (-3.505) (4.860) (1.310) R2= .5454 Ra2= .5144 F= 17.597 ^ U=1041.2+.005Bt+1.04GDP+152.6I (9) (2.371) (4.696) (1.966) R2= .4844 Ra2= .4492 F= 13.779 where: U≡ Unemployment

Xt≡ Total Exports Mt≡ Total Imports Bt≡ Total Net Exports GDP≡ Real GDP I≡Real Prime Interest Rate

Equations (7), (8) and (9) show lower unemployment with increased trade. Trade balance had no discernable effect, maybe due to deterministic relation of GDP to net exports. ^ UR=2.77-4.3E-6Xt+.049GDP+.12I (10) (-3.22) (.696) (2.92) R2=.288 Ra2=.237 F=5.65


72

^ UR=3.37-2.8E-6Mt+.063GDP+.09I (11) (-2.82) (.888) ( 2.45) R2=.254 Ra2=.20 F=4.76 ^ UR=4.28+3.2E-6Bt+.105GDP+.03I (12) (1.88) (1.48) (1.26) R2=.181 Ra2=.123 F=3.10 where UR ≡ Unemployment Rate (%) PRO=19.04-.009469YR (15) Xt ≡ Exports ($ million) Mt ≡ Imports ($ million) Bt≡ Net Exports ($ million) GDP ≡ GDP ($ million 2000=100) I ≡ Interest Rates (% quarterly) Equations (10), (11) and (12) show the unemployment rate dropping with greater trade (shown by exports or imports), and an indeterminate effect of trade balance. The source papers for these results may be found in the web files Cotherman, Schmaeder and Dunkin. The three students working on this topic came to roughly the same conclusion. Serious studies, such as those cited by the Slaughter and Swagel (2009) and the Tax Foundation (2009), have also found what we should expect from economic theory. That is, there is no evidence that increased trade causes unemployment or wage depression in the U.S. As noted in the Economist (2005), the loss in manufacturing jobs comes from capital substitution, increasing productivity and making the economy wealthier. Once again, solutions were debated to a nonproblem.

THE COST OF HEALTH CARE IS RISING In discussions of this topic, the concepts of cost and price are often indistinct. There is an inference that the costs of constant health care are rising, i.e., that the price is going up. Headlines ask why health care costs rise; stories pose questions by politicians and citizens; and reasonable explanations may be given but a short paragraph. ^ MEDVA=-15394+0.047932 GDPVA (13) (66.926) R2 = .9914 Ra2 = .9911 F= 336.3 where MEDVA ≡ Medical value added millions of 2000 US dollars GDPVA ≡ Total value added. Student input: Equation (13) regresses value added of medicine and value added of GDP. The results, despite autocorrelation, show value added in medicine is increasing with GDP. That is, we are getting more health care as income increases. Equations (14) and (15) deal with medical prices.

^ MEDP=14.02+0.046941 CPI (14) (66.926) R2 = .9438 Ra2 = .9428 F= 336 where MEDP ≡ Consumer Price Index for Medical Services

CPI ≡ Consumer Price Index for the whole economy

^

(-28.24) R2 = .93 Ra2 = .9289 F = 797 where PRO ≡ MEDP/CPI YR ≡ YEAR These results show that medical prices change roughly proportionally to general prices, but fall slightly as a proportion over time. Together, these suggest that we pay more for health care because we are demanding more health care. The prices don’t seem to be increasing. Again, despite the autocorrelation, results were quite strong. The source paper for these results may be found in the web file Stegman. These also parallel serious market studies. Periodically, inquiries about health care costs are begun with great fanfare, and ended with a quiet press note. The results of such inquiries are consistent. We pay more for health care, because we keep asking for, and getting more health care. New tests, new research expenditures, new procedures and new drugs explain the increasing costs, rather than higher price. Savings from technology, etc., are offset by tighter quality control during drug development, etc. As the Singletary (2009) and Brown (2009) articles summed up the problem, we must pay more to get more.

CONCLUSION Students found that statistical methods can be used to explore the importance of issues of the day. Later refinements in their models cemented their conclusions. For the specific topics chosen, they found that politicians and news reporters sometime debate nonissues, and that economic topics are used to disguise political topics or drama.

REFERENCES

1. AFL-CIO ”Wal-Mart’s Imports Lead to U. S. Jobs Exports,” from http://www.aflcio.org/corporatewatch/walmart/walmart_5.cfm 2. Alliance for American Manufacturing “With 3.2 Million U.S. Manufacturing Jobs Lost, Issues Facing Industry Should Be Discussed in Tonight’s Debate,” from http://www.americanmanufacturing.org/newscenter/pressreleases/2007/05/15/with-32-million-us-manufacturing-jobs-lost-issues-facing-industry-should-be-discussed-in-


73

http://www.aflcio.org/corporatewatch/walmart/walmart_5.cfm

http://www.americanmanufacturing.org/newscenter/pressreleases/2007/05/15/with-32-million-us-manufacturing-jobs-lost-issues-facing-industry-should-be-discussed-in-tonight%E2%80%99s-debate/




74

tonight%E2%80%99s-debate/ 3. Baker, D. “Chevron’s Profit Soars to Record,” San Francisco Chronicle, from www.sfgate.com/cgi- bin /article.sgi?f=c/a/2008/BU6AUQMT9.DTL 4. Brown, D. “We All Want Longer, Healthier Lives. But It's Going to Cost Us,” The Washington Post, Jan. 11, 2009, from http://www.washingtonpost.com/wp-dyn/content/article/2009/01/09/AR2009010902296.html

5. CNN.COM “Exporting America,” from http://www.cnn.com/CNN/Programs/lou.dobbs.tonight/popups/exporting.america/frameset.exclude.html 6. Term used in The Economist, Feb. 7th-13th, 2009 London, pp. 9 ff. 7. The Economist, London, U.K., September 29, 2005. 8. Ellis, David “Exxon Shatters Profits Record,” CNN on Money, from http://money.cnn.com/2008/02/01/news/companies/exxon_earnings/ 9. Energy Tomorrow “America’s Oil and Gas Industry; Putting Profits Into Perspective,” from http://www.energytomorrow.org/ViewResource.ashx?id=5335 10. The Everyday Economist, from http://everydayecon.wordpress.com/2006/04/26/oil-profit-margins-vs-other-industries/ 11. Mouwad, J. and Krauss, C. “Exxon Posts Record 2008 Profit Despite Slip in 4th Quarter,” New York Times, Jan 30, 2009 from http://nytimes.com/2008/02/01/business/olcnd-exxon.html 12. Singletary, Michelle “The High Cost of Being Healthy,” The Washington Post, Thursday, January 15, 2009, at http://www.washingtonpost.com/wp-dyn/content/article/2009/01/15/AR2009011501942.html 13. Slaughter, M. and Swagel, P. “Does Globalization Lower Wages and Export Jobs?,” International Monetary Fund, September 1997, from http://www.imf.org/external/pubs/ft/issues11/ 14. Tax Foundation “Do Multinational Corporations Export Jobs?” (staff paper), from http://www.taxfoundation.org/news/show/156.html 15. “U.N. Workshop on Integrated Economic Statistics and Informal Sector,” Teheran, 10-13 November, 2007, UN Statistics Division from: http://www.google.com/search?hl=en&client=firefox-a&rls=org.mozilla:en-US:official&q=world+manufacturing+statistics&start=0&sa=

N Data Sources for Quantitative Studies 1. Employment: Annual averages (1000) created by Economagic from monthly BLS data. http://economagic.com/ 2. Unemployed: Annual averages (1000) created by Economagic from monthly BLS data. http://economagic.com/ 3. Unemployment rate: Annual averages (%) created by Economagic from monthly BLS data. http://economagic.com/ 4. Labor force: Annual averages (1000) created by Economagic from monthly BLS data. http://economagic.com/ 5. Interest rate: Annual averages (%) by Economagic from quarterly Federal Reserve Board of Governors quarterly data. http://economagic.com/ 6. Real GDP: Annual ($billion) from Economagic from the Federal Reserve Board of Governors data. http://economagic.com/ 7. Exports, Imports, Balance ($ million): Foreign Trade Statistics, Historical Series, U.S. Dept. of Census. http://census.gov/foreign-trade/statistics/historical/index.html 8. CPI, Medical CPI: from BLS data. http://economiagic.com/ 9. Medical value added: ($million, 2000=100) Bureau of Economic Analysis, Industry Economic Accounts, Historic Data, NAIS Data. http://www.bea.gov/industry/gdpbyinf_data.htm 10. GDP value added: ($million, 2000=100) Bureau of Economic Analysis, Industry Economic Accounts, Historic Data, NAIS Data. http://www.bea.gov/industry/gdpbyinf_data.htm 11. Gasoline, oil prices: Averages ($.001/gal., ‘82-‘84=100) by Economagic from Dept. of Energy monthly data. http://economagic.com/ 12. US GDP and World GDP: ($million, 2000=100) 2008 World Development Indicators. The World Bank, Washington, DC 2008. http://go.worldbank.ofg/U0FSM7AQ40 13.US Manufacturing % of GDP and World Manufacturing % of GDP: 2008 World Development Indicators. The World Bank, Washington, DC 2008. http://go.worldbank.ofg/U0FSM7AQ40


http://www.sfgate.com/cgi-%20bin%20/article.sgi?f=c/a/2008/BU6AUQMT9.DTL

http://www.sfgate.com/cgi-%20bin%20/article.sgi?f=c/a/2008/BU6AUQMT9.DTL

http://money.cnn.com/2008/02/01/news/companies/exxon_earnings/

http://money.cnn.com/2008/02/01/news/companies/exxon_earnings/

http://www.energytomorrow.org/ViewResource.ashx?id=5335

http://www.imf.org/external/pubs/ft/issues11/

http://www.taxfoundation.org/news/show/156.html

http://economagic.com/






http://census.gov/foreign-trade/statistics/historical/index.html

http://economiagic.com/

http://www.bea.gov/industry/gdpbyinf_data.htm

http://www.bea.gov/industry/gdpbyinf_data.htm


http://go.worldbank.ofg/U0FSM7AQ40

http://go.worldbank.ofg/U0FSM7AQ40

KATYUSHA COMPUTATIONS OR ROCKET SCIENCE: USING ECONOMIC ORDER QUANTITY TO ASSESS SUSTAINABILITY OF HOSTILE ROCKET OFFENSIVES

Johnnie B. Linn III

Division of Business Concord University Athens, WV 24712

ABSTRACT In the 2006 conflict of Hezbollah against Israel, Hezbollah launched a more or less constant number of rockets per day from southern Lebanon into Israel. This paper models Hezbollah as a “firm” that “delivers” launched rockets at a constant rate. Standard inventory theory is used to generate an economic order quantity (EOQ) of replacement rockets required to sustain the rate of launch indefinitely. Countermeasures to replacement deliveries affects Hezbollah’s order cost and inventory holding costs, so changes in EOQ can be used to measure impact of changes in countermeasures. Furthermore, Hezbollah begins with an extra inventory of missiles that can be drawn upon. Expected changes in EOQ are calculated for the cases of straight-line and accelerated draw. If measures of observed EOQ begin to fall behind predicted values, Hezbollah cannot sustain its rate of launch.

INTRODUCTION

In the 2006 conflict between Hezbollah and Israel, Hezbollah launched a more or less constant number of rockets per day from southern Lebanon into Israel. Hezbollah would get the maximum psychological impact from its rocket strikes if it fires off the same number of rockets each day and never runs out of them. The problem then is, if Hezbollah is modeled as a “firm” that “delivers” launched rockets at a constant rate, given the number of rockets in stock and the frequency and size of additions to that stock, how many rockets ought Hezbollah fire off each day? In business-speak, this is an economic order quantity problem. A business wants to select an order size, or economic order quantity, that minimizes its inventory-related costs, which is a sum of ordering costs and inventory holding costs, subject to a fixed demand each day. For Hezbollah’s suppliers, the math is the same, the only difference in their having previously selected a level of “demand” that they hope to sustain indefinitely.

THE MODEL In the standard economic order quantity (EOQ) model, as developed by Wilson (1934), a firm’s total cost for producing a good for which there is con nt d and is sta em

2 , 1

where Q is the order size, p is the price of the product, D is the daily demand, c is the cost of placing a new order, and h is daily unit holding cost. It will be noted Q for both order costs and holding costs is denominated in units per shipment; the Q for holding costs is multiplied by an implied number of one shipment received before draw on inventory begins. In the case of Hezbollah, the demand is the daily number of rocket strikes, which is assumed to be constant. The economic order quantity is ominated in rockets per shipment and is given by

den

2 , 2

The values of parameters c and h will be known only to Hezbollah and its suppliers, but we need know only their ratio in the right hand side of (2), which we can learn if we find the size of Q. Let us assume that all costs are associated with the rockets only, not launchers or launch crews, etc. The product is a launched rocket. The price is the opportunity cost of a launched rocket, that is, the amount of psychological impact that is not inflicted if the rocket is not launched. In addition, part of the opportunity cost of a launched rocket is associated with the possibility that it will be found and destroyed on its launcher. Let us express this as a probability z. For every successful launch there will have been 1/(1-z) attempts, so the feed, or number of rockets made available, needed to sustain a daily demand D would be D/(1-z). The cost of placing a new order (sending a shipment of rockets) does not include the cost of the rockets themselves, but of the logistics costs involved. Let the cost of a shipment when there is no threat of it being intercepted be c, and the proportion of shipments that are intercepted is n, then the order cost is c/(1-n) for each successful delivery. The values


of the rockets destroyed are not included in the order costs because the total number of rockets shipped (destroyed and undestroyed) is a constant regardless of the economic order quantity. The unit holding cost is the daily natural cost of deterioration in storage plus the daily probability that a rocket in storage is targeted and destroyed. Let us express this probability as s. The inventory holding c /(1-s). Our modified economic orde

ost is then hr quantity equation is

2 11 1

. 3

One final adjustment to the equation must be made. In steady-state, the economic order quantity is twice the average size of inventory. This accounts for the “2” in the formula. There is no extra inventory, and each shipment of rockets arrives just as inventory is exhausted. In the case at hand, though, Hezbollah has extra inventory that can be drawn upon, so, in an intermediate state the economic order quantity is less than twice the average size of inventory. Let us suppose that Hezbollah chooses to draw down its inventory while it awaits its first (or next) shipment of rockets after hostilities have begun, and that the amount of its reserve inventory that it releases is Q1. The daily order cost term will have Q + Q1 in its denominator instead of Q, indicating the longer period over which it is possible to distribute order costs. The holding cost term will likewise have Q replaced with Q + Q1 since the release of inventory is a new variable and its holding cost will h counted for. The economic or n

ave to be acder qua tity will be given by

2 11 1

. 4

When the next shipment arrives, the EOQ is recalculated at that time based on a new amount of reserve inventory to be released, likely a lesser amount. The size of EOQ increases as Q1 decreases, but if maintainable feed has been selected and there have been no changes in costs, the sum of EOQ and reserve inventory release will be invariant. Incidentally, Equation (4) also gives us the solution for an optimal release of reserve inventory for the current EOQ and a constant demand. We can also calculate the shipment interval L needed to sustain the demand. That would be the ratio of the sum of economic order quant tory release to feed or

ity and reserve inven

2 1 11

. 5

Now what happens if Hezbollah has selected an unsustainable level of feed? Shipment size will not increase

enough to cover planned reductions of release from reserve inventory. Release from reserve inventory would have to be increased and eventually there would have to be a reduction in feed.

TWO STRATEGIES FOR HEZBOLLAH Let us explore two mathematically tractable strategies for Hezbollah. We can call these the straight-line method and the accelerated method, suggestive of methods for depreciation. In the straight-line method, Hezbollah sets an expected number of days T to exhaust reserve inventory. We count the value of T from the time the first shipment arrives (assuming that Hezbollah had a sufficient number of extra rockets, not counted in inventory, to maintain feed before the first shipment arrived). Afterwards, for the remainder of period T, order size and release from inventory are constant. Also, during that time, some of the rockets in inventory will have been destroyed. The sustainable release from inventory is the expected nu er o d rockets allocated over T days, and is appr i

mb f undestroyeox mated by

1 23 , 6

where K is the initial inventory of rockets. The effective hazard rate in storage for a given rocket is s when inventory is large because the probability of it being withdrawn from inventory that day is tiny, but the effective hazard rate on the last day of release is s/2, so the best estimator of the hazard rate over the life of the inventory is the geometric mean of s and s/2, or s/1.5. For the straight-line method to be viable, Hezbollah will have to show a constant shipment size. If it starts to decrease, the offensive will falter. In the accelerat thod, inventory declines in propo ry,

ed me release fromrtion with invento as follows:

2 11 1

, 7

where K(t) is inventory at the end of period t. So shipment size starts at zero and increases at a decreasing rate. If it levels off, Hezbollah will have made a seamless transition from the period of working off inventory to the period of sustaining feed indefinitely from shipments.

A SAMPLE CALCULATION Figure 1 shows the EOQ’s for the straight-line and accelerated methods for an initial inventory of 10,000 rockets, 100 strikes per day, a ratio of order cost to unit holding cost of 100, an offensive length of 50 days, a 5 percent casualty rate each for launches and shipments, and a 5 percent daily casualty rate for rockets in inventory. The daily feed is 105 rockets, the EOQ for the straight-line method is 94 rockets per shipment, the release from


76


77

inventory is about 51 rockets per shipment, and the interval between shipments is about 1.4 days. Hezbollah can maintain the offensive with an EOQ of 94 for about 50 days. Afterwards, it would need an EOQ of 145 rockets per shipment. Under the accelerated method, Hezbollah would need about 90 days for its deliveries to come fully on stream.

SO WHAT REALLY HAPPENED? The offensive began on July 12, 2006 and lasted 34 days (2006 Lebanon War, 2009). About 4,000 rockets were fired into Israel during the offensive, at a rate of more than 100 per day except during the 2-day hiatus after the Israeli attack on Qana. Most of the rockets fired were the short-range Katyushas, of which Hezbollah’s inventory was about 13,000 (Gardner, 2006). A week before the end of the offensive, the Israeli Defense force claimed to have destroyed about a third of the inventory of Katyushas (Hezbollah armed strength, 2009). At a briefing On July 18, Israeli Air Force Commander Brig. Gen. Amir Eshel stated that a “number of trucks” transporting rockets from Syria to Hezbollah had been destroyed (Greenburg, 2006). What can we make of these numbers? From the expression containing s in Equation (6), for a value of T of 27 at the time of the BBC reference, we obtain a value of s of about 0.023. A “truck” is likely not larger than a 2.5-ton cargo capacity

light medium military tactical vehicle, as Gen. Eshel emphasized that targeted trucks were disguised as civilian vehicles. The smallest Katyshua rocket weighs 92 pounds (Katyusha rocket launcher, 2009), so about 32 of them could be carried in a 2.5-ton payload. It is hard to assess how many trucks a “number” of trucks is in the report given 6 days into the campaign. It is reasonable to assume at least one truck was destroyed per day. If at least one truck survives the trip per day, that would put an upper limit on L, the interval between shipments, of half of a day. Using 13,000 for K, 0.023 for s, and a lower limit of 34 for T, we get an upper limit of release from inventory, from Equation (6), of about 87. Adding to this an EOQ of 32 (if we assume that a shipment comprises only one truck, not a convoy), we get an estimate of combined release and EOQ of 119. Assuming that feed is 125 and n is 0.5, we get a rough estimate of c/h from Equation (4) of about 116. This estimate seems reasonable, as one would expect that the cost of placing an order would be much greater than the daily hazard of one rocket in inventory.

CONCLUSION From the sketchy data available, estimates can be made of the ratio of order cost to holding cost for Hezbollah, the casualty rates, and release from inventory. The data are not solid enough for us to make an assessment of whether EOQ is a viable model for Hezbollah; for that, we will have to wait for the belligerents to weigh in.

Figure 1. Straight-line and Accelerated Release

0

50

100

150

200

0.0 20.0 40.0 60.0 80.0 100.0

Rockets/Shipmen

t

Days

Hezbollah EOQEOQ (str) EOQ (acc)

REFERENCES 2006 Lebanon War, 2009. In Wikipedia, The Free Encyclopedia. Retrieved 14:40, July 19, 2009, from http://en.wikipedia.org/w/index.php?title=2006_Lebanon_War&oldid=302953706 Gardner, 2006. Hezbollah Missile Threat Assessed. http://news.bbc.co.uk/2/hi/middle_east/5242566.stm Greenburg, 2006. IAF foils rocket transport, http://www.ynetnews.com/articles/0,7340,L-3278029,00.html#n

Hezbollah armed strength. (2009, June 11). In Wikipedia, The Free Encyclopedia. Retrieved 17:06, June 11, 2009, from http://en.wikipedia.org/w/index.php?title=Hezbollah_armed_strength&oldid=295812892 Katyusha rocket launcher. (2009, July 14). In Wikipedia, The Free Encyclopedia. Retrieved 11:58, July 14, 2009, from http://en.wikipedia.org/w/index.php?title=Katyusha_rocket_launcher&oldid=302031094 Wilson, R. H. (1934) "A Scientific Routine for Stock Control" Harvard Business Review. 13: 116-128.


78

http://news.bbc.co.uk/2/hi/middle_east/5242566.stm

http://en.wikipedia.org/w/index.php?title=Hezbollah_armed_strength&oldid=295812892

http://en.wikipedia.org/w/index.php?title=Hezbollah_armed_strength&oldid=295812892

http://en.wikipedia.org/w/index.php?title=Katyusha_rocket_launcher&oldid=302031094

http://en.wikipedia.org/w/index.php?title=Katyusha_rocket_launcher&oldid=302031094

DRUG APPROVALS AND DRUG SAFETY: PRELIMINARY RESULTS

Natalie Reaves Department of Economics

Rowan University Glassboro NJ 08028

ABSTRACT These preliminary results are an attempt to address the impact of government regulation on the safety on new drugs. Many studies have used drugs withdrawals or Black Box warnings as indicators of drug reliability. Adverse drug event data are supplied to the FDA by consumer and medical providers. This paper examines whether the Prescription Drug User Fee Act (PDUFA) led to an increase in adverse drug events reported by consumers and medical providers. This significant change in drug regulation was designed to speed up the drug approval process. This purpose of this paper is to determine what impact, if any, this regulatory change had on drug safety.

INTRODUCTION

There has been much publicity within the last few years regarding the safety of prescription drugs after they have been made available to the public. Some drugs have been withdrawn from the market while others were required to give additional notice of possible severe side effects. For example, Vioxx, a prescription painkiller, was withdrawn from the market in 2004 due to cardiovascular concerns leading to stroke and congestive heart failure. In addition, certain antidepressants were discovered to increase the suicide risk in children and adolescents, and now they must warn users regarding these possible side effects. At the same time, there is pressure for the Food and Drug Administration (FDA) to approve drugs and provide relief to patients with medical conditions in as timely a manner as possible. Patient groups and the pharmaceutical industrial have long complained about the long, expensive drug approval process in the U.S. that they believe is keeping needed remedies off the market. So, there is a tradeoff that regulatory agencies, such as the FDA, face when they attempt to get remedies to patients as soon as possible while accounting for potential, even deadly side effects. Researchers have long understood the tradeoffs between the speed of approval of a new drug remedy vs. the safety of the drug. On one hand, there is the incentive to get new drugs to the market as soon as possible to provide relief from health ailments. However, drugs must be tested regarding both their safety and efficacy before they can be approved by the Food and Drug Administration (FDA This paper attempts to answer a basic question: Does the speed in

which the FDA reviews a new drug application for marketing approval affect the drug’s safety?

BACKGROUND

Firms in the pharmaceutical industry are required to undergo a very lengthy and costly regulatory process before a drug is approved by the FDA. Drug firms spend, on average, $800 million to $1 billion dollars just to bring one successful drug to market (DiMasi et al 2003). In addition, the drug development process averages about 10-15 years from a drug’s synthesis to marketing approval by the FDA. Finally, for every 5,000-10,000 compounds that drug firms test, only one of these will become a marketed drug (DiMasi et al 2003). The high cost and risk of pharmaceutical R&D is associated with the firm’s need to develop drugs that are both “safe” and “effective” against the targeted illness. As a result, drug companies seek patentable “blockbuster” remedies to compensate for the high attrition, costs and risks associated with drug development.

PRESCRIPTION DRUG USER FEE ACT (PDUFA)

In 1992, the U.S. Congress passed the Prescription Drug User Fee Act (PDUFA) to provide additional resources to the Food and Drug Administration (FDA) to speed up the drug approval process. Under the PDUFA, the FDA collects fees from drug companies to offset the costs of the drug review and approval process. The original Act expired in 1997, and was extended in 1997, 2002 and 2007. The FDA uses these funds to hire additional resources and reviewers to expedite the drug approval process. The goal was to make new drugs available more quickly to patients by using the fees collected from industry to offset review costs. The passage of the PDUFA required firms to pay fees with their new drug applications, annual fees on each establishment and renewal fees for approved products. The drug industry pays these fees in exchange for “performance goals” set by the FDA, namely a shorter FDA approval time. However, recent controversies regarding adverse drug outcomes, has led to criticism that the FDA traded speed for safety. So, this paper is an attempt to measure the impact, if any, of the PDUFA on drug safety.


79

The FDA classifies new drug applications as either Priority or Standard. Priority drugs represent a significant therapeutic advantage over existing remedies. However, drugs designated as Standard, represent little additional benefit over existing remedies. The “performance goals” set by the FDA under the PDUFA warrants that 90% of all Priority new drug applications should have their FDA reviews completed within 6 months after submission of the New Drug Application to the FDA. In addition, the FDA’s performance goal for Standard drugs is for 90% of these new drug applications to be reviewed within 12 months. Table-1 reveals that the median time for the FDA to review a Priority drug has fallen over the past decade and approaches the performance goal set by the FDA of 6 months. The FDA review time needed to review a Standard drug has fallen considerably, since 1993 and is approximately 11-12 month.

LITERATURE REVIEW

The pharmaceutical industry is a heavily regulated industry that has been studied extensively by economists. In recent years, there has been an attempt to determine the impact of the PDUFA on approval times and drug safety. One study found that the PDUFA was responsible for 42% reduction in approval times for new drugs from 1991-2002 (Berndt et al 2004). Other studies have attempted to ascertain the impact of the PDUFA on drug safety. An issue that develops when attempting to quantify drug safety issues is how do you measure a drug’s safety? A common measure of safety, in previous studies, was to examine “safety based withdrawals” of drugs after market approval. A study done by the U.S. Government Accounting Office (GAO) in 2002 found that drugs withdrawn following approval increased from 1.56% for 1993-1996 to 5.35% for 1997-2001. However, a Tufts Center for the Study of Drug Development study (2005) found that faster approval times did not correlate with increased drug safety withdrawals. Another measure of drug safety used in previous research was to examine the number of Black Box Warnings (BBW’s) given to approved drugs that warn of potentially serious side-effects. A 2006 study compared new drug approvals from before and after the 1992 PDUFA and found no statistically significant difference in the rates of BBW’s in the two periods (Begosh et al 2006). Finally, an alternative measure used to assess safety of approved drugs is to examine adverse drug events (ADE’s) reported to the FDA (Olson 2008).

ADVERSE DRUG EVENTS

The FDA maintains an Adverse Events Reporting System (AERS). These reports are filed electronically by medical providers such as physicians, hospitals. pharmacists and patients can file them as well. The FDA defines a serious adverse event as an outcome that includes either death,

hospitalization, is life-threatening, can lead to disability or congenital anomaly and/or other serious outcomes. The FDA reports on three types of serious ADE’s (a) ADE’s that require hospitalization (b) ADE’s that result in death and (c) Total serious ADE’s. Table-2 is a listing of all adverse events reported to the FDA. There has been as almost doubling of adverse events reported to the FDA. Table 3 is a breakdown of who reports these adverse events. It appears that health care providers provide the bulk of complaints to the FDA. Overall, the number of complaints received by both types of informants has grown significantly. Finally, Table 4 is a further breakdown of the types of events reported. The number of deaths reported has more than doubled during this period. However, caution is advised since these numbers are self-reported and do not necessarily mean the suspected product was the cause of the outcome.

METHODOLOGY AND DATA A negative binomial model can be used when the dependent variable involves counts that are over-dispersed and do not fit the Poisson distribution (Cameron & Trivedi 1998). Adverse Drug Events are highly skewed in the first two years after a drug is approved, so this type of estimation has been used in previous studies of this type (Olson 2008). The negative binomial model will model the occurrence of adverse drug events (ADE’s) for new molecular entities introduced between 1990 and 2005 in the U.S. The dependent variable is the count of ADEs for each drug i in years 1 and 2 after marketing approval in year t. Yi = ∑2

j=1 ADEi, t+j (1)

There were 455 New Molecular Entities approved between 1990-2005 by the FDA. Several types of data were excluded including vaccines, salts, chemicals, or variations of existing remedies. Also, products withdrawn within the first two years or where there was missing information was excluded. So, the final dataset included 411 NME’s approved from 1990-2005.

MODEL

DEPENDENT VARIABLE: Adverse Drug Events (ADE)

ADE = f (NDA phase, PDUFA, orphan drugs status, priority

rating, black box warning, year specific effects) (2) Most ADE’s are reported within the first two years after a drug is launched and then trail off. So, this paper will confine itself to New Molecular Entities (NME’s) adverse drug events that occur within the first 2 years after the launch of


80


81

the product for newly approved. NME’s are drugs that include an active ingredient that has not previously been approved for marketing in the United States in any form. These data were obtained from the FDA’s Adverse Event Reporting System (AERS).

EXPLANATORY VARIABLES PDUFA: This represents the new regulation in 1992. This will equal 1 for all drugs marketed in 1992 and later, 0 otherwise. NDA phase: The FDA review time (in months) from the date of first submission of the new drug application until the date of FDA approval. Rating: The drug rating given by the FDA. If a drug is given a Priority rating it receives a “1.” A standard rating will receive a “0.” Black Box Warning (BBW): Did this drug receive a Black Box warning at approval? If the drug received a Black Box warning, the value is “1.” Orphan drug: Was this drug classified as an orphan drug? Orphan drugs are designed to treat patient populations less than 200,000 persons in the U.S. If the drug received an orphan drug designation, this will take the value of “1.”

PRELIMINARY RESULTS What follows are preliminary results for the data. The model will be extended, in the future, to include other variables. These preliminary results suggest that faster review times are linked to increased ADE’s since the coefficient is negative and significant. For example, a reduction in the NDA review phase by one month is correlated with a 1.5% increase in ADE’s. The regulatory variable, PDUFA, appears insignificant and this indicates that the impact of drug industry fees on adverse events is immaterial once you control for the FDA review period. The Black Box Warnings

are correlated with adverse outcomes and this coefficient is positive suggesting that drugs that received a BBW at approval are more likely to incur an adverse event. Moreover, if a new drug received a Priority rating from the FDA, it is associated with an increase in adverse drug events. The Orphan drug status coefficient is negative but insignificant, suggesting this special status had no impact on ADE’s, after controlling for other factors.

FURTHER RESEARCH TO BE COMPLETED

At this stage, the dataset needs to be expanded to include both drug and patient characteristics. For example, adding the therapeutic class that a particular drugs belongs to will further this analysis since some types of drugs may experience more adverse drug events than others. In addition, patient characteristics such as gender and age, may affect the possibility of adverse drug events. Moreover, it would be useful to disaggregate the adverse events by examining the impact of the above variables on ADE’s that resulted in death and/or ADE’s that resulted in hospitalization in comparison to overall serious ADE’s.

CONCLUSION

These preliminary results are an attempt to address the impact of the PDUFA on the safety on new drugs using adverse event data. Previous studies have used drugs withdrawal or Black Box warnings as indicators of drug reliability. However, these actions are in the hands of the regulator, FDA. Adverse drug events data are driven by the consumer, and medical providers allowing a broader array of outcomes. Previous studies using adverse event data used older data and/or used approval dates as the time in which the drug was available to consumers, instead of marketing launch date which gives a more accurate picture of the time frame in which to monitor adverse outcomes. This paper will continue to investigate the impact of the PDUFA on drug safety in future research.

TABLE 1 FDA APPROVAL TIMES FOR PRIORITY AND STANDARD NEW DRUG APPLICATIONS 1993-2008 Calendar Year PRIORITY

DRUGS STANDARD

DRUGS

# Approved Median FDA Review Time (months)

# Approved Median FDA Review Time (months)

1993 19 16.3 51 26.9 1994 16 13.9 45 21.0 1995 16 7.9 67 18.7 1996 29 7.8 102 17.8 1997 20 6.3 101 15.0 1998 25 6.2 65 12.0 1999 28 6.1 55 13.8 2000 20 6.0 78 12.0 2001 10 6.0 56 14.0 2002 11 13.8 67 15.3 2003 14 7.7 58 15.4 2004 29 6.0 89 12.7 2005 22 6.0 59 13.1 2006 21 6.0 80 13.0 2007 23 6.0 59 10.4 2008 18 6.0 55 13.1

Source: FDA

TABLE 2

TOTAL ADVERSE EVENTS REPORTED TO THE FDA

Year Total Received

2000 266,866

2001 284,762

2002 322,493

2003 370,843

2004 422,930

2005 463,819

2006 471,394

2007 482,154

2008 526,527


82

TABLE 3

ADVERSE DRUG EVENTS BY TYPE OF INFORMANT: 2000 -2009

Year Healthcare Provider: Physician

Healthcare Provider: Pharmacist

Other Healthcare Provider

Total Healthcare Provider

Consumer

2000 59,090 18,794 20,191 98,075 46,249 2001 66,001 19,050 23,685 108,736 39,517 2002 67,967 17,184 27,944 113,095 30,282 2003 79,793 19,410 39,392 138,595 48,352 2004 92,737 20,329 46,056 159,122 74,644 2005 106,362 21,540 43,820 171,722 105,308 2006 113,444 21,512 49,827 184,783 127,475 2007 121,000 21,343 60,343 202,686 174,216 2008 154,044 27,070 88,651 269,765 226,647 2009 (Q1) 42,632 7,023 24,800 74,455 55,752

Source: FDA

TABLE 4 PATIENT OUTCOMES FOR AERS FROM 2000 – 2009 (1Q)

Year Death Serious* 2000 19,445 153,818 2001 23,988 166,384 2002 28,181 159,000 2003 35,173 177,008 2004 34,928 199,510 2005 40,238 257,604 2006 37,465 265,130 2007 36,834 273,276 2008 49,958 319,741 2009 (Q1) 15,373 87,676

Source: FDA


83


84

TABLE 5 NEGATIVE BINOMIAL REGRESSIONS RESULTS FOR SERIOUS ADE’s 1990-2005

VARIABLE RESULTS

PDUFA -0.06(0.45)

NDA PHASE -0.015(0.007)**

RATING .72(0.13)**

BBW .83(0.20)**

Orphan -0.45(0.43)

DUM91-DUM 05 ___

____________________________________

Dependent variable = Adverse drug events Count; N = 411

Standard errors are in parentheses. Significant at the 0.1 level. Significant at the 0.05 level. Significant at the 0.01 level.

REFERENCES

1. Baciu, A., K. R. Stratton, S. P. Burke, and Institute of Medicine (U.S.). Committee on the Assessment of the US Drug Safety System. 2007. The future of drug safety: Promoting and protecting the health of the public. Washington, D.C: National Academies Press.

2. Begosh, A., J. Goldsmith, E. Hass, R. W. Lutter, C. Nardinelli, and J. A. Vernon. 2006. Black box warnings and drug safety: Examining the determinants and timing of FDA warning labels National Bureau of Economic Research, Inc, NBER Working Papers: 12803.

3. Berndt, E. R., A. H. B. Gottschalk, T. Philipson, and M. W. Strobeck. 2004. Assessing the impacts of the prescription drug user fee acts (PDUFA) on the FDA approval process. National Bureau of Economic Research, Inc, NBER Working Papers: 10822.

4. Cameron, A. C., and P. K. Trivedi. 1998. Regression analysis of count data Econometric Society Monographs, no. 30; Cambridge; New York and Melbourne: Cambridge University Press.

5. Dickson, M., and J. P. Gagnon. 2004. Key factors in the rising cost of new drug discovery and development. Nature Reviews.Drug Discovery 3, (5) (05): 417-29.

6. DiMasi, J. A., R. W. Hansen, and H. G. Grabowski. 2003. The price of innovation: New estimates of drug development costs. Journal of Health Economics 22, (2) (03): 151-85.

7. DiMasi, J. A. 2002. The value of improving the productivity of the drug development process: Faster times and better decisions. PharmacoEconomics 20 Suppl 3, : 1-10.

8. Food and drug administration: Effect of user fees on drug approval times, withdrawals, and other agency activities: GAO-02-958. 2002. GAO Reports (09/17): 1.

9. Grabowski, H. 2002. Patents, innovation and access to new pharmaceuticals. Journal of International Economic Law 5, (4) (12): 849-60.

10. Olson, M K. 2008. The risk we bear: The effects of review speed and industry user fees on new drug safety. Journal of Health Economics 27, (2) (3): 175-200.

11. ———. 2004. Are novel drugs more risky for patients than less novel drugs? Journal of Health Economics 23, (6) (11): 1135-58.

12. Pharma. (2009). 2009 Industry Profile. Pharmaceutical Manufacturers Association of America. Washington, DC.

13. Philipson, T. J., and E. Sun. 2008. Is the food and drug administration safe and effective? Journal of Economic Perspectives 22, (1) (Winter): 85-102.

14. ———. 2008. Cost-benefit analysis of the FDA: The case of the prescription drug user fee acts. Journal of Public Economics 92, (5-6) (6): 1306-25.

15. Tufts Center for the Study of Drug Development, 2005. Drug Safety withdrawals in the U.S. not linked to speed of FDA approval. Tufts CSDD Impact Report 7 (5), 1-4.

16. U.S. General Accounting Office (GAO) Food and Drug Administration: Effect of User Fees on Drug Approval Times, Withdrawals and other Agency Activities. 2002. Publication GAO: 02-958 (220).


FACTORS IMPACTING THE THROWAWAY SOCIETY: A SUSTAINABLE CONSUMPTION ISSUE

John McCollough Department of Business and Economics

The Pennsylvania State University-Lehigh Valley 8380 Mohr Lane

Fogelsville, PA 18051

ABSTRACT The purpose of this paper is to gain a deeper understanding behind the positive correlation that exists between income and the generation of residential municipal solid waste from a theoretical context. This paper theorizes that there are deeply ingrained micro behaviors which are responsible for driving the relationship between income and residential municipal solid waste. More specifically, a theory is presented in this paper which asserts that as a consumer’s income increases, time becomes more valuable. The increase in income enables the consumer to purchase more goods which, in turn, contributes to an increase in solid waste. However, the increase in solid waste is magnified due to the fact that consumers, in order to economize on time, begin to substitute out of and away from reusable goods and into disposable goods. This is how societies become labeled as “throw-away societies.”

INTRODUCTION

The purpose of this paper is to gain a deeper understanding behind the positive correlation that exists between income and the generation of residential municipal solid waste from a theoretical context. Indeed, many scholars who write on issues of solid waste have noted that there has been little contribution to the subject in the way of economic reasoning and theory which links the two variables.

i Obviously as

incomes grow consumers will consume more. This has been shown to be true both in theory and in empirical research. It has also been shown empirically that higher income households do, in fact, generate more municipal solid waste (MSW).

ii But the empirical evidence has never addressed

why this relationship exists. To say that higher incomes lead to greater amounts of solid waste is not so straight forward. Even to say that increased consumption generates more solid waste is not altogether apparent. For instance, households with higher incomes may consume more tangible products than households with lower incomes, but unless those products are discarded thus entering the waste stream, the higher income and increased consumption does not translate into solid waste. Households with higher incomes may simple hold a larger inventory of the same kinds of products or simple have a greater array of different products or goods stored on their premises.

Therefore, the key to understanding the relationship between income and MSW is to understand how much of a household’s typical bundle of goods is represented by disposable goods and how much is represented by reusable goods. This paper theorizes that as incomes increase, households will shift purchases away from reusable goods to disposable goods.

iii As this happens, the generation of gross

residential MSW will increase.iv

The paper demonstrates, therefore, that it is not necessarily income growth that is responsible for the buildup of residential municipal solid waste, rather it is the way in which consumers choose to spend their incomes. If consumers repair and maintain a product for reuse, there is no significant impact on residential municipal solid waste. However, if they continue to purchase disposable products and if they continue to replace products that could be reused via repairs and maintenance, the impact on residential municipal solid waste is much greater. Thus income has only an indirect role in helping to create the buildup of residential municipal solid waste, since it is really income that drives the consumer to substitute out of and away from reusable goods and into disposable goods as they attempt to economize on their opportunity cost of time. The presentation of the model developed in this paper in conjunction with the existing empirical evidence supporting the positive correlation between income growth and residential MSW should be of interest to those engaged in the Environmental Kuznets Curve debate. Despite the fact that the empirical evidence shows the generation of residential MSW to be positively correlated to income at the local, national, and international levels it does not rule out a non-linear, Environmental Kuznets Curve type, relationship for the residential MSW environmental indicator. It could simply mean that the wealthiest of societies (or municipalities) have yet to reach that turning point in the future. Only a more enhanced theoretical model on the generation of solid waste could answer this question. Empirically testing of the model, specifically linking the increase in disposable goods to an increase in either gross residential MSW or net residential MSW will be left for further research due to the difficulty in determining exactly how much of the waste stream consists of disposable goods and how much consists of reusable goods.

v


86

There are societal problems in regards to the environment when consumers shift their mix of purchases away from reusable goods and into disposable goods. This phenomenon creates a “throw-away society”. Indeed, many advocates for the reduction of MSW use the “three Rs mantra” which is reduce, reuse, and recycle. They feel that in terms of impact on solid waste, reduction is preferable to both reuse and recycling, and that reuse is preferable to recycling. Reuse and recycling are thought to help conserve natural resources and reduce resource extraction while saving energy and landfill space which helps alleviate the environmental problems associated with net MSW. However, the reuse of a product is, in general, more environmentally friendly than recycling

vi

since recycling has to go through a secondary production stage to bring the product back into a reusable form.

vii

THEORETICAL MODEL

The model developed in this paper starts with the assumption that every type of tangible or physical good available for purchase can be classified as either a disposable good (X

1) or

a reusable good (X2) and that individuals seek to maximize

utility from the consumption of a bundle of disposable and reusable goods subject to their own budget constraint. Disposable goods are defined as goods that are used n times and then discarded. Reusable goods are those goods that are used n times and then some form of maintenance is performed on the good in order for the reusable good to be used again. Therefore, in most cases, a reusable good performs the same function as a disposable good, but has a longer life span. At some point in time the reusable good eventually wears out and then it must also be discarded just like the disposable good. In this respect, each good eventually (the disposable and the reusable good) enters the waste stream. However, since the reusable good is used more times than its disposable counterpart, the impact on municipal solid waste (on a per use basis) is less than that of a disposable good. It is assumed in this model that there are no significant technological differences between a reusable or disposable good. Both the reusable and disposable good have the same functionality and same design features. To clarify, the distinguishing feature between the two goods is that some form of maintenance is performed on the reusable good and then the reusable good is used again. No maintenance is required of the disposable good. Therefore, the disposable good is used n times and then discarded while the reusable good is used n + m times (where m is the number of further uses a consumer gets from the product due to repairs and maintenance) before it is discarded. For instance, a disposable paint brush is used once and then discarded. A reusable paint brush is used once and then

maintenance is performed (the paint brush is cleaned with soap and water, or turpentine, then dried) before the paint brush is used again. Another example would be a disposable lighter. Once the fuel is used in the disposable lighter the lighter is discarded. With a reusable lighter, once the fuel is spent, maintenance must be performed on the reusable lighter by refilling the lighter with fuel before the lighter can be used again. As a final example, disposable paper cups, plates, or utensils are used once and then discarded. The reusable counterparts to these products are also used once and then some form of maintenance is performed on them so that they can be used again (i.e., washing and cleaning of the cups, plates, utensils). In essence, all physical and tangible goods can be thought of in terms as being either disposable or reusable regardless of how many uses it was designed to provide. For instance, assume that a long lasting, durable product such as a washing machine was engineered to provide X hours of service and that after that time the washer either wears out, breaks down, or is in need of some form of maintenance, at that point it can then either be repaired and reused or it can be replaced and discarded without being repaired. If the washing machine is replaced and discarded without being repaired, then it is considered disposable. If the washing machine is repaired after it breaks down so that it can be reused, it is considered reusable. Thus, washing machines are also considered to be either disposable or reusable. Indeed many household products that are now commonly thought to be only disposable had a reusable counterpart at one time. Strasser (1999) gives a good historical account of how the evolution towards a throw-away society took place. For example, at one time in America all towels and napkins were made of cloth and were reusable. Eventually disposable paper towels and napkins were introduced onto the market. These products are now so commonplace that it is hard for consumers to think of anything but disposable paper towels and napkins. Gross residential MSW, is generated by both disposable goods (X

1) and reusable goods (X

2) and can be expressed as

S = S1 + S

2. Where S

1 = r

1X

1 and S

2 = r

2X

2. Furthermore, r

1

= (a1)*(b

1)*(c

1), where (b

1) is the proportion of the good, by

weight, that ends up as solid waste, (a1) is one divided by the

number of uses extracted from the good in question, and (c1)

is simply the number of pounds per unit the good weighs.viii

The same applies for r

2, where r

2 = (a

2)*(b

2)*(c

2). For

example, a tangible good, such as a cigarette lighter (disposable or reusable) has a (b) factor of 1, since, after the product is finally fully used, the entire content is discarded. If a disposable lighter has only twenty uses then the (a

1) factor

= .05 (i.e., 1/20), whereas, the reusable lighter that has 2000


uses before finally wearing out would have an (a2) factor of

.0005 (i.e., 1/2000). Assuming that the weight of a disposable or reusable lighter is .3 lbs, then the total impact on solid waste from a disposable lighter after each use would be (a

1)*(b

1)*(c

1)X

1 = r

1X

1= (.05)*(1)*(.3lbs)X

1, or .015 lbs of X

1. The total impact on solid waste from each use of a

reusable lighter would be (a2)*(b

2)*(c

2)X

2 = r

2X

2 =

(.0005)*(1)*(.3lbs)X2, or .00015 lbs of X

2.ix

This example illustrates that the impact on solid waste from a reusable good is less than that from a disposable good because the reusable good lasts longer. Another example would be disposable and reusable paint brushes, both of which have a (b) factor of 1 since the entire good is disposed of once it is fully used. The disposable paint brush would have an (a

1) factor of 1 since it is only used once

and then discarded.x Assuming the reusable paint brush is

used 20 times and then discarded, its (a2) factor will then be

.05. If the (c1) factor equals .5 lbs (still wet with paint or

water/turpentine) for either a disposable or reusable paint brush, then the total impact on solid waste from one use of a disposable paint brush is (a

1)*(b

1)*(c

1)X

1 =r

1X

1= (1)*(1)*(.5

lb)X1, or .5 lbs of X

1 . The total impact on solid waste from

one use of a reusable paint brush would be (a2)*(b

2)*(c

2)X

2

=r2X

2= (.05)*(1)*(.5 lbs)X

2, or .025lbs of X

2 . Again, this

example demonstrates that the impact from reusable goods on solid waste is less than disposable goods since the reusable good lasts longer. In every instance, r

1 > r

2 . The

packaging of disposable and reusable goods will be assumed to be incorporated into r

1 (via b

1) and r

2 (via b

2). This

reinforces r1 > r

2 since disposable goods are purchased more

frequently then reusable goods and, therefore, require more packaging on a per use base. Of course, some waste is generated when a repair or maintenance is made to certain reusable goods, such as the replacing of a belt in a washing machine. The old belt enters into the waste stream. This waste also is part of the b

2 factor

and is assumed to be negligible. Therefore the inequality of r1

> r2

would still hold. By expressing S1 = r

1X

1 =

(a1)*(b

1)*(c

1)X

1 and S

2 = r

2X

2 = (a

2)*(b

2)*(c

2)X

2, the impact

that all goods (disposable and reusable) have on solid waste can be calculated. It is noted here that not all products have a reusable counterpart. Examples of these would be service products or food products.

xi

The full price of a disposable or reusable good is denoted as Π

i. The full price consists of a direct price (cost) denoted as

pi and two separate indirect prices (or costs). The first

indirect price is denoted as wti m

, where w equals the wage rate one can earn by working in the marketplace and t

i m

equals the total time it takes to consume and maintain a product. The second indirect price is denoted as wt

i s, where

w, once again, equals the wage rate and ti s

is the total search time spent shopping for a product. Therefore, Π

i =(p

i + w(t

i m+ t

i s ) ).

The first indirect cost, wt

i m , is simply the value of time it

takes to consume and maintain a product. For instance, the consumer can consume a domestically prepared meal at home or consume a meal in a fast food restaurant. It takes considerably more time to prepare and consume a meal at home than it does to consume a meal at a fast food restaurant. The time spent preparing a meal at home is valuable and could have been spent working for a wage in the market place. Conversely, the time spent maintaining a reusable good (such as repairing or maintaining either a television set or washing machine when either breaks or cleaning a reusable paint brush) could be spent working for a wage in the marketplace. Therefore the higher the wage rate, the greater the indirect cost of consuming and maintaining a product.

xii Disposable goods are thought to be purchased

because of their time saving/convenience qualities. At the same time there is also a search cost involved for the consumer when purchasing a product. The consumer must be educated as to which product is the best bargain. The consumer must also drive to the store, park the car, select the product, and stand in line to make the purchase. All this takes time and again the consumer could have spent this time working for a wage in the marketplace, hence the term “search cost”. It is assumed that the search cost is higher for disposable goods than reusable goods because they are purchased more frequently. However, it is also assumed that the search cost decreases the more frequently a particular disposable good is purchased because the consumer has more experience at purchasing the product. Furthermore, search cost is also assumed to decline as the price of the product and/or the complexity of a product declines.

xiii

Typically, as a consumer’s wage increases both a substitution and an income effect will transpire to change the optimal mix of consumption for both disposable and reusable goods. First the consumer will purchase more of both disposable and reusable goods since they have more income to spend on both items (the income effect). However, as a consumer’s wage increases, relatively more and more disposable goods will be purchased compared to reusable goods. This substitution effect occurs because the consumer is seeking to economize on the indirect cost (wt

im) that is associated with

repairing and maintaining each good. To repeat, the indirect


repair and maintenance cost of reusable goods is higher than for disposable goods. The Lagrangian for this model is:

max L= U(X1, X

2 , S, e )+λ( wT +Y - Π

1 X

1 - Π

2 X

2

- h (r1 X

1 +r

2 X

2 ) ) (1)

Where X

1 = disposable good and X

2 = reusable good. X

1 has

a lower maintenance cost than does X2 (i.e.; wt

1m < wt

2m) but

X1 has a higher search cost than does X

2 (i.e.; wt

1s < wt

2s). It

is assumed in this paper that the search cost is less than the maintenance cost for both reusable and disposable goods. In other words, it is assumed that it takes less time to purchase a product than to repair and maintain the product.

xiv

Π

1 = (p

1 + w(t

1m+ t

1s ) ) = total cost of one unit of a

disposable product Π

2 = (p

2 + w(t

2m+ t

2s ) ) = total cost of one unit of a

reusable product S = S

1 + S

2 = (r

1X

1 +r

2X

2 ) = total amount of solid

waste e = environmental awareness by the consumer h = unit cost of solid waste removal I = income = wT+Y Where wT = total wage income and Y = non wage income Although S and e enter the utility function, they are not choice variables. The only choice variables that face the consumer are X

1 and X

2. The consumption of these goods

generates their own waste, which in turn impacts the utility function by way of S and e. The first order conditions are:

U1 + Us

1*r1 - (Π

1 + hr

1) λ = 0 (2)

U

2 + Us

2*r2 - (Π

2 + hr

2) λ = 0 (3)

wT +Y - Π

1 X

1 - Π

2 X

2 - h(r

1 X

1 +r

2 X

2 ) = 0 (4)

Where U1 and U

2 are both > 0 and U

s is < 0.

The reason that U

s is < 0 is that solid waste must be stored on

the premises until it is disposed of. This is a nuisance for the consumer. In addition, it is assumed that consumers care about the impact that their waste (as well as all waste generated by society) has on their own health and the

environment. This implies that Use

is < 0 (i.e., the marginal disutility of solid waste increases with environmental awareness).

xv

Solving equation 6 – 8 yields the demand function for X

i :

X

i = X*

i (p

1, p

2, w, t

1m, t

1s, t2m, t2s , r

1, r

2, h, Y, e ) (5)

Where: X

1 = quantity of disposable goods sold (purchased)

X

2 = quantity of reusable goods sold (purchased)

p

1 = price of disposable good

p

2 = price of reusable good

w = consumer wage rate ti m

= time value of consuming and maintaining a disposable good and/or a reusable good ti s

= time value of searching for and purchasing a disposable/reusable product r

i = proportion of a disposable good and/or a reusable good

that ends up as solid waste h = cost of solid waste removal y = non labor income e = environmental awareness

The comparative statics of ϑX u1/ ϑw derived from the

expenditure minimization model yield pure substitution effects that are different from the standard model. This results from the fact that a change in w impacts both X

1 and

X2 at the same time.

The substitution effects for X

1 is assumed to be positive, thus

this would lead to the following result:

ϑXu1/ ϑw = -(t1m + t1s)D11 * -(t2m + t2s)D21 > 0

xvi D (6)

Where D11

and D21

are the first two cofactors of the first column of the bordered Hessian matrix and D is its Determinant.


89

ϑXm1/ ϑI derived from the Utility maximization model is

assumed positive for normal goods.

Finally, ϑXm1/ ϑw can be broken down into a pure

substitution effect and an income effect by way of the Slutsky equation.

ϑXm1/ ϑw = ϑXu - ( (t

1m + t

1s )X

1 + (t

2m + t

2s )X

2 - T)

ϑXm ϑw ϑI (7)

The assumed inequality of ϑXm1/ ϑw > ϑXm

2/ ϑw indicates

that consumers shift toward disposable goods as wages rise. If over the life of a disposable good versus its reusable counterpart, the direct cost plus the search cost of the disposable good is greater than or equal to the direct cost of the reusable counterpart, then the driving force behind the shift in purchases to disposable goods is the fact that consumers are trying to economize on the maintenance costs. Conversely, if over the life of both a disposable good versus its reusable counterpart, the direct cost plus the search cost of the reusable counterpart costs more, consumers would not purchase the reusable good since, by definition, the maintenance cost is greater for the reusable item as well and therefore, the total cost of a reusable good would be greater. The change in solid waste with respect to X

1 is simply:

ϑS/ϑX1 = r

1

Therefore:

ϑS/ϑX1* ϑXm

1/ϑw = r

1* {ϑXu

1/ ϑw - ( (t

1m + t

1s )X

1+

(t2m +

t2s

)X2 - T) ϑXm

1/ϑI} (8)

Whereas: r

1 = (a

1)*(b

1)*(c

1)

This formula shows how the composition and amount of solid waste is impacted by a change in w. As w increases, there is a pure substitution away from reusable goods to more disposable goods. In addition, since w has increased, consumers will purchase more of both goods (the income effect), but assuming the income elasticity is higher for disposable goods, relatively more of these goods will be purchased through the income effect. Therefore r

1X

1

increases by more than r2X

2 does .

However, this is not the end of the story. Various other variables impact S as well, such as: the price to haul away solid waste (which is represented in the model as h), the level

of waste collection services provided by the municipality, length of time that recycling programs have been operating, pricing structure of waste disposal as set by the municipality, and the environmental awareness of the consumer. These factors are other explanatory variables which help to determine the overall level of solid waste generation As environmental awareness increases, the consumer feels “guilty” about the purchase of disposable goods and purchases reusable counterparts instead. In this respect, the higher the level of environmental awareness, the more a disposable good acts like a “bad”, reducing marginal utility. Thus U

se < 0. This relationship will be referred to as a

“greening” effect. Which means that the greater the level of ecological awareness (which is assumed to be a function of education and not income), the more likely it is that the consumer will change the optimal mix of disposable and reusable goods so that the mix will be weighted more heavily with reusable goods.

xvii

Therefore, the overall change in municipal solid waste, overtime, is assumed to result from two factors. The first is from a change in w. The second is from a change in environmental awareness, which is an indirect effect, or a “greening” effect. This paper hypothesizes that in the long run the direct impact from a wage increase will outweigh the indirect impact (or “greening” effect).

CONCLUSIONS The model developed in this paper shows how gross residential municipal solid waste changes with a change in income as consumers attempt to maximize their utility by economizing on their indirect costs (i.e., the time variable). More specifically, the model proposes that as incomes rise, consumers will purchase more of both reusable goods and disposable goods. In addition, as incomes rise consumers will naturally substitute their purchases out of and away from reusable goods and into disposable goods. The reason for the shift towards disposable goods as incomes rise is that it becomes too costly for consumers to spend their time repairing and maintaining products. Their time is better spent in more productive endeavors. It is simply cheaper (in terms of opportunity cost of time) to dispose of old products and replace them with new products. Disposing of products, of course, creates solid waste. It is not evident that repairing and maintaining products for reuse actually has an impact on residential MSW. These reasons are why we see residential MSW per capita continue to grow as incomes increase. If the above argument is true, then it demonstrates that the Environmental Kuznets Curve cannot possibly exist for the MSW environmental indicator.


90


91

As stated previously, other factors, such as price of waste removal and level of service will impact solid waste as well. But these are policy issues to be put into place by local governments and are outside the control of the consumer. Comparative statics can be derived for these variables as well. These comparative statics would closely parallel the comparative statics derived in previous models. In summary, it is not necessarily income growth that causes the growth of residential MSW, instead it is the way in which consumers spend their incomes. To make this point clear, assume hypothetically that consumers saved their entire incomes. In this case, no waste would be generated. Alternatively, they could use their incomes to purchase only reusable goods, thus minimizing their impact on residential MSW. However, they won’t because as their incomes grow it is not worth their time and effort to maintain and repair reusable goods. This is true even if the explicit cost of a reusable good is cheaper than the disposable good (on a per use basis). Instead, consumers will seek out and purchase disposable goods in a never ending quest to economize on their time.

ENDNOTES 1 See Cooper (2005), Strasser (1999), and Fernandez (2001), Thomas (2003) 2 See Wertz (1976), Petrovic, W. M. and Jaffe, B.L. (1978), Richardson, R.A., and Havlicek, J. (1978), and Kemper and Quigley (1976) 3 It is assumed that once a reusable or disposable good has been fully used it then enters into the waste stream and does not remain on household premises. 4 Gross residential MSW is the amount of waste discarded by a household before any products are recycled or sent to thrift stores. Any discarded waste leftover after both recycling and/or having the waste sent to a thrift store is considered “net residential MSW”. Net residential MSW is the amount of solid waste that ends up in the landfills or at incineration plants. 5 See McCollough (2007) for empirical evidence of the relationship between income growth and disposable goods 6 For instance, see King, Burgess, Ijomah, and McMahon (2005) 7 Although not the focus of this paper, it should be noted that at times disposing of a product, such as an appliance, can be of more benefit to the environment than reusing the product. This, of course, depends on how much more energy efficient the new product is. 8 Municipal solid waste is typically measured by weight.

9 Obviously, disposable goods and reusable goods will not always have the same weight. Sometimes a reusable might weigh more than a disposable good since in, many cases, it is made to be more durable. 10 Even products such as a paint brush that are designed to be disposable can become reusable with some additional maintenance 11 Still the impact can be calculated. The viewing of a movie at the local theater has a (b

1) factor of 0, a (c

1) factor of 0,

and an (a1) factor of 1. The impact on solid waste from the

purchase of a movie is (a1)*(b

1)*(c

1)X

1= r

1X

1= 0 lbs of X

1.

The “consumption” of a movie does actually generate solid waste. Ignoring concession sales, the waste generated by viewing a movie would come from the disposal of the ticket for the movie. Another example is that a meal at McDonald’s has a (b

1) factor of .18 (packaging content only, the food is

consumed), a (c1) factor of .75 lbs, and an (a

1) factor of 1.

The impact on solid waste from the purchase of a meal at McDonald’s is (a

1)*(b

1)*(c

1)X

1= r

1X

1= .135 lbs of X

1.

12 Typically, consumers will pay others, such as mechanics and repairmen, to maintain and repair reusable products for them if they find that the repairman or mechanic can provide this service at a cost that is lower than the consumer’s indirect cost. 13 Beatty and Smith (1987) empirically test the above assumptions and finds that search costs is negatively related to the frequency of purchase and positively related to price and/or complexity of a product. 14 This, of course, depends on the frequency with which the product is purchased as well as the price and complexity of the product being purchased. 15 In this model, the consumer feels guilty about his purchases of disposable goods and the impact disposable goods have on the environment. Also, the consumer feels better about the environment when he purchases reusable goods. This assumes away the free rider dilemma associated with an individual’s contribution to the buildup of municipal solid waste versus the entire population’s contribution to the buildup of municipal solid waste. 16 Since they are compensated changes, it must be that ϑXu

2/ ϑw < 0 17 See Saltzman, C., Duggal, V., and Williams, M. (1993) for evidence that recycling is negatively related to income. Also, see Bayus (1991) for evidence that early replacers of automobiles have higher incomes but less educational attainment than late replacers of automobiles. References available upon request from John McCollough.

DISCUSSANT COMMENTS WAGE GROWTH AND EMPLOYMENT GROWTH

Lynn A. Smith

Department of Economics Clarion University of Pennsylvania

Clarion, Pennsylvania 16214

This study makes a significant contribution to our understanding of the link between earnings growth and employment growth in regional labor markets. It uses a data collection approach that could be compared to the “survey method” employed decades ago to input-output analysis by William Miernyk of West Virginia University. In the absence of Miernyk’s “survey method” coefficients in regional input-output models were derived from the national model. Likewise, Professor Tannery argues that disaggregated data gives us a better setting than aggregate data to test the central hypothesis from this study - earnings growth in a particular sector will lead to lagged employment growth in that sector. According to the hypothesis developed by the author, earnings growth can be a predictor of employment growth in a regional economy or a local economy. Growth in earnings will not lead to employment growth initially because of long run cost-benefit analysis carried out by employers. Because employers want to be certain that the recovery is under way, they first extend the hours of the current workforce rather than employ more workers. This behavior of employers explains in part why unemployment is a lagged economic indicator. According to Professor Tannery, it is important to use disaggregated data in this study given that Pennsylvania is a diverse state with urban counties in the southeastern part of the state, and rural counties in most other regions.

An important contribution is made by this study in that, “[the] results indicate that earnings growth can be used as a predictor of within-industry employment changes at the county level.” Also, the author correctly notes that, “The use of the Quarterly Census of Employment and Wages to predict future employment gains may be much stronger than we are able to estimate with aggregate data.” The following comments are made with regard to style and format of this paper. These comments, if adopted, may assist the reader in understanding the author’s arguments. First, adding a section titled, “The Empirical Model” would be helpful. In the text of this section the author should include each regression equation employed in the study, and the explanatory variables and dependent variables for each equation should be identified. Second, the level of significance should be selected a priori in The Empirical Model section. While it is acceptable to add footnotes in the tables that report the results to indicate “significant at the one percent level” and “significant at the five per cent level” it s important to select levels of significance a priori. Of course, these levels of significance will be selected in large part based on the probable occurrences of Type I and Type II errors, and the trade-off between making these errors. Also, it may be beneficial for the author to consider adding a brief comment that the link between earnings growth and employment growth as presented in the paper is in no way inconsistent with Marshallian demand analysis where there is an inverse relationship between price and quantity demanded.


92

OPTIMUM CURRENCY AREAS AND SYNCHRONIZATION OF BUSINESS CYCLES IN SUB-SAHARAN AFRICA

Yaya Sissoko

Department of Economics Indiana University of Pennsylvania

Indiana, PA 15705

ABSTRACT This paper investigates the Theory of Optimum Currency Areas in Sub-Saharan Africa (SSA). This issue is examined in a context of small open economies of SSA (CFA and non-CFA countries) using a structural Variance Auto Regression (VAR) approach with limited capital mobility and a weak-banking system in Africa. The finding suggests similar terms of trade and trade balance disturbances in the CFA and non-CFA countries in contrast to the supply and demand shocks which tend to influence the non-CFA zone to a greater extent. The sizes of the disturbances and the speed of adjustment confirm that the CFA and non-CFA countries are suitable of forming a monetary union. In addition, these results suggest the creation of smaller monetary arrangements in the CFA and non-CFA regions as preliminary steps in creating one monetary union in Africa. However, there is weaker business cycle synchronization in Sub-Saharan Africa.

INTRODUCTION The idea for monetary union has been around for a while. The globalization and the internationalization of the world economy push countries to get together and create a monetary unification. But, there are costs and benefits of forming a monetary union. Mundell (1961, 1968), rightly considered as the father of the Optimal Currency Areas theory, discusses the conditions for the realization of a monetary unification with a single currency. To achieve an Optimum Currency Area (OCA), countries should be economically integrated and have some experiences of flexible exchange rate regime. Factors of production should move freely within the area with stable relative prices. The size of the economy is a key factor. The theory of OCA also requires a single currency with a single central bank without losing reserves and impairing convertibility. This means national central banks have to give up their sovereignty over their own currency. Furthermore, Mundell (1999) underlines seven criteria to realize a monetary union. First, there should be a large transactions area in order to have a low, flat transaction cost. Second, monetary policy should be stable. An unstable monetary policy could result in an unstable real money balances. Third, there should be no controls by the governments of the monetary union. Of course, this is a difficult task given the political situation in each country

member of the union. The fourth feature is the need of a strong central state to avoid the collapse of its currency when the country is invaded. The fifth factor is a central bank committed to stabilize the prices. To achieve this criterion, the central bank has to hold substantial gold reserves and foreign exchange reserves. The sixth feature is a sense of permanence. There should be a belief that the monetary union and the currency are going to be here forever, not just for a limited period of time. Finally, the OCA should provide low interest rates for the single currency to dominate and prevail as a currency bloc. In the end, there should be a Euro bloc, which has already started in the European Monetary Union1. McKinnon (1963) suggests that the achievement of an OCA depends mostly on the openness of the economy and the achievement of conflicting economic objectives such as full employment, balanced international payments and stable internal average price level. Moreover, factor mobility is an important feature in the theory of OCA. Furthermore, Kenen (1969) finds that one criterion of an OCA is the degree diversification of the economy of the members. The desirability of fixing the exchange rate increases with the level of diversification of the economy. A more diversified production base might imply perhaps higher net benefits from fixed exchange rate. Bayoumi and Eichengreen (1992, 1994) empirically investigate the Optimum Currency Areas (OCA) theory in light of a monetary union in Western Europe and the United States. They divide the United States divided into seven geographical regions namely, New England, Mid-East, Great Lakes, Southeast, Plains, Far West and West. They also examine the feasibility of a monetary unification in Western Europe, East Asia and the Americas including the United States and Canada. Bayoumi and Eichengreen use a structural VAR to identify the incidence of aggregate supply and demand shocks. Following Mundell (1961), a monetary unification should lower transaction costs and eliminate the variability of exchange rates but with the loss of monetary policy independence. Countries experiencing symmetric or similar disturbances of aggregate supply and aggregate demand would form a monetary union. Of course, the size of the disturbances and the speed of adjustment will matter. They find that there should be a Northern European Bloc, a


93

Northeast Asian Bloc and Southeast Asian Bloc. The Americas region is less plausible as candidate for a monetary unification but the United States, Canada and possibly Mexico may get together to form the North American Bloc. Moreover, Horváth and Grabowski (1997) find that asymmetric supply disturbances across African countries and symmetric demand shocks across African regions during the 1960-1992 period. These regions include Northern, Western, Eastern and Southern Africa. Their findings make the African Continent a less plausible candidate for monetary union but monetary arrangements might be possible at a smaller scale. Bayoumi and Eichengreen (1997) suggest that economic integration and monetary integration go together. They apply the OCA theory to the European countries by computing an OCA index. The OCA index is driven by the relative size of the country and the degree of economic integration. Eichengreen (1997) argues that Europe is an OCA since European countries experience region-specific shocks and a higher variability of real exchange rates than the United States and Canada. Furthermore, Bayoumi and Eichengreen (1998) show that the patterns of exchange rate variability and intervention across countries using the theory of OCA. Countries experiencing larger asymmetric shocks are countries with more flexible exchange rates. Bayoumi, Eichengreen and Mauro (1999) investigate the feasibility of monetary arrangements for ASEAN2. They identify gradual steps such as standard economic criteria, higher level of intra-regional trade and firm political commitment for the ASEAN to achieve a monetary unification. More importantly, Frankel and Rose (1998) outline four criteria of forming a potential OCA. The first factor is the extent of trade, that is, the trading intensity with other potential members of the currency union. The degree of openness of potential members of an OCA depends of their economic integration, which leads to low transaction costs and risks associated with different currencies. The second criterion is the similarity of the shocks and the cycles. Countries might experience closer international trade linkages when they do a lot of intra-industry trades. This, in turn, leads to similar business cycles. The degree of mobility of labor is the third criterion. Factors of production should move freely between regions according to their marginal productivity. The last feature is the system of risk sharing with respect to fiscal transfers. This is known as the balanced budget or fairness criterion. Countries with a huge budget deficit should not transfer their burdens to countries with a balanced budget or small budget deficit. Frankel and Rose focus on the former two criteria known as the “Lucas Critique”. They consider that international trade and international business cycles are endogenously correlated. That is, trade integration leads to more international trade, which in turn will result in high correlations of business

cycles across countries. Countries with the same or close international trading partners will benefit from a monetary unification. The benefits of joining a monetary union have to be higher than the associated costs, namely transaction costs and loss of monetary independence costs. Sub-Saharan Africa (SSA) already has two monetary unifications. The first one consists of 7 countries in West Africa and 6 countries in Central Africa plus the Islamic Federal Republic of Comoros. The single currency used is the CFA Franc (CFAF)3. The Comoros Franc (CF) is the currency in Comoros Island. The CFA Franc Zone was created in 1946 after the Second World War4. The CFA Franc was originally pegged to the French Franc. The French Treasury provides foreign exchange reserves to the CFA Franc Zone and maintains a freely convertibility vis-à-vis the French Franc (FF). However, the CFA Franc is now pegged since 1999 to the Euro, the currency of the European Union5. The second currency union in SSA is the Common Monetary Area (CMA) in Southern African, which includes South Africa Republic, Lesotho, Namibia and Swaziland. Bank notes issued by these countries are freely convertible into the South African Rand. Bayoumi and Ostry (1997) apply the theory of OCA to SSA countries by investigating the size and correlation of the real disturbances across countries and the level of intra-regional trade. According to the theory of OCA, the benefits of forming a monetary union are lower transaction costs and the elimination of the exchange rate variability while the cost is the loss of monetary sovereignty. Assuming the same speed of adjustment, if countries face similar or symmetric disturbances then they will gain from forming a monetary union. Of course, the benefits depend upon the degree of diversification of their export commodity base. That is, the desirability of monetary unification should decrease with the degree of specialization of production. Other factors such as the poor quality of the data, the complement of the production structure, the poor local and intra-regional transportation and communications networks might explain the asymmetric of African trade disturbances. A monetary union requires that the participating countries give up their sovereignty over the national currency and monetary policy. A unified monetary system implies some costs and benefits shared by the member countries. The main objective of this paper is to examine whether SSA countries should form one or more monetary arrangements in the light of the theory of an OCA. Another objective is the synchronization of business cycles in SSA countries. The economic integration of SSA countries will tend to raise the inter-linkages of their business cycles. The remaining of this paper is as followed. The first section discusses the introduction and review literature. The second


94

section contains the specific objectives and the data. The methodology, model identification and unit root test results are mentioned in the third section. The fourth section reports the empirical results and the last section concludes and draws the policy implications of the study.

MODEL AND METHODOLOGY The analysis of the theory of an OCA is used to determine potential members of a monetary union. Potential members of an OCA tend to have synchronized business cycles (Bayoumi and Eichengreen, 1993 and 1994). Roughly speaking, the partners of synchronization depend upon the mechanism of transmission via common shocks or a weaker form of transmission. A structural Variance Auto Regression (VAR) methodology is used first to assess the possibility of one or more OCAs in SSA countries. This methodology follows Blanchard and Quah (1989) and Bayoumi and Eichengreen (1993, 1994). Consider a small open economy with limited capital mobility. Political instability and weak financial/institutional infrastructure in SSA suggest that it is inappropriate to assume uncovered interest parity. In what follows, we adapt a small open economy aggregate supply/aggregate demand (AS/AD) model to reflect exogenous capital mobility that may be more appropriate for SSA. The following equations provide the elements of such a model that will provide the restrictions to identify the shocks within a structural VAR framework (please refer to Equations 1-11 in the Appendix).

Where = terms of trade as proxied by the relative price of the primary export commodity,

hy =real GDP,

( y = capacity output, = nominal interest rate, s = nominal exchange rate (e.g., CFA Franc per dollar), = domestic price level, m = money stock, d = autonomous aggregate demand, all variables except the interest rate are in logarithms,

ip

Et is the conditional expectations parameter, and all Greek parameters are positive. The observed movements in the variables are due to five mutually uncorrelated “structural” shocks with finite variances. These are terms of trade shocks εt

h , aggregate

supply shocks, ε ts , trade balance shocks, ε t

z , aggregate

demand or real demand shocks, εtd , and money supply

shocks,ε tm .

Equation 1 is the evolution of the world oil or export commodity price, which is assumed to be exogenous. Equation 2 is an aggregate supply equation, where Aggregate Supply depends on capacity output and terms of trade (world

oil price or export commodity price). Capacity output in equation 3 is a function of the productive capacity of the economy (e.g., capital stock and human capital or employment), and for simplicity, is assumed to be a random walk process. The balance on goods and services (equation 4) is assumed to be a function of the real exchange rate, (st − pt) and domestic real income. For simplicity, we normalize the foreign price level to unity so that (st − pt) measures the relative price of foreign goods in terms of domestic goods. Although we label a trade balance shock, it can capture capital flows shocks, or exogenous shifts in imports or exports. Equation 5 implies that the exogenous part of the trade balance shocks follow a random walk.

zt

Equation 6 is a conventional aggregate demand (IS) equation where aggregate spending depends on the expected real interest rate and the exogenously given level of the trade balance. The autonomous portion of aggregate demand, d , is assumed to follow a random walk in equation 7. Equation 8 is a conventional money demand equation with unitary income elasticity. Equation 9 is the evolution of money supply, which for simplicity, is assumed to follow a random walk. Finally we close the model by postulation of goods and money market equilibrium relationships (equations 10 and 11).

t

In order to solve the model, we eliminate the interest rate from equation (6) using equation (8) to get:

pt = (λγ

1 + λγ)Etpt +1(

λ1 + λγ

)(dt − zt) + (1

1+ λγ)mt − (

1+ λ1 + λγ

)yt (12)

This is a first order expectational difference equation in the price level. Note that for finite values of the parameters, and assuming that λγ ≠ 1, the forward-looking solution is convergent. With rational expectations, and given the stochastic processes for the exogenous variables in equations 1, 3, 5, 7, and 9, the forward looking solution for the price level is given by

pt = mt + λ(dt − zt ) − (1 + λ )yt (13)

From equation (13), Equilibrium real money balances are

The equilibrium real exchange rate, which is compatible with trade balance, is obtained using (4)

st − pt =η2

η1

yt −1

η1

zt (15)

mt − pt = λ(zt − dt ) + (1 + λ )yt (14)


It can be shown that the long run impact of the structural shocks on the endogenous variables has a peculiar triangular structure6. In order to show the long run impact of the five structural shocks ε t = [ε t

h,ε ts ,εt

z ,ε td ,ε t

m ] on the system of endogenous variables we express the solution to the model in first differences:

'[ , , ( )( ), ] ,t t t t t t t th y s p m p pΧ = − −

Δht = ε t

h (16)

Δyt = θε th + ε t

s (17) Δ ( s t − p t ) = (

η 2

η 1

)( θ th + ε t

s ) − (1

η 1

) ε tz (18)

Δ(mt − pt) = λ(ε tz −ε t

d ) + (1+ λ)(θε th +ε t

s) (19) Δpt = λ (ε t

d − ε tz ) − (1 + λ )(θε t

h + ε ts ) + ε t

m (20)

Note that although endogenous variables have unit roots, all are difference stationary. The long run impact of the structural shocks on the endogenous variables is triangular. Specifically, all shocks except terms of trade shocks have no long effect on the oil price or the relative price of primary commodity. Real demand, trade balance, and monetary shocks have no long run impact on output. Real demand and monetary shocks have no long run impact on the real exchange rate, and monetary shocks have no long run effect on real money balances. Given the model structure above, the long run effects of the shocks of the endogenous variables are given by:

(21)

More importantly, SSA countries seem to lag behind other regions of the World in term of output growth performance even though they might have an edge in the area of inflation. Bayoumi and Eichengreen (1994) find that the average annual growth rates of output and inflation are respectively 3.3% and 7.2% for Western European Countries during the 1960-1990 period against 6.0% and 8.4% for East Asia and 3.1% and 4.9% for the Americas including the United States and Canada. The output and inflation variability is somewhat higher across SSA countries than across the countries of the regions of the world considered above10.

where aij(1) represents the cumulative long-run effect of shock j on variable i. The zero entries in equation (21) provide the 10 (long run) restrictions needed to identify the shocks.

EMPIRICAL RESULTS

The study covers 30 SSA countries from both the CFA Franc and the non-CFA Franc zones. The CFA countries covered in the study include Benin, Burkina Faso, Cameroon, Central

African Republic, Chad, Congo, Côte D’Ivoire, Gabon, Mali, Niger, Senegal and Togo. The non-FCFA countries are Botswana, Burundi, Ethiopia, Gambia, Ghana, Kenya, Lesotho, Madagascar, Malawi, Mauritius, Nigeria, Rwanda, South African Republic, Swaziland, Tanzania, Uganda, Zambia and Zimbabwe7. The data used in the sample consists of 40 annual observations from 1960 to 2000 taken from the International Financial Statistics (IFS) CD-ROM published by the International Monetary Fund (IMF).The data set includes the following series: terms of trade, output, real exchange rates, real money balances and price level8.

The proper specification of the VAR requires testing for times series properties of the data. The variables are tested for unit roots using the Augmented Dickey-Fuller (ADF) and Kwiathowski-Phillips-Schmitt-Shin (KPSS) test statistics9. The ADF test statistics show the null hypothesis of a unit root cannot be rejected at the log level for the data for most of the series in question at 5 percent significance level. The ADF test statistics also indicate that the variables are stationary in the first differences at the 5 percent significance level. This makes the use of a VAR appropriate. Moreover, the KPSS test statistics confirm the results of the ADF test. That is, the acceptance of the null hypothesis of the KPSS test makes the use of a VAR in first differences appropriate.

Table 1 shows the mean of the annual average growth rate of output and inflation across the CFA and non-CFA countries for the full period of the data. The non-CFA countries have grown on average faster than the CFA countries during the period of 1960 to 2000. The ratio is 1 to 2 in favor of the non-CFA countries with Botswana leading with a growth rate of output of 9.1% in the non-CFA zone compared to only 6.0% for the Republic of Congo in the CFA zone. However, the CFA countries outperform the non-CFA countries with an annual average inflation rate of 6.7% against 13.1% for the latter. The inflation variability is somewhat smaller across the CFA countries than across the non-CFA countries.

( )( )

( )( ) ( ) ( )( ) ( ) ( ) ( )( ) ( ) ( ) ( ) ( ) m

t

dt

zt

st

ht

t

tt

tt

t

t

aaaaaaaaa

aaaaa

a

ppmps

yh

εεεεε

11111011110011100010000

5554535251

44434241

333231

2221

11

=

Δ−Δ−Δ

ΔΔ

,


Correlation of Supply and Demand Disturbances Table 2 is the country codes used for the CFA and non-CFA countries. Table 3 displays the results of the correlations of supply shocks above the diagonal and the correlations of demand disturbances below the diagonal. The supply correlation coefficients within the CFA countries and between the CFA and the Non-CFA countries do not suggest a clear regional pattern among these countries. Indeed, most of the supply correlations are insignificant within the CFA countries. The Non-CFA countries have more significant positive supply correlations than the CFA countries but still do not show any clear geographical pattern. The correlations of supply shocks between Gambia, Ghana, Malawi, Nigeria and Uganda are positive and significant suggesting a regional pattern. One possible explanation is perhaps the differences in the economic structures within the CFA and Non-CFA countries. These results are similar with the findings of Bayoumi and Eichengreen (1994) that the supply correlations within Europe and the Americas do not feature a clear geographic pattern in contrast to the ones of Asia. The demand disturbances exhibit significant positive coefficients within the CFA countries and the Non-CFA countries with a clear geographic pattern. However, the correlations between the CFA and the Non-CFA countries show a number of significant positive coefficients but with no clear regional pattern. The CFA countries except Congo and Côte D’Ivoire might be good candidates for monetary union. The correlations of demand shocks between Burundi, Ghana, Malawi, Mauritius and Mali also suggest a coherent regional pattern. In the non-CFA countries, the best clear geographical pattern consists of Swaziland, Botswana, Ethiopia, Ghana, Madagascar, Malawi, Mauritius, Nigeria, Rwanda and South Africa. The demand disturbances are highly correlated within the CFA countries. These results are also similar with the findings of Bayoumi and Eichengreen (1994) in their study about the prospects of monetary unification around the world.

Size of Disturbances and Speeds of Adjustment and Synchronization of Business Cycles Besides the level of correlation between countries, the size of the shocks and the speeds of adjustment are also important in defining a monetary unification. Bayoumi and Eichengreen (1994) identify three criteria to define an Optimum Currency Area. Related to the country’s macroeconomic disturbances, these criteria are: the size of shocks, the cross-country correlation and the speed of adjustment. Countries relatively highly correlated and with similarly sized shocks and speed of adjustment are suitable to form a monetary union. The results of the sizes of the disturbances and the speeds of adjustments are given in table 4.

The size of the shock is measured by the standard deviation of each disturbance. Larger sized disturbances are costly to the economy to offset the shocks. On average, the Non-CFA countries experience a larger terms of trade shock than the CFA countries. CFA and Non-CFA countries face similar sized supply shocks on average with the largest disturbance in Chad for the CFA countries and Nigeria and Burundi for the Non-CFA countries. Moreover, the CFA and Non-CFA also display similarly sized trade balances disturbances on average. Burkina Faso in the CFA zone and Burundi in the non-CFA zone are the two countries with the largest trade balance shocks. The size of the monetary shocks is smaller on average in the CFA countries than in the non-CFA countries. As one should expect, this is the discipline effect of the fixed exchange rate regime. However, the CFA countries display on average larger demand shocks than the non-CFA countries with Central African Republic in the CFA zone and Uganda in the non-CFA experiencing the largest demand disturbances. Overall, even though the CFA and non-CFA countries face similarly sized disturbances, the trade balance shocks are far the largest shocks on average for both the CFA and non-CFA countries. Indeed, trade balance disturbances represent on average twice the terms of trade shock size, five times the size of supply shocks, three times the size of monetary shocks and four to six times the size of demand shocks.

A simple measure of the speed of adjustment is the ratio of the impulse responses function in a chosen year, say the third year divided by its long run level11. A low value of the speed of adjustment indicates a relatively slow adjustment while a high value indicates a large amount of adjustment. Note that there are high costs to the economy associated with a relatively slow adjustment. The non-CFA countries have on average faster adjustment speed in terms of trade, supply, and trade balance, monetary and demand shocks than the CFA countries. Indeed, only one third of the terms of trade adjustment occurs within three years while the adjustment of terms of trade shocks is two thirds in the non-CFA countries. The fastest adjustment in terms of trade happens in Togo for the CFA countries and Burundi for the non-CFA countries where all the adjustments occur within three years. Three fourths of the adjustment of supply shocks occurs on average within three years in the non-CFA countries. In the CFA countries, in contrast, the change or adjustment is only two thirds. Cameroon and Botswana in the CFA and non-CFA countries respectively achieve the fastest supply shock adjustment. The adjustment speed in trade balance disturbances within three years is respectively half and two thirds for the CFA and non-CFA countries. The non-CFA countries achieve two thirds of the adjustment in monetary shocks within three years while the change is only half in the CFA countries. Finally, 75% of the adjustments of demand disturbances occur on average within three years in the non-CFA countries. In contrast, the change is only half in the


97


98

CFA countries. Gabon and Uganda in the CFA and non-CFA countries respectively achieve all the adjustment of demand shocks within three years faster than anyone else in their respective bloc. The non-CFA countries face on average faster adjustment speed of real shocks within three years than the CFA countries. In the CFA countries, only one third of the adjustment of real disturbances occurs within three years in contrast to the non-CFA where the adjustment is one half. In contrast, the adjustment speed of nominal disturbances is faster on average in the CFA countries than in the non-CFA countries. Indeed, three fourths of the nominal adjustment speed occurs within three years in the CFA countries while the adjustment is only one half for the non-CFA countries. Bayoumi and Eichengreen (1994) investigate the speed of adjustment in West Europe, East Asia and the Americas including Canada and the United States. Within two years, they find that Asia has the fastest adjustment of all the change of output and prices, followed by the Americas and Europe where only 80% and 50% of the adjustment is completed respectively. Overall, the CFA and non-CFA countries experience similarly sized disturbances and almost same speed of adjustment for the different shocks. However, there is no clear geographical pattern for the correlations between CFA and non-CFA countries. Nevertheless, this may be an indication of a partial synchronization of business cycles in SSA. Countries facing the similar shock sizes with same speed of adjustment of the disturbances might get together to form a monetary union with less opportunity costs.

CONCLUDING REMARKS The Theory of Optimum Currency Area (OCA) is applied to SSA countries to indentify feasible monetary arrangements. This study focuses on the correlations of aggregate supply and demand disturbances, the sizes of the disturbances and the speed of adjustment as the necessary conditions of forming a monetary union. Countries with high disturbance correlations, with same shock sizes and same speed of adjustment may be a strong evidence for currency unification. The results of the supply disturbances do not show a strong evidence of common currency area in the CFA and non-CFA countries. The correlations of supply shocks between the CFA and non-CFA countries do suggest a clear regional pattern among these countries. The results, however, favor smaller SSA blocs such as the one between South African Republic, Cameroon, Cote D’Ivoire and Niger. In contrast, the correlations of demand shocks feature significantly positive coefficients among the CFA countries, the non-CFA countries and between the CFA and non-CFA countries. These results suggest a clear geographical pattern within the CFA and non-CFA countries or between the CFA

and non-CFA countries. The results of the supply and demand disturbances are very similar to the findings of Horvath and Grabowski (1997) about African regions. Notwithstanding their economic structure disparity, CFA and non-CFA countries experience on average similarly sized disturbances. These results hold for the different shocks considered in the study, namely terms of trade, supply, demand, monetary and trade balances disturbances and real and nominal shocks. CFA and non-CFA countries also feature on average similar speed of adjustment within three years after experiencing macroeconomic disturbances. These results may suggest a possible partial synchronization of business cycles in Sub-Saharan African countries. The management of the exchange rate policy and monetary policy will be much easier even though there is loss of monetary sovereignty. Further studies may investigate the level of intra-regional trade within the CFA and non-CFA countries and between the CFA and the non-CFA countries. It will be interesting to check the findings of intra-regional trade disturbances in SSA countries in the light of the theory of OCA.

ENDNOTES

1The European Monetary Union, created by the Maastricht treaty (1993), includes currently 27 members. 2ASEAN consists of ten countries, namely Brunei Darussalam, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Thailand and Vietnam. 3The CFA stands for the “Communauté Financière Africaine” in the West African Economic and Monetary Union (WAEMU) and for the “Coopération Financière en Afrique Centrale” in the Central African Economic and Monetary Union (CAEMC). 4See Clement et al. (1996). 5The exchange rate is 1 French Franc (FF) for 50 CFAF before the 1994 devaluation and 100 CFAF after and 1 Euro for 655.957 CFAF. 6See Weber (1997) and Wehinger (2000) for similar Aggregate demand/Aggregate Supply triangular long run impact structures. 7This sample covers the bulk of SSA except for the Comoros Islands in the CFA group and Angola, Equatorial Guinea, Guinée, Guinea Bissau, Liberia, Mozambique, Sierra Leone, Somalia, Sudan, and Democratic Republic of Congo (i.e. former Zaire) in the non-CFA group. These dropped from the study because of insufficient data. 8See Sissoko and Dibooglu (2006) for a detailed explanation about the construction of the data. 9These results are available from the author upon request. 10See Bayoumi and Eichengreen (1992) for similar comparative findings between the European countries and the U.S. regions. 11The choice of the third year is somewhat arbitrary but the

APPENDIX

Equations 1 – 11:

httt hh ε+= −1 Terms of trade (1)

ttst hyy θ+= (

Aggregate Supply (2)

( y = ( y t −1 + εt

s Evolution of Capacity output (3)

ttttt zypsnx +−−= 21 )( ηη = 0 Trade balance (4)

zt = zt −1 + εt Trade balance shock (5)

ytd = dt − γ [it − Et (pt +1 − pt )]− zt Aggregate Demand (AD)/IS (6)

dt = dt −1 + εtd Evolution of autonomous AD (7)

mtd = pt + yt − λit Money demand (8)

mts = mt −1

s + ε tm Money supply (9)

yts = yt

d = yt Goods market equilibrium (10)

mts = mt

d = mt Money market equilibrium (11)


99

Table 1: Annual Average of Output and Inflation Growth Rates -1960 to 2000

GROWTH INFLATION Mean Std. Deviation Mean Std. Deviation CFA COUNTRIES Benin 0.031 0.047 0.108 0.098 Burkina Faso 0.018 0.077 0.048 0.073 Cameroon 0.019 0.055 0.071 0.068 C.A.R* 0.016 0.106 0.058 0.068 Chad 0.046 0.141 0.053 0.067 Congo 0.030 0.589 0.111 0.102 Cote D’Ivoire 0.027 0.084 0.065 0.062 Gabon -0.014 0.214 0.063 0.080 Mali 0.022 0.096 0.101 0.091 Niger -0.005 0.260 0.059 0.091 Senegal 0.016 0.058 0.086 0.108 Togo -0.015 0.096 0.048 0.135 Average 0.015 0.140 0.067 0.080 NON-CFA COUNTRIES Botswana 0.091 0.069 0.121 0.072 Burundi -0.006 0.238 0.114 0.093 Ethiopia 0.028 0.044 0.062 0.077 Gambia 0.020 0.256 0.090 0.103 Ghana 0.037 0.083 0.240 0.216 Kenya 0.060 0.157 0.096 0.080 Lesotho 0.028 0.085 0.129 0.053 Madagascar 0.017 0.035 0.126 0.092 Malawi 0.060 0.084 0.207 0.256 Mauritius 0.044 0.070 0.100 0.098 Nigeria 0.038 0.302 0.157 0.147 Rwanda 0.026 0.134 0.106 0.102 South Africa 0.030 0.027 0.087 0.046 Swaziland 0.054 0.083 0.100 0.051 Tanzania 0.049 0.087 0.173 0.082 Uganda 0.078 0.178 0.232 0.330 Zambia 0.034 0.136 0.233 0.284 Zimbabwe -0.020 0.158 0.117 0.102 Average 0.035 0.117 0.131 0.120

Note: *Central African Republic


100


101

Table 2: Country Codes for CFA and non-CFA countries Country Country Code

CFA Countries BENIN BE BURKINA FASO BF CENTRAL AFRICAN REPUBLIC CA CAMEROON CR CONGO CO COTE D’IVOIRE CI GABON GA MALI MI NIGER NI SENEGAL SE TOGO TO

Non-CFA Countries BOTSWANA BO BURUNDI BU ETHIOPIA ET GAMBIA GM GHANA GH KENYA KE LESOTHO LE MAGADASCAR MA MALAWI MW MAURITIUS MU NIGERIA NG RWANDA RW SOUTH AFRICAN REPUBLIC SA SWAZILAND SW TANZANIA TA UGANDA UG ZAMBIA ZA ZIMBABWE ZI

Table 3: Correlation of Supply and Demand Shocks

Notes:

CFA COUNTRIES NON-CFA COUNTRIES BE BF CA CR CD CO CI GA MI NI SE TO BO BU ET GM GH KE LE MA M

W MU NG RW SA SW TA UG ZA ZI

BE --- -.22 -.07 .08 .06 -.07 -.03 -.22 -.10 -.07 -.03 -.11 -.08 -.01 -.03 -.31 -.29 -.09 -.11 -.07 -.08 .01 -.16 .12 -.10 .05 -.04 -.07 .18 .06 BF .30 --- -.05 .18 .18 .12 .23 .05 .18 -.17 .16 -.24 -.01 .07 -.03 -.27 .06 -.16 -.11 .02 -.22 .00 -.10 -.07 .10 .06 -.03 -.03 -.03 .44 CA .42 .17 --- -.23 -.32 -32 .06 -.14 .14 -.05 -.21 .12 -.02 .01 -.07 .26 -.07 .20 -.01 -.30 .09 .16 -.32 -.35 -.18 -.07 .01 .00 .33 .22 CR .29 .38 .34 --- .29 -.00 .21 .09 .04 -.08 -.01 -.05 .19 .56 -.04 -.35 -.12 -.09 .19 -.40 -.17 .28 -.23 -.25 .43 -.07 .22 -.07 .19 .01 CD .07 .27 .24 .17 --- -.41 -.19 -.11 -.38 .03 .07 -.33 .08 -.20 -.09 -.33 -.12 -.15 .20 .06 -.10 .06 -.25 .05 .28 .03 .17 -.20 -.14 .15 CO .36 .46 .36 .13 .09 --- .10 .86 .10 -.00 -.08 .49 .04 -.12 .01 .08 -.16 .32 .26 -.14 .11 -.12 .01 -.27 -.12 .10 .01 .06 .14 -.03 CI -.00 .27 .01 -.04 -.24 .23 --- .05 .37 -.19 -.15 -.05 .22 .03 -.08 .08 .09 .13 -.03 -.21 -.08 .10 .13 -.29 .09 .10 .13 -.05 .36 .16 GA .54 .50 .51 .43 .20 .43 .06 --- .27 -.13 .19 -.03 .06 .22 .18 .02 .47 -.01 -.11 -.02 -.14 -.10 .15 .18 -.05 -.08 -.20 .37 -.11 .14 MI .46 .42 .46 .40 .43 .25 .11 .66 --- .01 .03 .21 .12 .33 -.06 -.01 .17 -.04 -.12 .02 .07 .08 .08 -.15 -.01 .07 -.19 .22 -.06 .25 NI .43 .48 .33 .19 .32 .32 .15 .49 .58 --- -.05 .07 -.35 -.08 -.01 -.14 -.01 -.26 -.05 .39 -.10 -.14 .15 -.02 .30 -.03 -.06 .34 -.19 .17 SE .41 .68 .49 .41 .25 .52 .17 .72 .68 .60 --- -.20 -.20 .11 .14 -.25 .06 -.08 -.34 .11 .04 .03 -.07 .38 .20 .25 -.16 -.13 -.07 .00 TO .12 .17 -.05 .07 .14 .28 .10 .07 .31 .16 .31 --- .27 -.06 .02 .29 -.04 .22 .01 .11 -.07 -.05 .18 -.15 .05 -.08 .02 .24 -.04 -.03 BO -.37 .26 .11 -.04 .42 .11 .01 -.16 -.10 .02 -.01 .22 --- -.08 .10 .03 .17 .30 .18 -.10 .05 .21 -.00 .12 -.08 .26 .08 .16 -.02 -.10 BU .03 .01 .25 .20 .42 -.08 -.00 -.03 .35 .25 .23 .33 .11 --- -.49 .08 .04 -.16 .02 -.22 .11 .19 -.13 .01 -.34 .20 -.36 -.00 -.08 .17 ET -.31 .21 .05 -.17 -.02 -.17 .25 -.23 -.15 .08 -.04 .06 .38 .10 --- .07 .26 .02 .01 .15 .20 -.31 -.05 .21 .28 -.17 -.07 .10 .13 -.18 GM -.07 .02 -.42 -.24 -.28 .06 -.02 -.03 -.30 -.23 -.13 .02 .03 -.51 .19 --- .27 .21 .23 -.14 .28 .36 .20 .00 -.45 .09 .16 .09 -.07 -.20 GH .05 .38 -.04 -.16 .25 .12 .12 .19 .30 .21 .26 .16 .35 -.04 .26 .25 --- -.15 -.24 .04 .08 -.02 .30 .03 -.09 .15 .18 .33 -.11 .17 KE -.23 .19 -.23 -.20 -.19 .16 .07 -.24 -.23 -.13 -.01 -.13 .12 -.20 .31 .37 .14 --- .14 .09 .07 .15 .01 .17 .26 .05 .12 .37 .19 .02 LE -.23 -.23 -.23 .01 -.00 -.21 .03 -.03 .00 -.18 -.25 -.24 .19 -.27 .01 .03 -.03 .21 --- -.19 .21 .10 -.16 .10 -.24 .17 .11 -.11 -.02 -.07 MA -.13 .30 .13 .08 .22 .32 -.16 .28 .27 .32 .38 .11 .27 .22 -.07 -.18 .31 -.08 -.38 --- .01 -.37 .45 .41 -.03 -.05 -.06 .33 -.35 .18 MW -.19 .27 -.06 -.01 .03 .42 .23 -.10 .01 .19 .16 -.02 .17 .04 .11 -.25 .18 .16 -.08 -.04 --- -.01 -.04 .09 -.10 .11 -.09 -.09 .03 -.16 MU .06 .29 .19 .12 .21 -.02 .02 .28 .31 -.07 .14 .15 .03 -.06 .12 -.21 .28 -.06 .08 .45 .15 --- -.08 -.17 -.12 -.10 .10 -.04 .09 -.01 NG -.20 .15 .18 -.03 -.01 .20 -.00 .03 .07 .18 .09 .02 .26 -.02 .35 -.07 .26 .13 .09 .40 .37 -.01 --- .15 -.06 -.17 -.08 .33 -.15 .02 RW -.09 .44 .08 -.17 .05 .13 .38 -.01 .22 .06 .18 .12 .14 .05 .43 .06 .17 .13 -.00 -.10 .24 .25 .10 --- -.27 .32 -.27 .10 -.21 -.13 SA .03 .14 .11 -.02 .09 .02 -.00 .29 .14 .08 .02 -.17 .09 -.27 -.21 -.02 .41 -.08 .22 .08 -.03 .16 .07 -.02 --- -.26 -.05 -.15 -.01 -.04 SW -.30 .25 .01 -.13 -.00 .18 .20 -.14 -.01 -.01 -.05 .15 .31 .08 .30 .03 .40 .24 -.11 .34 .52 .27 .44 .27 .28 --- .01 -.25 -.12 -.19 TA -.25 .01 -.50 .23 -.12 -.13 .01 -.10 -.25 -.19 -.11 -.05 -.00 -.11 .18 .26 -.08 .09 .27 .06 .06 -.16 .08 .13 -.05 .05 --- .00 .42 -.05 UG -.20 -.07 -.04 -.00 -.05 .15 .23 -.16 -.2 -.24 -.04 .11 .22 -.05 -.06 .18 .14 .10 -.02 -.02 .20 -.12 -.04 -.15 .07 .15 -.03 --- -.09 .40 ZA -.13 -.25 -.27 .09 -.16 -.08 -.05 -.22 -.31 -.35 -.30 .18 .10 -.15 .26 .39 .02 .01 .15 -.05 .04 -.09 .21 -.12 -.20 .09 .42 .25 --- .08 ZI -.13 .08 .06 -.11 -.02 -.12 .23 -.07 .06 .02 .15 .11 .07 .13 .20 -.12 -.04 .09 -.03 .02 -.16 -.22 .15 .46 -.21 .10 .07 -.08 .07 ---

-Correlations of supply disturbances are above the diagonal and correlations of demand disturbances are below the diagonal. -Bold denotes positively significant coefficients at the 5 percent level. At 5%, the critical value of the correlation coefficient, r is 0.26.


102

Table 4: Shock Sizes and Adjustment Speed in the CFA and NON-CFA Countries CFA COUNTRIES COUNTRIES εt

h ε ts ε t

z εtd ε t

m SS AS SS AS SS AS SS AS SS AS Benin 0.26 0.87 0.04 1.83 0.70 0.38 0.32 2.76 0.10 2.10 Burkina Faso 0.20 1.35 0.06 0.93 0.72 2.43 0.10 2.44 0.05 0.81 Cameroon 0.29 0.25 0.07 3.79 0.61 2.54 0.19 0.97 0.14 0.24 Central African Republic 0.29 0.41 0.11 0.07 0.65 3.26 0.11 2.73 0.06 0.30 Chad 0.28 0.34 0.36 2.30 0.68 0.08 0.11 0.25 0.77 0.38 Congo 0.24 0.12 0.11 3.69 0.69 0.81 0.16 0.38 0.06 0.56 Cote d'Ivoire 0.29 0.10 0.06 1.76 0.69 2.22 0.09 0.76 0.04 2.39 Gabon 0.25 2.22 0.14 0.22 0.69 1.29 0.10 0.52 0.06 3.84 Mali 0.25 0.40 0.10 3.75 0.67 1.20 0.18 0.07 .08 3.10 Niger 0.16 0.86 0.20 0.10 0.60 1.21 0.13 1.89 0.07 0.11 Senegal 0.26 1.40 0.08 0.09 0.66 2.38 0.10 3.79 0.08 3.00 Togo 0.28 2.54 0.06 3.21 0.61 0.26 0.29 0.17 0.11 0.77 Average CFA 0.25 0.91 0.12 1.81 0.66 1.51 0.16 1.39 0.14 1.47 COUNTRIES NON-CFA COUNTRIES Botswana 0.20 1.74 0.06 3.59 0.61 0.58 0.11 0.51 0.02 1.10 Burundi 0.27 3.58 0.28 0.80 0.64 0.44 0.23 1.17 0.09 3.02 Ethiopia 0.31 2.07 0.04 2.64 0.76 2.10 0.15 1.06 0.04 2.45 Gambia 0.24 1.43 0.23 3.00 0.74 0.79 0.37 2.96 0.07 2.86 Ghana 0.40 0.72 0.09 2.16 0.75 2.57 0.11 1.20 0.16 2.52 Kenya 0.27 1.26 0.14 0.64 0.72 1.81 0.13 0.76 0.25 0.20 Lesotho 0.13 1.03 0.07 2.12 0.65 1.84 0.08 1.70 0.02 0.96 Madagascar 0.26 1.71 0.04 3.11 0.70 3.16 0.11 1.88 0.11 2.23 Malawi 0.29 0.76 0.07 1.31 0.61 1.84 0.19 1.99 0.14 2.30 Mauritius 0.31 1.62 0.08 0.51 0.66 0.80 0.20 2.79 0.11 1.57 Nigeria 0.60 1.15 0.28 1.69 0.62 2.78 0.25 0.40 0.13 1.83 Rwanda 0.36 0.97 0.11 1.15 0.60 1.36 0.15 3.89 0.05 0.82 South African Republic 0.15 0.58 0.11 0.13 0.67 1.72 0.07 1.18 0.03 0.69 Swaziland 0.18 0.76 0.07 0.23 0.67 3.83 0.11 1.24 0.05 2.45 Tanzania 0.42 1.62 0.08 0.51 0.71 0.80 0.13 2.79 0.09 1.57 Uganda 0.52 1.71 0.15 1.41 0.57 1.31 0.26 1.18 0.27 3.41 Zambia 0.47 2.48 0.16 0.50 0.74 0.95 0.18 0.97 0.18 1.06 Zimbabwe 0.22 0.71 0.10 0.45 0.68 0.79 0.79 3.40 0.05 0.02 Average Non-CFA 0.31 1.66 0.12 1.91 0.67 1.64 0.20 1.74 0.10 1.91 Note: SS = Shock Size; AS = Adjustment Speed.

REFERENCES Bayoumi, T. and B. Eichengreen. 1997. Ever closer to heaven? An optimum-currency-area index for European countries. European Economic Review. 4:761-70.

Bayoumi, T., B. Eichengreen, and P. Mauro. 1999. On regional monetary arrangements for ASEAN. Working Paper. 1-30. _____. 1998. Exchange rate volatility and intervention: Implications of the theory of optimum currency areas. Journal of International Economics. 45:191-209.


103

Bayoumi, T, and O. D. Jonathan. 1997. Shocks and trade flows within sub-Saharan Africa: Implications for optimum currency arrangement. Journal of African Economies. 6(3): 412-44. Blanchard, O. and D. Quah. 1989. The dynamic effects of aggregate demand and supply disturbances. American Economic Review, 79, September: 655-73. Clément, J. A. P., J. Mueller, S. Cosse, and J. L. Dem. 1996. Aftermath of the CFA franc devaluation. International Monetary Fund, Occasional Paper. May, 138. Dickey, D. D., and Fuller, W. A. 1981. Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica. 49:1057-72. _____. 1981. Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica. 49:1057-72. Dickey, D. 1979. Distribution of estimates for autoregressive time series with a unit root. Journal of American Statistical Association, 74:427-431. Frankel, J. A. and A. K. Rose. 1998. The endogeneity of the optimum currency area criteria. The Economic Journal. 108, July: 1009-25. Fuller, W. A. 1976. Introduction to statistical time series. Wiley, 2nd ed., New York.

Horváth, J. and R. Grabowski. 1997. Prospects of African integration in light of the theory of optimum currency areas. Journal of Economic Integration. 12(1), March: 1-25. International Monetary Fund. 1999. Direction of trade statistics yearbook, (various years), International Monetary Fund, Washington, DC. International Monetary Fund. International financial statistics, CD-ROM, 6/2001, International Fund, Washington, DC. Kenen, P. B. 1969. The theory of optimum currency areas: An eclectic view. In Monetary problems of the international economy, eds. R. Mundell and A. Swoboda. University of Chicago Press, Chicago. McKinnon, R. I. 1963. Optimum Currency Area. American Economic Review, 53: 717-24. Mundell, R. A. 1961. A theory of optimal currency areas. The American Economic Review. 51 (3):657-65. _____. 1968. A Theory of optimum currency areas. International Economics. New York, Macmillan. _____. The Euro: How important? Cato Journal. Winter, 18(3):441-44. Sissoko, Y. and S. Dibooglu. 2006. The Exchange Rate System and Macroeconomic Fluctuations in Sub-Saharan Africa. Economic Systems, 30(2): 1-16.


104

DOES CONSUMER SENTIMENT AFFECT HOUSEHOLD SAVING?

I-Ming Chiu Department of Economics

Rutgers University, the State University of New Jersey 311 North 5th Street Camden, NJ 08102

ABSTRACT

Low household saving not only affects the welfare of retired individuals, it also decelerates the growth of an economy via its contribution to lower national saving. The objective of this paper is to identify macroeconomic variables that may help explain the decline in household saving starting in the mid 1990s. While many economic/financial variables appear to be deciding factors to household saving in past empirical studies, the popular and readily available consumer sentiment index from survey results is not on the list. Applying the dynamic ARDL model with Bounds Testing approach, we confirm the consumer sentiment as an important factor that affects household saving in both short- and long-run.

INTRODUCTION Since the mid 1990s, low household saving1 in the United States has attracted a lot of attention from economists and policy makers. Figure 1 reveals that the household saving rate has been decreasing over the sample period (1978:Q1 ~ 2008:Q4). This downward trend in household saving can be troublesome for two reasons. First, national saving (of which private saving - business and household - is a component) has been insufficient to support domestic investment. Investment fuels growth in output and wages, leading to a higher standard of living. The result is an increasing dependence on foreigners’ willingness to finance US investment. Second, household saving may be inadequate to support retirement of a substantial proportion of the population. Social Security, as it exists, will not be able to fill the gap. In 1950, the Social Security system had a comfortable worker-to-retiree ratio of 16 to 1. That number stood at 3 to 1 in 2005 and is expected to fall to 2 workers for every retiree by 2030. Given the potential unfavorable consequences of low household savings, there is a strong need to uncover its causes. Hopefully, with a better understanding of households’ saving decisions, this decreasing saving trend can be stopped and reverted. While there have been a significant amount of empirical studies (DeSerres and Pelgrin, 2002; Juster et al, 2004) that aim to explain this deteriorating saving trend by using various macroeconomic variables, none of them as far as we know of consider the very popular and readily available consumer sentiment index (CSI) as a part of the explanatory

variables. The consumer sentiment index appears to be an important indicator that explains current consumption and predicts future consumption (Carroll et al, 1994; Ludvigson, 2004). Though the role of consumer sentiment in households’ spending decisions is not well understood, it shouldn’t be excluded in the saving study since consumption and saving are simultaneously decided. Besides failing to include the CSI in a saving study, two other empirical difficulties arise. First, differenced method is usually required in time series study to transform nonstationary variables to stationary2 ones. This transformation process is essential to obtain desirable statistical estimates. Second, there is no way to explore the possibility that a long-run relationship may exist between household saving and its macroeconomic determinants if the difference method is used. An appropriate estimation approach is needed to distinguish how the relevant macroeconomic variables affect the household saving decision at different time horizons (short vs. long run). In this paper we attempt to tackle the above two issues in the household saving study by including the consumer sentiment index as an additional explanatory variable and utilizing an updated empirical approach. The ultimate goal is to better understand how households optimize their spending and saving decisions at an aggregate level. The remainder of the paper is organized as follows. Based on literature review, Section Two provides a set of macroeconomic variables that may help explain household saving decisions. Special attention will be paid to the consumer sentiment index variable. Section Three addresses model and data specifications. A preliminary data examination is also provided in this section. Section Four briefly explains the empirical approach, an autoregressive distributed lag model (ARDL) with Bounds Testing Approach (BTA). Section Five addresses empirical test results. Section Six offers summaries and conclusions.

MACROECONOMIC DETERMINANTS OF

HOUSEHOLD SAVING Modern consumption theory is based on a mixture of “Keynesian theory” and “Life-Cycle /Permanent Income theory”. The former focuses on how consumer and investor sentiment – what Keynes (1936) referred to as “animal spirits” – may influence the real economy. The latter is


105

proposed by Modigliani (1971) and Friedman (1957, 1963) and focuses on lifetime consumption decision making. The life cycle theory emphasizes how to maintain a stable standard of living in the face of changes in income over the course of life. Permanent income theory focuses on forecasting the level of income available to a consumer over a lifetime. Today, these two have largely merged3. Campbell and Mankiw (1989) have used a unique empirical approach to confirm the validity of both theories. Guided by theories and based on a rich amount of empirical literature, we consider the following six factors to be the influential macroeconomic determinants to household saving. Consumer sentiment Despite the widespread attention given to surveys of consumer confidence, the mechanisms of how consumer confidence affects the real economy is still a puzzle. According to Van Raaij and Gianotten (1990), consumer sentiment index is a “balance of opinion” that is calculated based on survey results. The survey questions concern the present and expected financial situation of households, the present and expected economic situation, and the advisability of making major purchases. Many authors have successfully incorporated consumer sentiment in a consumption function (Katona 1975; Carroll et al, 1994; Ludvigson, 2004). Many others are quite suspicious about the usefulness of using the consumer sentiment index to explain/predict consumption expenditures. They argue that the explanatory power of the consumer sentiment index is limited once traditional economic/financial variables are introduced (Acemoglu and Scott, 1994; Santero and Westerlund, 1996; Fan and Wang, 1998). The contrasting opinions about the consumer sentiment index suggests that whether the index’s information contents are purely reflecting consumers’ answers to survey questions or revealing consumers’ independent and subjective mind on the economy. If the first suggestion is correct, then consumer sentiment is a redundant variable; it can be replaced by a bunch of relevant macroeconomic variables. Katona (1968, 1975), who undertook pioneering research on consumption/saving decisions of households, argues that consumer sentiment represents households’ “willingness” to consume and should be distinguished from their “ability”. A household’s willingness refers to its attitude and expectations about personal finances and the economy as a whole, while ability pertains to its income and wealth. These two interact but may evolve separately. In other words, consumer sentiment may include some independent information (e.g., subjective minds of consumers) that cannot be captured by some macroeconomic measurements.

Given the mixed results about the usefulness of consumer sentiment in a household’s spending decisions, it would be an important contribution to the current saving literature if we can empirically explore and find whether it is an additional deciding factor to households’ saving decisions. Demographics There are two views regarding how demographical factor affects household saving. According to LC-PIH, the main purpose of saving is to finance retirement. Retired people dissave. Therefore, a rise in the old-age dependence ratio4 decreases household saving. The second view concerns the balance between consumers’ impatience on spending and their precautionary need for saving. According to this view, younger people have less wealth for drawing on during rainy days and they are inclined to save more. On the other hand, when consumers reach their middle ages, say between 40 and 45, and they have accumulated wealth. Therefore, they become more impatient to save and tend to spend additional income. Therefore, when the middle-aged population accounts for a larger portion of the total population, the household saving may decrease as a result. We take the second view in our model specification since empirical evidence (Carroll, 1997) indicates that retired people rarely actually dissave. They tend to live off the income (e.g., interest payments and dividends) from their wealth. Real interest rate The effect of (real) interest rates on household savings is often ambiguous due to the competing substitution and income effects that take place in a standard intertemporal consumption model. Bosworth (1993) found a positive interest rate coefficient in time-series estimation for individual countries, but then found a negative coefficient in a panel study. DeSerres and Pelgrin (2002) found that interest rates were negatively related to savings in a panel study that focuses on OECD countries. It is possible that the substitution effect of higher interest rates, which makes current consumption more expensive relative to future consumption, was offset by the income effect. If the private sector is a net creditor, a rise in rates lifts income and consumption, while lowering savings. The more that households in that sector are liquidity constrained, the more likely the income effect is to outweigh the substitution effect. Given that the US has a very sound financial market, it isn’t difficult for households to require credit when needed. We expect that higher real interest rates may induce more savings due to a more dominant substitution effect. Net wealth The standard LC-PIH models predict that consumers will increase consumption by a fraction of their capital gains in


106

housing or equities, and that households may offset real capital gains through a reduction in traditional savings. According to some estimates (Parker, 1999; Lettau and Ludvigson, 2004), if the marginal propensity to consume out of wealth is 0.05, the increase in wealth could account for 5 percentage points of the decline in the saving rate. Using a net wealth to disposable income ratio as a relative wealth measure, Figure 3 displays that when this relative wealth measure (symbol in the figure: NW_Y) rises household saving rate (HSR) tend to move in the opposite direction. Productivity growth In a LC-PIH framework, a rise in per capita income is likely to benefit workers more than retirees. Given the higher saving rates of workers, this would generate a larger aggregate savings. The works of Carroll and Weil (1994) and DeSerres and Pelgrin (2002) have found a positive correlation between productivity growth and saving rates. However, if productivity is seen as an indication of better times ahead, LC-PIH predicts that workers will reduce savings in response to higher growth rates. Public sector saving According to the Ricardian equivalence proposition, private saving increases if government debts (indicates negative public savings) are financed by issuing bonds. The reason is that government bonds do not represent additional wealth held by households. Instead, these bonds merely point to a postponement of taxation. Forward looking households will save more for future higher taxation if they observe the government has run continuous budget deficits. While no researcher has been able to demonstrate strict Ricardian equivalence, some have found a negative relationship between government saving and personal saving. Bernheim (1987) examined industrial countries and found that a unit increase in government deficits was matched by a 0.5 to 0.6 decrease in consumption. Corbo and Schmidt-Hebbel (1991) found similar results for developing countries.

SPECIFICATION OF THE MODEL AND DATA According to the discussion from section 2, an aggregate household saving function can be represented in the following, HSR = f (CSI, P_RATIO, r, NW_Y, PRD, PUB_S) (1)

The household saving rate (HSR) depends on six explanatory variables. They are the consumer sentiment index (CSI), the proportion of population that is above age 45 (P_RATIO), the real interest rate (r), the net wealth to disposable income ratio (NW_Y), productivity (PRD), and the public saving rate

(PUB_S). Based on the literature review from the previous section, the sign under each determinant in the function indicates expected relationship of each with HSR. While we expect CSI, P_RATIO, NW_Y, and PUB_S each have a negative impact on household saving rate, the effect of r and PRD on household saving rate is positive. The household saving rate is defined as saving out of disposable income and data is obtained from the Bureau of Economic Analysis (BEA), an agency of the US Department of Commerce. The consumer sentiment index comes from Survey Research Center/University of Michigan. Net wealth data comes from The Federal Reserve Statistical Release: Flow of Funds Accounts of the United States and disposable income is from the BEA. The real interest rate is calculated by using the difference between the 10-year government bond rate and a consumer-based inflation rate, both of which are obtained from the BEA. Productivity is defined as real GDP divided by total employment. Real GDP and employment data are both obtained from the BEA. Public saving rate is defined as net government saving to real GDP ratio. Where net government saving is the net taxes (taxes exclude transfers) subtracted by government consumption and then added to government investment. The components that construct public saving rate are all obtained from the BEA. All the data begins in the first quarter of 1978 and ends in the last quarter of 2008. We utilize the data till the last quarter of 2006 for estimation purpose. Predicted household saving rate derived from the empirical results is then compared to the actual household saving rate in the last eight quarters (2007 Q1 ~ 2008:Q4) to evaluate the model’s out-of-sample explanatory power. Table 1 provides pair-wise correlation statistics between household saving rate and its six macroeconomic determinants. It can be found that, after a quick examination on the first numerical column, all of the factors except for real interest rate are negatively correlated with the household saving rate. Among these six determinants, population ratio, real interest rate, and net wealth to disposable income ratio have stronger correlations with the household saving rate. Moving to the second numerical column, it seems that consumer sentiment index doesn’t have strong correlations with the rest of the five determinants. It has strongest correlations with net wealth to disposable income ratio and public sector saving rate, however, both are below 0.5 (0.413 and 0.429 respectively). This may indicate that the consumer sentiment index provides unique information contents to household saving decisions and cannot be captured by other macroeconomic determinants.

- - + - + -


EMPIRICAL TEST APPROACH: THE AUTOREGRESSIVE DISTRIBUTED LAG MODEL

(ARDL) WITH BOUNDS TESTING APPROACH (BTA)

∑=

−Δβn

1jjtj HSR*To better understand how household saving decisions are

made at the aggregate level, the Autoregressive Distributed Lag Model (ARDL) is used. The ARDL model includes current and lagged independent variables (or includes lagged dependent variables as well) to capture the dynamic structure of economic variables. Though six candidate macroeconomic determinants of household saving is identified earlier, we are not able to differentiate how these determinants affect household saving at different time horizons. To be more specific, we cannot decide whether these determinants only have short-run impact on household savings or they have lasting (long-run) effect on household savings as well. In order to overcome this problem, the concept of cointegration will be incorporated in the ARDL model. Cointegration is used to capture the notion that nonstationary variables may possess long-run equilibrium relationships and thus have a tendency to move together in the long run (Engle and Granger, 1987). Cointegration implies an “error correction model” (ECM), where dependent variable not only responds to short-run changes from its determinants but also responds to long-run adjustments between itself and its determinants. With this modified ARDL model, the macroeconomic determinants’ short-run and long-run effects on household saving can be analyzed simultaneously. The ARDL model can be represented in the following equation,

HSRt = a0 + + +

+ +

∑=

−n

1jjtj HSR*a

tj )RATIO_P(*c

∑=

−n

0jjtj CSI*b

− j ∑=

−n

0jjtj r*d∑

=

n

0j

+ +

+ εt (2)

∑=

−n

0jjtj )Y_NW(*e

∑=

−n

0jjtj )S_PUB(*g

∑=

−n

0jjtj PRD*f

Equation (2) assumes a linear dynamic structure of the variables defined in the household saving model in section 2. It states that the household saving rate at current time t, denoted by HSRt, depends on the past household saving rate (HSRt-j, j ≥ 1) as well as other current and past macroeconomic determinants; this includes the consumer sentiment index (CSI), the (middle-aged) population ratio (P_RATIO), the real interest rate (r), the net wealth to disposable income ratio (NW_Y), the productivity (PRD),

and the public saving rate (PUB_S). Equation (2) can be re-parameterized and specified as follows,

ΔHSRt = α + + +

+ +

+ +

+ [δ0 + δ1* HSRt-1 +

δ2 * CSIt-1 + δ3 * (P_RATIO)t-1 + δ4 * rt-1 + δ5 * (NW_Y)t-1 + δ6 * PRDt-1 + δ7 * (PUB_S)t-1] + εt (3)

∑=

−Δγn

1jjtj CSI*

− j ∑=

−Δλn

1jjtj r*

∑=

Δρn

1jtj PRD*

∑=

Δθn

1jtj )RATIO_P(*

∑=

−Δωn

1jjtj )Y_NW(*

∑=

−Δηn

1jjtj )S_PUB(*

− j

Where “Δ” is the first difference symbol, which indicates a “short-run” changes in variables. Equation (3) points out that the short-run household saving rate is dependent on the short-run macroeconomic determinants plus a long-run factor that includes the same determinants in levels (all the variables in the square bracket in Equation (3)). This “long-run” factor is termed the cointegrating vector (Engle and Granger, 1987), which is used to explain the long-run relationship between household saving and its determinants in an “econometric” sense. Equation (3) can also be derived and estimated in a vector error correction model (VECM) setting, in which household saving rate and its macroeconomic determinants are all treated as endogenous. However, the disadvantages of using VECM includes significant loss of degree of freedom and the need to pretest the time series property of each variable (i.e., whether they are stationary or nonstationary). Pesaran et al. (1995, 1996, and 2001) introduced a bounds testing approach (BTA) to overcome the problems addressed above. Equation (3) can be estimated by using the BTA, which involves two steps. First, after selecting the appropriate number of lags (i.e., choosing j in equation (3)), the null of no long-run relationship (H0: all δs equal zero) is tested against the alternative hypothesis (H1: not all δs equal zero). Pesaran et al. (1996) provides an appropriate F-statistic table to help determine whether there exists a long-run relationship among variables. This test includes two sets of critical values: one set assuming all the variables used in the model are nonstationary or I (1); another assuming all variables used are stationary or I (0). For each application, these two sets of values provide a band to cover all the possible classifications of variables into I (0) or I (1). If the computed F-statistic lies above the upper level of the band, the null is rejected. Thus, a long-run relationship exists (variables are cointegrated). If the computed F-statistic falls below the lower level of the


band, the null cannot be rejected. There is no long-run relationship. Otherwise, the inference is inconclusive. If the long-run relationship cannot be rejected, Equation (3) can be reorganized by normalizing the coefficient of HSR in the square bracket and is written as follows,

ΔHSRt = α + + +

+ +

+ +

+ δ1 [HSRt-1 + φ0 +

φ1 * CSIt-1 + φ2 * (P_RATIO)t-1 + φ3* rt-1 + φ4 * (NW_Y)t-1 +φ5* PRDt-1 + φ6 * (PUB_S)t-1] + εt (4)

∑=

−Δβn

1jjtj HSR*

Δθ tj )RATIO_P(*

−Δω jtj )Y_NW(*

−Δη jtj )S_PUB(*

∑=

−Δγn

1jjtj CSI*

− j ∑=

−Δλn

1jjtj r*

∑=

Δρn

1jtj PRD*

∑=

n

1j

∑=

n

1j

∑=

n

1j

− j

The square bracket part in Equation (4) is the error correction term or cointegrating vector, which represents a long-run equilibrium relationship between the household saving rate and its determinants. It is denoted by ECM and a final representation of the ARDL model yields,

ΔHSRt = α + + +

+ +

+ +

+ δ1*ECMt-1 + εt (5)

∑=

−Δβn

1jjtj HSR*

Δθ tj )RATIO_P(*

−Δω jtj )Y_NW(*

−Δη jtj )S_PUB(*

∑=

−Δγn

1jjtj CSI*

− j ∑=

−Δλn

1jjtj r*

∑=

Δρn

1jtj PRD*

∑=

n

1j

∑=

n

1j

∑=

n

1j

− j

In Equation (5), δ1 is an adjustment coefficient and is expected to carry a negative sign; it indicates how fast the current change in household saving rate (ΔHSRt) responds to the long-run disequilibria from the previous period, which is captured in the error correction term (ECMt-1). Akaike information criterion (AIC) is used to select the optimal lag lengths in equation (5), while the maximal lag lengths is set at four given that quarterly data is used in our estimation.

EMPIRICAL TEST RESULTS Following the two-step Bounds Testing Approach, the long-run relationship test is conducted and the resulting F statistics (4.5546) is greater than the table value (upper band value =

3.6460; this value is based on a cointegrating vector with intercept but no trend) at the 95% of significance level. Therefore, we conclude that a long-run relationship exists. Table 1 reports the estimated coefficients and corresponding t-statistics/p-values for long-run determinants. First we notice that all the estimated coefficients carry the correct signs as predicted. While higher real interest rate and larger productivity encourages more household savings, an increase of the rest of determinants decreases household savings. If a 10% significance level is adopted, the consumer sentiment index (CSI), the population ratio (P_RATIO), and the real interest rate (r) have significant effects on the household saving rate. If we allow for a 20% of significance level, then all the macroeconomic determinants except for public saving rate5 (PUB_S) have a significant long-run relationship with household saving rate. According to the magnitude of the estimated coefficient of CSI, household saving rate will decrease by 0.0439 of one-percent if the CSI increases by one unit. Table 2 provides the entire estimates of the ARDL model, which includes both the short-run determinants (those in first differences) and the long-run adjustment term (the error correction term; ECMt-1). The AIC selects the current differenced term for the consumer sentiment index (CSI), the population ratio (P_RATIO), and the real interest rate (r). However, for net wealth to disposable income ratio (NW_Y) and public saving rate (PUB_S), one additional lagged differenced term is included. For productivity (PRD), two additional lagged differenced terms are selected. If a 10% significance level is allowed, the differenced consumer sentiment index (ΔCSI), population ratio (ΔP_RATIO), and the real interest rate (ΔR) carry the same expected sign like their long-run counterparts and have significant effects on the household saving rate (ΔHSR). For the rest of the three macroeconomic determinants, we calculate their accumulated net impact on household saving rate by focusing on significant determinants only, where the significant level is set at 10%. For the net wealth to disposable income ratio, while its current change (ΔNW_Y) has no significant impact, its lagged differenced term ((ΔNW_Y1 ≡ NW_Y (-1) – NW_Y(-2)) has a positive effect on the household savings. For productivity, both of its lagged differenced terms (ΔPRD1 ≡ PRD(-1) – PRD(-2) and ΔPRD2 ≡ PRD(-2) – PRD(-3)) have a net negative effect (-.3734 + -.3241 = -.6975) on the household saving rate. Be notice that households do not respond to the immediate changes in NW_Y and PRD in our empirical results. The last determinant, public saving rate, has a net negative (-.4096 + .2503 = -.1593) effect on the household saving rate, which is same as its long-run effect. According to the above empirical results, we conclude that consumer sentiment has significant short- and long-run effects on household saving decisions when other influential macroeconomic determinants are taken into considerations. This finding supports Katona’s


109


110

argument that consumer sentiment may contain independent information that is important to household spending/saving decisions. The estimated adjustment coefficient of ECM ( = -.4614) is significant and has the expected negative sign. This indicates that any disequilibrium in the long-run household saving rate that occurred from the previous period will be carried forward to the current short-run household saving rate (ΔHSR) and follows a downward adjustment. The adjustment speed is about an average of 46.14% quarterly.

1̂δ

In Table 4, we provide the predicted household saving rates based on the ARDL empirical results versus the actual household saving rates in the period of 2007:Q1 to 2008:Q4. The predicted saving rates are able to mimic the movement of the actual ones. The corresponding Fig. 1 provides a visual comparison. The predicted saving rate line is smoother than the actual saving rate line; the differences between these two lines indicate that some shocks are not captured in our empirical model.

SUMMARY AND CONCLUSIONS Low household saving not only affects the welfare of retired individuals, it may also decelerate the growth of an economy via its contribution to lower national saving. The main objective of this paper is to identify macroeconomic variables that may help explain the decline in household saving that started in the mid of 1990s. While many economic/financial variables appear to be deciding factors to household saving

decisions in the past empirical studies, the popular and readily available consumer sentiment index from survey results is not on the list. This ignorance may cause a model misspecification if consumer sentiment index provides unique information that is independent from the rest of saving determinants. The Bounds Testing Approach (BTA) with Autoregressive Distributed Lag (ARDL) model is utilized to determine the short-run and long-run effects of these macroeconomic determinants on household saving. The test results indicate that, while other macroeconomic determinants are important to household saving decisions at different time horizons, consumer sentiment index does provide additional information in explaining the variation of household saving rate. Its impacts on household saving are significant in both short- and long-run. The estimated household saving rate equation (i.e., equation (4)) reveals that the quarterly change in household saving rate not only responds to period by period changes of its macroeconomic determinants, it also absorbs any deviations from long-run equilibrium in the system. The empirical results are useful in the sense that it is able to predict the household saving in an out-of-sample range. Given the importance of consumer sentiment in saving decision, we would like to further investigate why its information contents are independent from other economic and financial determinants in the near future. This may involve a reliance on additional survey data on how households respond to survey questions.

Table 1. Correlation _________________________________________________ HSR CSI P_RATIO r NW_Y PRD PUB_S HSR 1.000 CSI -0.374 1.000 P_RATIO -0.862 0.110 1.000 r 0.910 -0.286 -0.757 1.000 NW_Y -0.908 0.413 0.836 -0.767 1.000 PRD -0.070 0.210 0.006 -0.062 0.043 1.000 PUB_S -0.484 0.429 0.335 -0.346 0.555 0.034 1.000 _________________________________________________ Table 2. Estimation of Cointegrating Vector Estimated long-run coefficients in the cointegrating vector (ECMt-1) using the bounds testing approach _________________________________________________ Dependent variable is HSR 112 observations used for estimation from 1979Q1 to 2006Q4 -------------------------------------------------------------------------- Regressor Coefficient Standard ErrorT-Ratio[Prob] CSI -.0439 .0162 -2.7062[.008] P_RATIO -.3228 0.1158 -1.7419[.0850] r .6745 .0781 8.6421[.0000] NW_Y -.9298 .6761 -1.3754[.1720] PRD .9336 .5928 1.5749[.1190] PUB_S -.0740 .1047 -.7069[.4810] CONST 22.1990 3.9702 5.5914[.0000] _________________________________________________ Table 3. Estimation of ARDL model w/Error correction term Error correction representation for the selected ARDL model The optimal lags selected is based on the Akaike information criterion _________________________________________________ Dependent variable is ΔHSR 112 observations used for estimation from 1979Q1 to 2006Q4 -------------------------------------------------------------------------- Regressor Coefficient Standard ErrorT-Ratio[Prob] ΔCSI -.0203 .0070 -2.8884[.0050] ΔP_RATIO -.1489 .0571 -2.6063[.0110] Δr .3112 .0515 6.0375[.0000] ΔNW_Y -1.0168 .0832 -.4100[.6830] ΔNW_Y1 1.0020 .0887 1.9988[.0480] ΔPRD .0290 .1062 .2732[.7850] ΔPRD1 -.3734 .1654 -2.2573[.0260] ΔPRD2 -.3241 .1013 -3.1970[.0020] ΔPUB_S -.4096 .1545 -2.6509[.0090] ΔPUB_S1 .2503 .1481 1.6904[.0940] ΔCONST 10.2434 2.4456 4.1885[.0000] ECM (-1) -.4614 .0688 -6.7018[.0000] --------------------------------------------------------------------------


111

List of additional temporary variables created:

ΔCSI = CSI – CSI(-1*) ΔP_RATIO = P_RATIO – P_RATIO(-1) Δr = r – r(-1) ΔNW_Y = NW_Y – NW_Y(-1) ΔNW_Y1 = NW_Y(-1) – NW_Y(-2) ΔPRD = PRD - PRD(-1) ΔPRD1 = PRD(-1) - PRD(-2) ΔPRD2 = PRD(-2) - PRD(-3) ΔPUB_S = PUB_S - PUB_S(-1) ΔPUB_S1 = PUB_S(-1) - PUB_S(-2) ΔCONST = CONST-CONST(-1) ECMt-1 = HSRt-1 + .0439*CSI t-1 + .3228*(P_RATIO) t-1 -.6745*r t-1 + .9298*(NW_Y) t-1 - .9336*PRD t-1 +.0740*(PUB_S) t-1 -

22.1990*CONST -------------------------------------------------------------------------- R-Squared .5495 R-Bar-Squared .4845 S.E. of Regression .5987 F-stat. F(11, 100)10.7556[.0000] Residual Sum of Squares 34.7761 Equation Log-likelihood -93.4253 Akaike Info. Criterion-108.4253 Schwarz Bayesian Criterion -128.8140 DW-statistic 2.1094 * “-j” indicates the time lag; -1 indicates the previous period. _________________________________________________ Table 4. Actual vs. predicted household saving rate Forecasts for the level of HSR ________________________________________________ Observation Actual Prediction Error -------------------------------------------------------------------------- 2007Q1 1.1000 .5740 -.5260 2007Q2 .3000 .4035 .1035 2007Q3 .5000 .3648 -.1352 2007Q4 .4000 .3610 -.0390 2008Q1 .2000 .8465 .6465 2008Q2 2.5000 1.8868 -.6132 2008Q3 1.3000 1.0372 -.2628 2008Q4 3.2000 2.0593 -1.1407 _________________________________________________ Figure 1

Household Saving Rate (hsr)

-202468

101214

1978

Q1

1979

Q4

1981

Q3

1983

Q2

1985

Q1

1986

Q4

1988

Q3

1990

Q2

1992

Q1

1993

Q4

1995

Q3

1997

Q2

1999

Q1

2000

Q4

2002

Q3

2004

Q2

2006

Q1

2007

Q4

Time

Per

cen

t (%

)

hsr


112


113

Figure 2 Household Saving Rate vs. Consmer Sentiment Index

-202468

101214

1978

Q1

1980

Q1

1982

Q1

1984

Q1

1986

Q1

1988

Q1

1990

Q1

1992

Q1

1994

Q1

1996

Q1

1998

Q1

2000

Q1

2002

Q1

2004

Q1

2006

Q1

2008

Q1

Time

Saving Rate (%)

0

20

40

60

80

100

120

Sentiment Index

hsr

csi

Figure 3

Household Saving Rate (HSR) vs. Net Wealth to Disposable Income (NW_Y) Ratio

-202468

101214

1978

Q1

1980

Q1

1982

Q1

1984

Q1

1986

Q1

1988

Q1

1990

Q1

1992

Q1

1994

Q1

1996

Q1

1998

Q1

2000

Q1

2002

Q1

2004

Q1

2006

Q1

2008

Q1

Time

HSR

0.001.002.003.004.005.006.007.00

NW_Y

hsr

nw_y

Figure 4

Out-of-Sample Forecast

0.0000

0.5000

1.0000

1.5000

2.0000

2.5000

3.0000

3.5000

2007

Q1

2007

Q2

2007

Q3

2007

Q4

2008

Q1

2008

Q2

2008

Q3

2008

Q4

Time

Savi

ng R

ate

(%)

Actual_HSR

Predicted_HSR

ENDNOTES 3 See Dornbusch et al., 2004 for a quick reference; we will use LC-PIH to denote these two theories thereafter.

1 We prefer to use this term than “personal” saving since our focus is on how “households” allocate their disposable income between consumption and saving.

4 Defined as population who are aged 65 and above to total population ratio.

2 The use of standard regression approach, such as OLS, may result in spurious results (Granger & Newbold, 1974) if variables used in regression are nonstationary. Most of macroeconomic time-series variables are nonstationary.

5 This may indicate that there is no “Ricardian equivalence” in the long run.

REFERENCES

Acemoglu, D. and A. Scott. 1994. Consumer confidence and rational expectations: Are agents’ beliefs consistent with the theory? Economic Journal. 104, 1994: 1-19. Bernheim, B.D. 1987. Ricardian Equivalence: An Evaluation of Theory and Evidence. In NBER Macroeconomics Annual 1987, 263-304, Cambridge, MA: MIT Press. Bosworth, B.P. 1993. Savings and Investment in a Global Economy. Washington D.C.: Brookings Institute. Campbell, J. and G. Mankiw. 1989. Consumption, Income, and Interest Rates: Reinterpreting Time-Series Evidence. NBER Working Paper 2924. Cambridge, MA: National Bureau of Economic Research. Carroll, C. 1997. Buffer-stock saving and the life-cycle/ permanent income hypothesis. Quarterly Journal of Economics. 112, 1997: 1-55. Carroll, C., J. Fuhrer, and D. Wilcox. 1994. Does consumer sentiment forecast household spending? If so, why? American Economic Review. 84, 1994: 1397-408. Carroll, C.D. and S.D. Lawrence. 1991. Consumption Growth Parallels Income Growth: Some New Evidence. In National Savings and Economic Performance, 305-43, ed. Bernheim and B. Shoven. Chicago, IL: University of Chicago Press. Carroll, Christopher D. and D.N. Weil. 1994. Saving and growth: A reinterpretation. Carnegie-Rochester Conference Series on Public Policy, Elsevier. 40(1), June 1994: 133-92. Corbo, Vittorio and K. Schmidt-Hebbel. 1991. Public policies and savings in developing countries. Journal of Development Economics 36, July 1991: 89-115. DeSerres, A. and F. Pergrin. 2002. The Decline in Private Savings Rates in the 1990s in the OECD Countries: How Much Can Be Explained by Non-Wealth Determinants? Working Paper, OECD Economics Department. No. 344, December 2002. Engelhardt, Gary. 1995. House prices and home owner savings behavior. Regional Science and Urban Economics 26, 1995: 313-36. Engle, R.F. and C.W.J. Granger. 1987. Cointegration and error correction: Representation, estimation, and testing. Econometrica 55, 1987: 251-76.

Fan, S., & P. Wong. 1998. Does consumer sentiment forecast household spending? The Hong Kong case. Economics Letters, 58, 1998: 77-84. Friedman, M. 1957. A Theory of the consumption function. Princeton. NJ: Princeton University Press. ______. 1963. Windfalls, the ‘Horizon,’ and Related Concepts in the Permanent-Income Hypothesis. In Measurement in Economics: Studies in Mathematical Economics and Econometrics in Memory of Yehuda Grunfeld, ed. Carl Christ et al. Stanford, Calif.: Stanford Univ. Press. Granger, C.W.J. and P. Newbold. 1974. Spurious regression in econometrics. Journal of Econometrics 2, 1974: 111-20. Juster, F. T., L. Lipton, J.P. Smith, and F. Stafford. 2004. The Decline in Household Savings and the Wealth Effect. RAND University of Michigan. Katona, G. 1968. Consumer behavior: Theory and findings on expectations and aspirations. American Economic Review, Vol. 58, No. 2, Papers and Proceedings of the Eightieth Annual Meeting of the American Economic Association.1968: 19-30. Katona, G. 1975. Psychological Economics. Amsterdam: Elsevier. Keynes, J.M. 1936. The General Theory of Employment, Interest, and Money. New York: Hartcourt, Brace. Lettau, M. and S. Ludvigson. 2004. Understanding trend and cycle in asset values reevaluating the wealth effect on consumption. American Economic Review. 94, 2004: 276-99. Ludvigson, S. 2004. Consumer confidence and consumer spending. The Journal of Economic Perspective. Vol. 18, No. 2, 29-50. Modigliani, Franco. 1971. Monetary Policy and consumption. In Consumer Spending and Monetary Policy: The Linkages. Conference Series No. 5. Boston: Federal Reserve Bank of Boston. Modigliano, Franco and R. Brumberg. 1954. Utility Analysis and the Consumption Function: An Interpretation of Cross-Section Data. In Post-Keynesian Economics, 288-436, ed. E.E. Kurihara. New Brunswick, NJ: Rutgers University Press.


114

Parker, Jonathan A., B.S. Bernanke and J.J. Rotemberg. 1999. Spendthrift in America? On Two Decades of a Decline in the U.S. Saving Rate. NBER Macroeconomic Annual 1999, 317-387, Cambridge, MA: MIT Press. Pesaran, M.H. and Y. Shin. 1995. An Autoregressive Distributed Lag Modeling Approach to Cointegration Analysis. DAE Working Paper, No. 9514, Department of Applied Economics, University of Cambridge. Pesaran, M.H., Y. Shin, and R.J. Smith. 1996. Testing for the existence of a long-run relationship. DAE Working Paper, No. 9622, Department of Applied Economics, University of Cambridge.

Pesaran, M.H., Y. Shin, and R.J. Smith. 2001. Bounds testing approaches to the analysis of level relationship. Journal of Applied Econometrics 16, 2001: 289-326. Santero, T. and N. Westerlund. 1996. Confidence indicator and their relationship to changes in economic activity. Working Paper OECD, pp. 170.

Van Raaij, F. and H. Gianotten. 1990. Consumer confidence, expenditure, saving, and credit. Journal of Economic Psychology, 10, 1990: 473-93.


115

THE EFFECTS OF URBANIZATION ON DEVELOPMENT: AN ANALYSIS OF LATIN AMERICA AND SUB-SAHARAN AFRICA

Chelsea Kaufman

Clarion University of Pennsylvania Clarion, PA 16214

ABSTRACT This study will attempt to discern the effects that urbanization has on human development rather than only economic growth. An analysis of several countries in Sub-Saharan Africa and Latin America were used in order to accomplish this goal. As one of these regions is undergoing urbanization and the other has a more established urban population, both the changes occurring during urbanization and the aftermath may be examined. Data was collected on these regions from 1990 – 2005 on variables such as urban growth, the proportion of the population living in urban areas, the age of the population, and the disparity between rural and urban areas. Then, these variables were analyzed in order to determine their association with human development.

INTRODUCTION

In developing countries, a recurring pattern is that much of the population will move from rural areas to urban areas. The motivation for this occurs when the country begins to industrialize and unemployed workers in the agricultural sector begin to move to urban areas in hopes of finding employment (Harris and Todaro, 1970). In the past, many studies have examined why people migrate, who migrates, or what effects the migration has on economic growth. The goal of this study is to take this research one step further, by attempting to determine the relationship between urbanization and human development. In order to accomplish this, two general hypotheses will be tested. The first of these is that as urbanization increases, development increases. The second is that if migration determinants are present, development will increase as well. The study will focus on two different regions: Latin America and Sub-Saharan Africa. These were chosen because both nations are experiencing the urbanization phenomenon, but Latin America has much higher human development. By keeping the regions separate rather than combining them, looking at the different characteristics of the regions will allow the analysis to examine the differences in the results. By doing the analysis in this manner, the study may possibly determine why these differences exist and what ramifications these differences may have for policy.

LITERATURE REVIEW

Migration Determinants Rural – Urban migration takes place for a multitude of reasons. One study of Ghana made an attempt to analyze the determinants of migration. First, regional differences were considered. These differences included types of agriculture, the relation of agriculture to the market, and the extent of socio-economic changes that have influenced traditional ways (Caldwell 1968). Differences in time were also considered, as the propensity to migrate increases in younger generations (Caldwell 1968). Also, climatic differences, which could lead to seasonal migration, were considered (Caldwell 1968). After making these determinations, several factors were analyzed in order to observe their relationship to the propensity for migration. A determinant of migration which several studies cite is a youthful population (Caldwell; Detang-Dessendre et al., 2002). Those who are between the ages of 15 and 24 have fewer family obligations and thus the opportunity cost of moving to an urban area for potential employment is much lower (Detang-Dessendre et al., 2002). One analysis in this study considers the effects of distance on the likelihood of migration. According to the results, the propensity to migrate was inversely related to the distance from the urban area at the .1% level (Caldwell 1968). The amount of education that a person had received also contributed to propensity to migrate: as one becomes more educated, he or she will move to the urban area to find employment in a non-agriculture job (Caldwell 1968). Furthermore, this propensity increases if the person in mind has familiarity with English (Caldwell 1968). Having relatives already living in urban areas and having a large family were other statistically significant factors (Caldwell 1968). Interestingly, some factors which had statistically significant relationships to migration did not have a large affect on the area at large. Economic living standards, for example, remained constant amongst those who had migrated; were seasonal migrants; or who never intended to migrate, even though those who had migrated had significantly higher economic living conditions (Caldwell 1968). Caldwell determined that this phenomenon occurred because migrants send money back to the rural areas; another study which examined the income of migrants suggested that the reason


116

migrants tend to have higher incomes could be due to the fact that those with lower incomes cannot afford to migrate (Price 1971). Additionally, although the previously mentioned relationship between youth and migration exists, the population age structure in urban areas should not become overly youthful due to other migration and growth trends (Caldwell 1968). Another study which analyzed the determinants of migration took place in the southern United States. Although many determinants were similar, this study also considered the reasons for which migrants would return to the rural areas. Some that were interviewed cited that they would prefer to raise families in the rural area (Price 1971). Others returned once they were able to find work in the rural area, an important reason due to the fact that their reason for migrating was a perceived increase of economic well-being of those living in urban areas (Price 1971). Additionally, this study placed a greater importance on economic condition as a determinant for migration. In order to analyze the determinants, those who were interviewed were classified by ethnic group. Not only were those in the ethnic group which had the lowest levels of economic conditions more likely to migrate, but they were also the least likely to return (Price 1971). Another important determinate considered in this study was whether migrants had already acquired jobs in the urban areas prior to migration. Although many moved to the urban areas because in order to obtain jobs, very few were certain of employment: the ethnic group with the highest percentage who had already obtained employment was only 35% (Price 1971). However, most of those who migrated were able to find jobs after only a week, and no long term unemployment persisted (Price 1971). Furthermore, the increases in economic well-being among urban residents were very large. For the very poorest ethnic group, the median income actually tripled (Price 1971). Though the price of living in urban areas was higher and the living conditions similar, most of the migrants reported that they were financially better off (Price 1971). Most of the gains made were by the poorest ethnic groups, who did have better housing, clothing, food and modern amenities than in the rural areas (Price 1971). Rural Living Conditions and Migration Although rural-urban migration typically occurs due to an expected increase of economic well-being in urban areas, increasing well-being in rural areas does not necessarily decrease the propensity to migrate. For example, as agricultural technology increases, those who own the land become richer, and the agricultural laborers are displaced by capital: thus inequalities increase (Rhoda 1983). Additionally, as increased technology causes prices to fall,

those who own small farms and do not have equal access to credit will be forced to stop producing (Rhoda 1983). Public works projects often yield similar results. Though building roads may provide jobs for the rural laborers, the increased transportation only makes farming more profitable for the large landowner (Rhoda 1983). Furthermore, the increased transportation makes migration to urban areas easier (Rhoda 1983). Thus, these improvements to rural living conditions have the potential to increase both inequalities and rural-urban migration, which are the very results the improvements were meant to diminish. Another problem with increased services to rural areas is that these services are not empirically related to a decrease in rural-urban migration produced by an increase in economic well-being (Rhoda 1983). Higher quality rural education is an example of a completely unintended consequence: the propensity to migrate from a rural area to an urban one is highly correlated with higher levels of education (Rhoda 1983). This trend probably occurs because of the limited opportunity for highly skilled occupations in rural areas (Detang-Dessendre et al., 2002) Other services provided to rural areas tend to be water, electrification, family planning and healthcare. Although these clearly increase the living conditions of rural areas, the effects which these services have on rural-urban migration are indeterminate or mixed (Rhoda 1983). However, it has been suggested that if access to these services is limited in rural areas, migration is more likely (Detang-Dessendre et al., 2002). However, it is has been suggested that once rural migrants arrive in the cities, they do not have equal access to resources. Typically, they find employment in jobs which those previously living in the urban areas find inferior (Jin et al. 2000). This could be due to the fact that in this study, many of the migrants were a “floating population.” These migrants only stayed in the cities temporarily, decreasing their personal risk by having the option to easily return to the rural area (Jin et al. 2000). These migrants were often denied access to subsidized housing, medical care, or schooling (Jin et al. 2000). However, urbanization persists. The urbanization may continue because while rural incomes are rising rapidly, they are still very low compared to urban incomes: this is consistent with the findings of both Rhoda and Todaro (Harris and Todaro, 1970; Rhoda 1983, Jin et al. 2000). Furthermore, even if migrants have unequal access to resources or are placed in inferior jobs, many are still more satisfied with their situation than in the rural areas (De Jong, Chamratrithirong and Tran, 2002). If workers are placed into jobs in the industrial sector rather than service or agriculture, this attribute is more likely (De Jong, Chamratrithirong and Tran, 2002). However, those who are highly educated or have a large number of contacts in the rural area may be


117

satisfied only with their employment, not with the living conditions (De Jong, Chamratrithirong and Tran, 2002). This finding supports both Rhoda and Caldwell. Migration and Growth

As previously mentioned, rural-urban migration takes place as rural residents of developing nations move to urban areas in order to increase economic well-being. Several studies have considered the effects which migration may have on growth. Despite expectations of better economic conditions in urban areas and the evidence presented by Price, areas with high levels of urbanization tend to have high levels of unemployment and underemployment. This was examined at length by Harris and Todaro, whose model predicts that people will migrate to the area with the highest expected income (Harris, Todaro 1970; Bencivenga and Smith 1997). Because urban wages are higher than rural wages, unemployment and underemployment result when rural-urban incomes are in equilibrium, assuming a fixed wage in the urban sector (Harris, Todaro 1970; Bencivenga and Smith 1997). However, Bencivenga and Smith cite that this is not the historically accurate case, as this phenomenon occurred in the absence of minimum wage legislation and labor unions, which were assumed to be keeping the urban wage at a certain level (1997). Bencivenga and Smith’s observations on the effects of rural-urban migration upon economic growth use a dual-economy model which also allows for an informal, low-wage production sector (1997). According to this model, rural-urban migration is cyclical, with rural-urban migration occurring when economic growth occurs and reverse migration occurring during recessions and downturns (Bencivenga and Smith 1997). While the findings concerning equilibrium between rural-urban wages and unemployment are similar to those of Harris and Todaro, this equilibrium fluctuates, causing a “development trap” to form (Bencivenga and Smith 1997). Additionally, the relationship between high urban wages, lower urban wages and unemployment persists not because of an urban wage that has been fixed at a high level, but because of high levels of capital stock that accumulate when in steady state equilibrium (Bencivenga and Smith 1997). Another issue that has been studied in conjunction with migration is the effect of remittances, or “migradollars” upon economic growth (Arango et al. 1996). One study which observed the effects of international migration, rather than rural-urban, found that remittances were beneficial to a nation’s economic growth, and often helped to offset balance of payments deficits (Arango et al. 1996). An important effect of these remittances is the increased capital they allow the residents of the receiving country to obtain, which may help the country to overcome barriers to growth (Arango et

al. 1996). Additionally, for every migradollar that is remitted, the individual household that receives more than one dollar in direct benefits, and other households receive a partial benefit (Arango et al. 1996). It is possible that this effect of migradollars could be applied to rural-urban migration instead of international migration: Remittances are sent from urban areas to rural areas, just as they are sent from one nation to another (Price 1971). Also, as Bencivenga and Smith cited, capital accumulation in urban areas is greater than in rural ones, implying that availability of capital is increased through rural-urban migration (1997). A study by Conway and Cohen confirms this theory (1998). According to their research, recipients of remittances, especially those who were well off, “hoarded cash, began savings accounts, or invested their remittances in land purchase and/or development (Conway, Cohen 1998).”

DATA AND METHODS The data for this study represents several countries in Latin America and Sub-Saharan Africa from 1990 – present, and was obtained from the World Development Indicators and the Penn World Tables. The method used to analyze the data will be multiple regression. The goal will be to evaluate the estimates for the following equations:

β 1

(2)

3 The three dependent variables, income, life expectancy, and gross enrollment are the three variables which are used to create the human development index. These were split up into three separate equations in order to observe the effects of urbanization on development rather than GDP growth alone, as previous studies have done. To represent urbanization, the independent variables chosen are intended to represent urban population growth, the amount at which the urban population is growing relative to the rest of the population, and the urban population size. Because these three variables may be highly associated, the correlation between the variables will be analyzed. If the


118

variables are highly associated, the regression will be run with them separately in order to observe what effects their correlation had upon the results. The choice to use all of these variables together was made for a number of reasons. Foremost, the goal of this study is to see how the growth of urban areas effects human development. Thus, the most logical variable to use would be urban growth. However, if urban areas are growing at a similar rate to the rest of the population, the results of that analysis could be questionable. For this reason, the urban growth relative to the population growth was included. Finally urban population size is included because even if the urban population is growing, the actual proportion of the population living in urban areas may still be very small, implying that the area is not urbanized. Other variables included are based upon migration determinants or are used as control variables. Determinants are included because if they are present, urbanization is more likely to be occurring. Because age is a major determinant for migration, it is included as a variable (Caldwell, 1968; Detang-Dessendre et al., 2002). Other determinants included are access to an improved water source (clean, safe water) and access to an improved sanitation facility (modern restrooms) in urban areas relative to urban. Recall that many who move from rural areas tend to have less satisfaction with access to services once they reach urban areas (De Jong, Chamratrithirong and Tran, 2002). Thus, if urban areas actually have better access to services, or at least these basic ones, those who migrate there may be satisfied not only with their employment, but also with their quality of life in general (De Jong, Chamratrithirong and Tran, 2002). Because this study is focusing on more than income alone, including a measure such as this is vital. The control variables used were the following: investment and education for income and GDP per capita PPP for both education and life expectancy.

EMPIRICAL RESULTS

Latin America The models for Latin America were slightly better than those for Africa, with r squared being higher and more independent variables being statistically significant (all results can be seen in the statistical tables attached). While many results were those that fit with the proposed hypotheses, some were unexpected. On the life expectancy model, urban population size and the difference in sanitation were both statistically significant. Thus, as the urban population size increases, life expectancy does increase. This is probably due to the higher incomes that are received as the urban population size increases: urban population size and GDP per capita PPP also have a statistically significant positive relationship. Recall that the variable for sanitation difference represents a disparity

between rural and urban living conditions. Because this has an inverse relationship, the result is as expected: as disparities decrease, life expectancy increases. Some of the results were different from those expected, however. Foremost, the control variable, GDP per capita PPP, was statistically significant at the .01 level alone, but lost significance when the other variables were added. If the correlation between the independent variables is examined, GDP is correlated to both urban population size and a youthful population, but the association was not large enough that the variables were run separately. The income model yielded similar results: the relationship between urban population size and income was already mentioned. There were some differences, however. Foremost, the difference in sanitation was not significant, as it is in both other models, and only approached significance in the final model. However, the difference in access to clean water was significant, and this variable measures a similar concept. Also, having a youthful population was significant. This relationship may have been present because young migrants come seeking employment opportunities, and tend not to have children, who would receive the improved education and healthcare necessary for the other dependent variables to be altered. An unexpected result in this model was that the control variables never had a significant relationship. An explanation for this anomaly may be that a lag should have been used; education and investment were used as the control variables, and increasing these does not instantaneously create income. For the education model, results were not as consistent as in the life expectancy or income models. While the difference in sanitation, the difference in access to clean water, and the presence of a youthful population were significant in the model which excluded associated variables, all but sanitation difference lost significance in the final model. This result is probably due to the correlations between urban population growth and excess urban growth altering the results of the model. Furthermore, the urban population size approached significance in the final model, but had not been significant in the previous model. Interestingly, urban population growth, which is slightly associated to urban population size, was significant when added to the final model. Perhaps the association between the two variables caused the correlation between urban population size and enrollment to increase. Examining the unexpected results, it is of note that the sanitation difference had a positive relationship with gross enrollment. A possibility is that because this disparity is very low, access to other services, such as education, may have a low difference as well. Also, as in the model for income, the control variable for this model, GDP, lacked significance again. Once more, using a lag may fix this recurring problem in future research.


119


120

Africa The results for the African region, while fairly consistent, were less accurate models than those for the Latin American region. The explanation for this result may be that while both the African region and the Latin American region are experiencing higher urban growth than population growth, the urbanization in Latin America is more progressed than it is in Africa. Note the descriptive statistics for both regions: while urban population growth and excess urban growth are far higher in Africa, Latin America’s proportion of people living in urban areas is higher. This implies that Latin America has mostly undergone their period of urbanization, and the data reflects the effects of the urbanization. Africa, however, is still becoming urban, meaning that changes may just be beginning to take place. For the life expectancy model, only urban population size was significant. As mentioned in the discussion of Latin America’s results, this is an expected relationship. Additionally, the difference in access to clean water approached significance: this is consistent with the Latin American models, which found the variables measuring disparity in resources to be significant. However, the difference in sanitation facilities was not even close to being significant, possibly due to the fact that urbanization has not progressed as far in this region. Interestingly, the control variable for this model lost significance when other variables were added for the Latin American model, but did not in this one. Possibly the smaller associations between the independent variables allowed for model to have fewer flaws. The education model had also had far different results than the one for Latin America. Foremost, recall that education and a disparity in sanitation facilities had a positive relationship in that model; in this one, it had a significant negative relationship. While this result is more expected, it is not consistent with the previous finding. A reason may be that the disparities are much higher in this region, following the same logic used to explain the previous positive relationship.

CONCLUSIONS AND RECOMMENDATIONS

A fairly consistent result throughout the models was that having a large proportion of the population living in urban areas lead to higher levels of development. Furthermore, the disparities between access to either improved sanitation facilities or access to clean water was significant in several of the models. Having a young population, while not significant in all models, was highly significant for income. Because some of the other significant independent variables were correlated with income, this result is important as well.

Based upon these results, my policy recommendations would be as follows. Foremost, it is of great importance to try to reduce disparities in access to services. This would imply that the government should invest in trying to provide these services to areas which least access. Additionally, incentives should be provided to those who wish to move to urban areas, especially those who are young and seeking improved employment opportunities. However, this policy should only be pursued until a certain point, as examining the relationship between urban population size and development suggests diminishing returns. That is, as the proportion of the population living in urban areas increases, development increases, but at a slower rate. Thus, the benefits of providing these incentives must be considered before they are provided. In the future, it would be interesting to analyze data from more regions, and possibly see the general implications which the relationships yield. Also, this study was limited by the availability of data for certain variables; some that may have increased the accuracy of the models had to be omitted. Perhaps this data would be more available for other regions, thus improving the models and the results obtained from them. Furthermore, the potential diminishing returns of urbanization could be analyzed, and closer analysis of the all the relationships could be examined. If these results held for several regions, they could possibly be seen as generally applicable for policy recommendations.

Table1 Regression Results:

Latin America Model 1: Control Variables Model 2: UrbPopSize, YouthPop, Dif Model 3: All Variables

Independent Variable

LifeExpect AvEnroll LogGDP LifeExpect AvEnroll LogGDP LifeExpect AvEnroll LogGDP

Intercept 20.0803 (0.0628)*

72.4739 (0.1025)

3.6401 0.0000***

44.6054 (0.0013)***

170.2929 (0.0014)***

3.7415 (0.0000)***

47.9698 (0.0007)***

118.8513 (0.0113)***

3.7018 (0.0000)***

UrbPopGrow -1.3591 (0.1703)

-7.8358 (0.0282)**

-0.0586 (0.3236)

OnlyUrbGrow 2.9922 (0.1307)

3.2861 (0.5863)

0.1490 (0.1266)

UrbPopSize 0.1750 (0.0012)***

0.2829 (0.1720)

0.0097 (0.0000)***

0.1814 (0.0009)***

0.3358 (0.0661)*

0.0098 (0.0001)***

WaterDif 0.0880 (0.4331)

-0.8748 (0.0011)***

-0.0118 (0.0003)***

0.1556 (0.1993)

-0.3686 (0.1533)

-0.0122 (0.0004)***

SanDif -0.4382 (0.0205)**

0.7785 (0.0000)***

0.0036 (0.1891)

-0.5433 (0.0086)***

0.3966 (0.0309)**

0.0048 (0.0896)*

YouthPop 0.0390 (0.4239)

-0.4787 (0.0064)***

-0.0059 (0.0283)**

0.0736 (0.2260)

0.0831 (0.7124)

-0.0050 (0.1270)

LogGDP 13.2566 (0.0000)***

5.1116 (0.6677)

4.1723 (0.2548)

-22.1616 (0.1214)

2.9308 (0.4313)

-11.2389 (0.3653)

GDPInvest 0.0063 (0.1997)

-0.0037 (0.3961)

-0.0046 (0.2876)

AvEnroll -0.0009 (0.7707)

-0.0024 (0.3765)

-0.0023 (0.4789)

R Squared 0.3234 0.0057 0.0562 0.5662 0.5026 0.6088 0.5918 0.6643 0.6474 n 46 35 35 46 35 35 46 35 35


121

Table 2

Regression Results: Africa

Model 1: Control Variables Model 2: UrbPopSize, YouthPop, Dif Model 3: All Variables

Independent Variable

LifeExpect AvEnroll LogGDP LifeExpect AvEnroll LogGDP LifeExpect AvEnroll LogGDP

Intercept 5.1359 (0.2880)

47.5433 (0.2097)

2.6214 (0.0000)***

-6.9203 (0.4727)

100.5489 (0.0260)**

3.8999 (0.0000)***

-11.3568 (0.2468)

79.3981 (0.0800)*

3.9068 (0.0000)***

UrbPopGrow 0.9196 (0.0460)**

5.3252 (0.0872)*

-0.0056 (0.9127)

OnlyUrbGrow -0.6726 (0.2110)

-3.4492 (0.3823)

0.0161 (0.7791)

UrbPopSize 0.0826 (0.0678)*

0.5488 (0.0546)**

-0.0047 (0.2072)

0.0900 (0.0442)**

0.3252 (0.3089)

-0.0048 (0.2179)

WaterDif 0.0329 (0.2288)

0.1855 (0.2801)

-0.0038 (0.0610)*

0.0487 (0.0833)*

0.2441 (0.1611)

-0.0039 (0.0726)*

SanDif 0.0095 (0.8058)

-1.3098 (0.0000)***

-0.0071 (0.1221)

0.0029 (0.9392)

-1.3405 (0.0000)***

-0.0071 (0.1524)

YouthPop 0.0210 (0.5880)

-0.0549 (0.8062)

-0.0091 (0.0054)**

0.0102 (0.7914)

-0.1614 (0.4800)

-0.0090 (0.0094)***

LogGDP 15.6486 (0.0000)***

1.2286 (0.9251)

18.1838 (0.0000)***

-13.4972 (0.2564)

18.6474 (0.0000)***

-9.2166 (0.4376)

AvEnroll 0.0031 (0.1484)

-0.0009 (0.7028)

-0.0009 (0.6989)

GDPInvest 0.0144 (0.1879)

-0.0061 (0.5797)

-0.0063 (0.5860)

R Squared 0.6496 0.0003 0.1081 0.6853 0.5000 0.5529

0.7148 0.5544 0.5547

n 49 34 33 49 34 33 49 34 33 * α = .10, ** α = .05, *** α = .01


122

Table 3

Descriptive Statistics: Africa Life Expectancy

LifeExpect UrbPopGrow OnlyUrbGrow UrbPopSize YouthPop LogGDP WaterDif SanDif Mean 49.6124 4.6455 1.7541 25.7122 57.8245 2.8422 39.2245 22.9184 Standard Error 0.6117 0.1554 0.1307 1.7804 1.9650 0.0315 2.5338 1.7605 Median 49.8590 4.4636 1.5965 22.3000 59.1000 2.8553 38.0000 22.0000 Standard Deviation 4.2816 1.0881 0.9152 12.4630 13.7549 0.2205 17.7368 12.3235 Sample Variance 18.3323 1.1840 0.8376 155.3273 189.1981 0.0486 314.5944 151.8682 Range 17.7340 6.2008 4.3050 47.0000 50.9000 1.0980 72.0000 51.0000 Minimum 41.6469 1.1336 -0.1438 11.1000 26.4000 2.1352 -2.0000 1.0000 Maximum 59.3809 7.3344 4.1612 58.1000 77.3000 3.2332 70.0000 52.0000 Sum 2431.0071 227.6304 85.9523 1259.9000 2833.4000 139.2677 1922.0000 1123.0000

Gross Enrollment AvEnroll UrbPopGrow OnlyUrbGrow UrbPopSize YouthPop WaterDif SanDif LogGDP Mean 51.0518 4.6899 1.6118 26.8818 56.1706 37.5294 22.5882 2.8557 Standard Error 3.0570 0.1960 0.1267 2.2102 2.3683 3.1767 2.0350 0.0417 Median 48.7619 4.4638 1.5714 23.2500 57.3000 37.5000 23.0000 2.8852 Standard Deviation 17.8250 1.1426 0.7391 12.8875 13.8092 18.5231 11.8657 0.2431 Sample Variance 317.7302 1.3056 0.5462 166.0885 190.6930 343.1052 140.7950 0.0591 Range 67.5839 6.0022 2.5716 42.7200 48.1000 72.0000 47.0000 1.0980 Minimum 18.9702 3.0391 0.4951 12.0200 26.4000 -2.0000 1.0000 2.1352 Maximum 86.5540 9.0413 3.0667 54.7400 74.5000 70.0000 48.0000 3.2332 Sum 1735.7627 159.4559 54.8022 913.9800 1909.8000 1276.0000 768.0000 97.0940


123

GDP Per Capita PPP LogGDP UrbPopGrow OnlyUrbGrow UrbPopSize YouthPop WaterDif SanDif AvEnroll InvestGDP Mean 2.8782 4.6128 1.6353 26.4679 56.9364 37.0303 21.6667 51.4620 6.8760 Standard Error 0.0373 0.1614 0.1355 2.2995 2.2960 3.2226 1.9698 3.1226 0.6069 Median 2.8822 4.3825 1.5134 23.4000 58.5000 37.0000 20.0000 52.1393 6.3228 Standard Deviation 0.2142 0.9273 0.7785 13.2098 13.1898 18.5126 11.3155 17.9377 3.4862 Sample Variance 0.0459 0.8598 0.6061 174.4976 173.9699 342.7178 128.0417 321.7608 12.1538 Range 0.8271 4.0514 3.4734 46.4000 47.5000 72.0000 41.0000 67.5839 15.6481 Minimum 2.4061 3.2830 0.4975 11.7000 26.4000 -2.0000 1.0000 18.9702 3.0824 Maximum 3.2332 7.3344 3.9709 58.1000 73.9000 70.0000 42.0000 86.5540 18.7305 Sum 94.9802 152.2236 53.9633 873.4400 1878.9000 1222.0000 715.0000 1698.2459 226.9067

Table 4 Descriptive Statistics: Latin America

Life Expectancy LifeExpect UrbPopGrow OnlyUrbGrow UrbPopSize YouthPop LogGDP WaterDif SanDif Mean 68.2814 2.3324 0.7458 56.2761 41.4022 3.6360 21.5217 11.0652 Standard Error 0.5056 0.1601 0.0715 1.7852 1.2601 0.0217 1.7958 1.0185 Median 69.0526 2.6438 0.7651 57.5000 42.1000 3.6841 22.5000 11.2500 Standard Deviation 3.4288 1.0858 0.4850 12.1075 8.5464 0.1471 12.1797 6.9078 Sample Variance 11.7570 1.1790 0.2353 146.5912 73.0411 0.0216 148.3440 47.7179 Range 15.7307 4.5534 2.2006 45.7000 40.5000 0.5502 48.0000 35.0000 Minimum 58.9393 -0.4190 -0.3490 28.2000 18.1000 3.3008 0.0000 0.0000 Maximum 74.6700 4.1344 1.8516 73.9000 58.6000 3.8510 48.0000 35.0000 Sum 3140.9456 107.2891 34.3071 2588.7000 1904.5000 167.2571 990.0000 509.0000


124


125

Gross Enrollment AvEnroll UrbPopGrow OnlyUrbGrow UrbPopSize YouthPop WaterDif SanDif LogGDP

Mean 91.1552 2.0166 0.6536 58.0680 41.1400 19.6000 28.6000 3.6547 Standard Error 1.6889 0.1810 0.0838 2.1540 1.6116 1.8792 1.9269 0.0248 Median 91.0033 1.9052 0.7140 59.8000 42.3000 23.0000 26.0000 3.6973 Standard Deviation 9.9920 1.0708 0.4956 12.7433 9.5341 11.1176 11.3997 0.1470 Sample Variance 99.8392 1.1466 0.2456 162.3906 90.8984 123.6000 129.9529 0.0216 Range 46.9737 3.9299 1.8908 46.0400 40.5000 38.0000 41.0000 0.5502 Minimum 67.2234 -0.5354 -0.3490 28.2000 18.1000 0.0000 8.0000 3.3008 Maximum 114.1971 3.3944 1.5417 74.2400 58.6000 38.0000 49.0000 3.8510 Sum 3190.4325 70.5814 22.8759 2032.3800 1439.9000 686.0000 1001.0000 127.9136

GDP Per Capita PPP LogGDP UrbPopGrow OnlyUrbGrow UrbPopSize YouthPop WaterDif SanDif AvEnroll InvestGDP Mean 3.6547 2.1432 0.6833 57.2029 40.9029 19.6000 28.6000 91.1552 14.7503 Standard Error 0.0248 0.1808 0.0822 2.1385 1.5583 1.8792 1.9269 1.6889 1.0272 Median 3.6973 2.2213 0.6774 58.6400 42.1000 23.0000 26.0000 91.0033 15.1215 Standard Deviation 0.1470 1.0697 0.4865 12.6513 9.2188 11.1176 11.3997 9.9920 6.0771 Sample Variance 0.0216 1.1444 0.2367 160.0562 84.9862 123.6000 129.9529 99.8392 36.9311 Range 0.5502 4.3804 2.1744 45.2600 40.5000 38.0000 41.0000 46.9737 24.6376 Minimum 3.3008 -0.5354 -0.3478 28.2800 18.1000 0.0000 8.0000 67.2234 5.9718 Maximum 3.8510 3.8450 1.8265 73.5400 58.6000 38.0000 49.0000 114.1971 30.6095 Sum 127.9136 75.0132 23.9144 2002.1000 1431.6000 686.0000 1001.0000 3190.4325 516.2596

REFERENCES

Arango, J., et al. 1996. "International migration and national development." Population Index 62(2): 181-212.

Bencivenga, V. R. and B. D. Smith. 1997. "Unemployment, migration and growth." The Journal of Political Economy 105(3): 582-608.

Caldwell, J.C. "Determinants of rural-urban migration in Ghana." 1968. Population Studies 22(3): 361-377.

Cohen, J. H. and D. Conway. 1998. "Consequences of migration and remittances for Mexican transnational communities." Economic Geography 74(1): 26-44.

De Jong, G., A. Chamratrithirong and Q-G Tran. "For better, for worse: Life satisfaction consequences of migration." 2002. International Migration Review 838-863.

Detang-Dessendre, C., et al. 2002. "Life cycle variability in the microeconomic determinants of rural-urban migration." Population (English Edition) 31-56.

Harris, J. R. and M. P. Todaro. 1970. "Migration, unemployment and development: A two-sector analysis." The American Economic Review 60(1): 126-142.

Heston, A., R. Summers and B. Aten. "Penn World Tables 6.2." September 2006. Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania. 27 March 2009 <http://pwt.econ.upenn.edu/php_site/pwt_index.php>.

Jin, Z., M. C. Seeborg and Y. Zhu. 2000. "The new rural-urban labor mobility in China: Causes and implications." The Journal of Socio-Economics 39-56.

Price, D. O. 1971. "Rural to urban migration of Mexican Americans, Negroes and Anglos." International Migration Review 5(3): 281-291.

Rhoda, R. 1983. "Rural development and urban migration: Can we keep them down on the farm?" International Migration Review 17(1): 34-64.

The World Bank. World Development Indicators 2008.


POSSIBLE RELATIONSHIPS BETWEEN THE THEORY OF GLOBAL WARMING AND STOCK MARKET DECLINES

David Nugent

School of Business Slippery Rock University Slippery Rock, PA 16057

ABSTRACT This paper presents a theoretical discussion of the possibility that part of the recent decline in stock market value may be due to investors’ perceptions that politicians may take draconian action to prevent global warming. Finance theory (Block and Hirt, 2005; Ross, Westerfield and Jordan, 2006) suggests that corporate stock value is affected by investors’ expectations of economic growth. The theory of global warming (Brown, 2007; Ochoa, Hoffman and Tin, 2005) suggests that the production of carbon dioxide and other greenhouse gases causes global temperatures to rise, and that the remedy would be the reduction of the production of greenhouse gases. If investors’ were to perceive that action to reduce greenhouse gases would also reduce economic growth, the result could be a decline in stock market value.

INTRODUCTION Recent declines in stock market value are probably due to a number of causes. Turmoil related to the bursting of the housing bubble, the collapse of financial institutions and widely fluctuating oil prices may be largely to blame. This paper suggests that there may be other contributory factors. The theory of global warming may be one of those factors. This paper addresses the potential role of the consequences of actions that may be taken to address the possibility that emissions of greenhouse gases increases global temperatures and how those actions may cause economic harm. The issue to be addressed is not whether the theory of global warming is correct or whether higher temperatures would affect economic activity. Rather, the issue is whether actions taken by politicians who believe the theory to be true will adversely affect economic activity and corporate stock values.

THEORY OF GLOBAL WARMING According to the theory of global warming (Brown, 2007; Ochoa, Hoffman and Tin, 2005) sunlight heats the surface of the Earth. In a manner comparable to the glass in a greenhouse retaining heat inside a greenhouse, the Earth’s

atmosphere reduces the radiation of heat into space. Certain gasses, called greenhouse gases, are theorized to be largely responsible for the retention of the Sun’s heart. These greenhouse gases include carbon dioxide, methane, nitrous oxide, water vapor and chlorofluorocarbons. It is suggested that increases in greenhouse gasses can cause average global temperatures to rise. Carbon dioxide in particular is theorized to pose a threat of global warming. When hydrocarbons (such as oil, coal, natural gas and wood) are burned, carbon combines with oxygen to form carbon dioxide. It is suggested that during the period since the industrial revolution, the burning of various hydrocarbons has caused the level of carbon dioxide in the atmosphere to increase significantly. For example, Brown (2007) states that between the beginning of the industrial revolution and 2007, concentrations of carbon dioxide in the atmosphere increased from 280 parts per million to 383 parts per million. It is suggested that if concentrations of carbon dioxide continue to increase, average global temperatures will rise. It is further suggested that to retard global warming, carbon dioxide emissions should be reduced.

ECONOMIC CONSEQUENCES OF EMISSION REDUCTIONS

Since a large proportion of economic activity is dependent, either directly or indirectly, on the burning of fuel, it seems likely that a drastic reduction in carbon dioxide emissions would result in a substantial reduction in economic output. If the advice to reduce carbon dioxide emission were taken to the absurd extreme of eliminating all burning, the result could be the reduction of economic activity to levels that existed before the discovery of fire. Hydroelectric power, nuclear power, wind turbines, solar power and other alternate sources of electricity could provide some level of economic activity. In the long run, such alternate energy sources may provide for a moderate level of economic activity. In the short run, however, the level of economic activity that could be produced by alternate energy sources would be a small fraction of current output.


It seems likely that voters would not tolerate the drop in standard of living that would be necessary to achieve zero emissions. Perhaps what is more likely would be a compromise that would limit emissions. Suppose that as a compromise, a law were passed that would limit carbon dioxide emissions to current levels. If a company’s growth were dependent on the use of energy, its inability to increase its supply of energy could result in a cessation of growth. If some companies were able to increase their energy use by purchasing other companies’ allocations, as might arise in a “cap and trade” system, those purchasers might be able to increase their output. However, it seems likely that the sellers of energy allocations would experience economic contraction. For the economy as a whole, the effect of emission limits could be great impediments to growth. For some proponents of emissions control, limiting emission to current level would not be enough. Continuation of current levels of emissions means that, theoretically, concentrations of carbon dioxide and global temperatures would continue to rise, albeit at a slower rate than if emissions were to accelerate. Even a substantial reduction in emissions may be inadequate. For example, Levy (2007, p 163) states “The reality remains that cutting emissions by half will only double the time for the damaging effects to take place; it will not reverse the trend. What is needed is a global commitment to a sea change that has as its goal a zero tolerance toward damaging emissions”. This seems to suggest that politicians should literally seek to reduce emissions of carbon dioxide and other greenhouse gases to zero.

POTENTIAL REACTIONS OF INVESTORS AND CHANGES IN STOCK VALUES

To investors observing the global warming debate, plausible outcomes may seem to range from moderate emission limits that would induce perpetual economic stagnation to tremendous economic decline resulting from draconian reductions in emissions. This range of potential outcomes could give rise to a range of reactions and a range of effects on stock market values. According to finance theory (Block and Hirt, 2005; Ross, Westerfield and Jordan, 2006), a corporate stock’s value can be calculated as the present value of future cash flows, typically in the form of cash dividends. If a company’s sales, profits and dividends were expected to grow, an investor would expect the stock value to grow as well. Such an investor would expect a rate of return that is the sum of 2 components: (1) a dividend yield and (2) a capital gain yield. For example, if expected dividend yield were

2% and expected capital gain yield were 8%, total expected rate of return would be 10%. To illustrate the calculation of a stock’s market value, consider the constant growth valuation model presented by Block and Hirt (2005, p 283). If a company’s dividend is expected to grow at a constant rate, the future value of future dividends can be expressed as the following formula: Po = D1 / Ke – g (1) Symbols are defined as: Po is the market price of the stock today. D1 is the dividend expected at the end of the coming year. Ke is the required total rate of return. g is the constant growth rate for dividends. To illustrate potential changes in stock values, consider first a hypothetical set of dividend, rate of return and growth rate figures that might seem reasonable in the absence of government restrictions on emissions. Suppose that the expected dividend (D1) were $1.00, the required total rate of return (Ke) were 10% and the expected growth rate (g) were 8%. The market price of the stock (P0) would be calculated as: P0 = D1 / Ke – g

= $1.00 / (.10 - .08) = $1.00 / .02 = $50.00 (2)

Consider next a scenario that might arise if politicians were to impose laws that would reduce emission close to zero. Suppose that economic output and dividends were to fall by 80%. In that case the expected dividend (D1) would fall from $1.00 to $0.20. Further suppose that the growth rate (g) were to fall to zero, while required total rate of return (Ke) remains at 10% The market price of the stock (Po) would be calculated as:

P0 = $0.20 / (.10 – 0) = $0.20 / .10 = $2.00 (3)

If politicians were to perceive that the preceding magnitude of economic decline would not be politically feasible, some compromise8 action might be taken instead. Suppose that such a compromise would result in economic output remaining the same, expected dividend (D1) remaining at $1.00, but growth rate (g) reduced to zero. Further suppose that required total rate of return (Ke) remains at 10%. The market price of the stock (P0) would be calculated as: P0 = $1.00 / (.10 – 0) = $1.00 / .10 = $10.00

(4)



If expectations of zero economic growth were also politically unfeasible, politicians might seek a compromise that would allow a modest rate of economic growth. Suppose that the growth rate in dividends (g) were to fall from 8% to 4%, while expected dividend (D1) were to remain at $1.00 and required total rate of return (Ke) were to remain at 10%. The market price of the stock (P0) would be calculated as: P0 = $1.00 / (.10 - .04) = $1.00 / .06 = $16.67

(5)

CONCLUSIONS Although recent stock market declines are probably due mainly to economic turmoil related to the bursting of the housing bubble, the collapse of financial institutions, widely fluctuating oil prices and other aspects of the recession, it seems plausible that part of the decline may be attributable to investors’ perceptions that politicians may impose greenhouse gas emission limits. If emission limits were perceived to have adverse economic consequences, the result could be declines in corporate stock values. The preceding calculations suggest that if stock value is dependent on expectations of growth in dividends, even a moderate decline in expectations of economic growth could result in a substantial decline in stock value. If actions taken to curb the perceived threat of global warming were to cause economic output to fall, the result could be a tremendous decline in corporate stock values. Stock market decline has the potential to affect large segments of the population. In addition to individuals’ investment portfolios, many others are affected by declines in pension fund assets. Retirement plans and the standard of living of retirees may be adversely affected by declines in stock values. Politicians may want to seek a balance between potential economic harm and perceived benefits of reducing emissions of greenhouse gases.

REFERENCES

Block, Stanley B., and Geoffrey A. Hirt. 2005, Foundations of Financial Management New York: McGraw-Hill Irwin. Brown, Paul, 2007. Global Warning: The Last Chance for Change Pleasantville: The Reader’s Digest Association.

Levy, Matthys, 2007. Why the Wind Blows: A History of Weather and Global Warming Hinesburg: Upper Access Book Publishers. Ochoa, George, Jennifer Hoffman, and Tina Tin, 2005. Climate: The Force That Shapes Our World and the Future of Life on Earth London: Rodale Books International. Ross, Stephen A., Randolph W. Westerfield, and Bradford D. Jordan, 2006. Fundamentals of Corporate Finance New York: McGraw-Hill Irwin.

ECONOMICAL VALUATION OF FORESTS: IRANIAN CASE STUDY

Sadegh B. Imandoust Department of Economics Payame Noor University

Mashhad, Iran

ABSTRACT One of the most important aspects of ecological economics is economic valuation. Natural resources such as forests and rangelands are the most important factors for human and environmental survive. Less and discounting on natural resource values has been caused that the actual value of it to be showed less. For this reason the broad objectives of this paper are to determine the Total Economic- Ecologic Values (TEEV) of forests and its different compounds such as Direct Use Value (DUV), Indirect Use Value (IUV), Option Value (OV) and Existence Value (XV) by the market and mean international prices in 2003, and make a comparative analysis of them with GNP. Results indicate that Total Economic- Ecologic Value of Iran’s forests is greater than the forestry sub-section value calculated by Central Bank. Thus, the application of Total Economic- Ecologic Value of forests in production and value added of the country for more investment and saving forests purposes is advised.

INTRODUCTION In the last quarter of the twentieth century, one of the major worries of the humankind has been deterioration in the quality of environment. Initially considered as a problem of industrial rich countries, environmental degradation is now recognized as an issue of concern in the developing world. Within the environmental agenda, loss of forests has been an important item as depletion of forests has been rather very fast, especially in the second half of the twentieth century. Apart from subsistence conversion, forests have yielded space to claims of developmental activities. During the last two decades Iran witnessed annual depletion of forest cover at a rate of 200 thousand hectares annually (Khazaei, 1997). The causes of environmental degradation, particularly forests, boil down to the market and policy failures. The reason behind these failures lies in the absence of proper road signals indicating the scarcity value of forests. As a number of benefits of forests accrue to the people, individually and collectively outside the market process, they failed to appreciate the true total economic-ecologic value of forests. The broad objectives of the paper are (a) to determine the total economic-ecologic value (TEEV) of the forests and its different components such as direct use value (DUV), indirect use value (IUV), option value (OV) and existence value (XV); and (b) to make a comparative analysis of Iran's Central Bank calculations about forests and TEEV.

BACKGROUND Heafle et.al (1992) studied the nonmarket benefits of protecting forest quality with using a Contingent Valuation Method in the southern Appalachian Mountains. They compared two willingness to pay question formats (discrete choice and payment card) and showed that there is a significant difference between them. Kling (1993) provides an empirical assessment of the magnitude of option values relative to expected surplus using a model presented by Larson and Flacco (1992). Results indicate that option values engendered by price and income uncertainty are generally quite a small percent of expected surplus. Clayton and Mendelson (1993) estimated use values associated with the opportunity to view grizzly bears on the Mc Neil River in Alaska, employing the contingent valuation method to estimate a willingness to pay that ranged from $ 227 to $ 277 per person for a four-day visit. Bello et. al (1997) analyze seven forest areas that selected on the basis of three main criteria: (i) the sustainability of the area for tourism; (ii) the actual flow of tourism; (iii) the regional plans for the creation of new parks. They estimated tourism value of Liguria Region of Italy as 77-85 dollars per hectare with Travel Cost Method. Anderson et. al (1997) compared the Total Economic Value of standing Amazonian rain forest with the Net Present Value of alternative agricultural land uses. Their finding showed that, at the current level of deforestation, the potential benefits of deforestation are higher than the expected costs. Learner and Poole (1999) estimated the defensive expenditures to protect water purification services of forested watershed of New Jersey forests as $ 55 million. Barnhill (1999) attributed a portion of the economic impact of tourism along the Blue Ridge Parkway to the scenic beauty of adjacent forests. The study found that visitors spend $ 1.3 billion in North Carolina and Virginia contiguous to the parkway, which these expenditures generate $ 98 million in tax revenues annually, and that visitor spending directly, supports more than 26500 jobs. Bogahawatte (1999) quantified and valued the NTFP obtained from the forests by the local communities in Seri Lanka. He estimated the NTFP value as 1-10 percent of forester's incomes. Loomis and Richardson (2000) calculated the Carbon Sequestration Values of 42 million acres of roadless area on U. S. national forests as $ 56/ton, $ 1 billion annually and $ 26.7 billion present value. Nayak (2001) studied the Economic- Ecologic values of Paralakhemundi forests in India with TEEV method and concluded that the total value of annual flow of productive and consumptive benefits, environmental services, option and existence preferences is far higher than government


130

revenue and estimated income, and offered the use of this new method (TEEV) for future studies. Anderson (2005) showed that even under uniform pricing, Ugandan’s profit from gorilla tracking in the Bwindi Impenetrable National Park alone could have been increased by between USD 30,000 and USD 220,000 (depending on assumptions about social costs) based on Data collected from a travel cost survey indicates that in 1997.

MATERIALS AND METHODS The TEEV of forests has been assessed for the forests all over the Iran. Iran's forests are geographically located at the North and West of it, has a forest area of 12.4 million hectares (Mesdaghi, 1995) and this much area was used for estimating the per hectare averages. The standards obtained from similar studies have been applied in the process wherever the data/information for country is not made available/quantified. The total economic-ecologic value (TEEV) includes the present use values, both direct and indirect, future use values or option values and non-use values. The components of TEEV however cannot be simply aggregative since trade-offs exist (Pearce and Moran, 1994) between different types of use values and non-use values. In practice the approach of TEEV should be used with caution. Nevertheless, exploring TEEV helps us in investigating the economic-ecologic valuation and formulating protective policy framework (Nayak, 2001). All the prices are based on the year 2003.

RESULTS A: Direct Use Value The components of the DUV of Iran's forests include timber, non-timber forest products (NTFPs) and recreation values. A-1: Timber The potential timber output of Iran's forests can be judged from the calculations of Central Bank of Iran. On the basis of Central Bank's Information which has been considered for estimation of the value of potential timber output, the possible wood production is 2870537 tons with the value of $ 3.75×10 8 ($ 0.37 billion) at 2003 (Jalayeri, 2004). A-2: Non-Timber Forest Products (i) Firewood

The foresters of Zagros region at the west forests of Iran, use timbers as firewood. Applying the parameters discussed for firewood, the sustainable production of firewood is estimated as 4.6 million cum (table 1). With the value of $ 22.4 per ton and change index of 0.75 (cum to

ton), the conventional market value of the potential firewood output of Iran's forests is worked out at $ 0.103 billion. (ii) NTFPs NTFPs are normally renewable resources. Iran's forests NTFPs are containing from Turpentine, Forest Pistachio, and Forest Walnut. These products are allowable products conditioned to owner's interest that the Forest & Rangeland Organization receives owner's interest from their users. Two of these products (Turpentine and Forest Pistachio) are for export uses too. Tables 2 and 3 show the values of NTFPs in Iran's forests. A-3: Forest Recreation Costanza Approach has estimated the recreational value of Iran’s forests. Costanza's adjustment is 0.558 (Khalilian, 1996) for Iran that by multiplying it at Costanza's estimated recreational prices for forests (66 dollar per hectare), the forest recreation value was result in $ 4.57×108 in 2003. Table 4 presents the value of recreation in Iran's forests in 2003 prices. Total Direct Use Value Table 5 sums up the values of the different components of DUV. According to this table, recreation comes out prominently as the most valuable item of Iran's forests. The Total Direct Use Value is estimated at 5770 billion Rials (Table5). B: Indirect Use Value Watershed protection, nutrient recycling, flood control, production of oxygen and pollution control, wildlife protection, soil conservation, carbon storage and so on are recognized as the ecological functions of a forest, the implicit values of which constitute the IUV. The researchers in this field have applied contingent valuation and ranking, averting behavior and replacement cost methods to determine this value. In the valuation exercise of the indirect uses of forests, the value of services such as Production of oxygen, Controlling of soil erosion and soil fertility, Recycling of water and controlling humidity and Sheltering of birds, squirrels, insects and plants are used. The value of ' Recycling of water and controlling humidity' is based on Jalayeri (2004) findings and the price of water per cum calculated by Ministry of Power as $ 0.0148 (Table 6). The value of Sheltering of birds, squirrels, insects and plants are based on Iran's Central Bank information in 2003 ($ 6.79×105). The other values of IUV (Controlling of soil erosion and soil fertility and Production of oxygen) are based on Costanza (1997) Approach, 96 and 88 dollars per



hectare, respectively that by multiplying them at the determined adjust index of 0.558 for Iran (Khalilian, 1996), their values are determined at $ 1.12×109 and $ 1.03×109 in 2003 prices respectively (Tables 7 and 8). Based on the results of tables 6-8 and Central Bank's Information, the IUV of Iran's forests are determined at $ 1.29×109 in 2003 prices. C: Future Use Value (FUV) or Option Value (OV) The part of the present use value that should be set aside to ensure sustainability has been stated (Chopra, 1993) as option value and can be assessed as

1)1(11

++−= nrV

U

(1)

The equation implies that the ratio of the usable income (U) to the present use value (V) is a function of the rate of discount and the assumed life of the forest (n). The present value (V) is nothing but the total use value, including both direct and indirect. If ‘V’ is to be maintained constant after the expected life,

1)1( ++ nrV

(2)

is the cost of preserving the forest which can be taken as the option value or the national value of the possible benefit obtainable from the preservation of forests in a situation where forests are considered as exhaustible (Chopra, 1993). Table 10 shows the OV of Iran's forests at various Scenarios of discount rate (r) and the life expectancy of the forests as 30 years. For example, in this formulation if the social discount rate (r) is taken as 5 per cent and the life expectancy of the forests as 30 years, then

OV amounts to 22 per cent of 'V'. Considering IUV at $ 6.81×1011, the 'V' has been estimated at $ 6.82×1011 and option value has been estimated at $ 4.895×108. D: Non-Use or Existence Value (XV) Considering similar surveys conducted for tropical forests, it has been assumed that the existence value of Iran's forests is equivalent to 'γ' per cent of the sum of the present use value and future use values. Since the existence value accrues to the global community it is not taken for granted for the assessment of the TEEV of the forests of Iran. The existence value and various discount rates (r) and various adjustment rate (γ) scenarios are showed at table 11. Total Economic-Ecologic Value Table 12 sums up the total economic-ecologic values of Iran's forests at various discount rates (r) and various adjustments rate (γ) scenarios. Findings showed that the Maximum and Minimum TEEV of Iran's forests are $ 5.42 and 4.12 billion respectively, that at comparison with Iran's GNP at 2003, they are 7.01 and 5.328 per cent respectively. Table 13 shows the various amounts of TEEV to GNP ratio percent at various discount rate (r) and adjustment rate (γ) scenarios.

CONCLUSIONS The Economic-Ecologic Values of Iran's forests calculated in this study were $ 4.12 billion, 5.328 per cent of Iran's GNP at the worst conditions (r=25 and γ=85 per cent); whereas the forestry sub-section production value calculated by Iran's Central Bank was 0.18 per cent of GNP at 2003. Because of valuing ecosystem services; option value and existence value, there is a big gap between Central Bank's calculations and finding results. Then, application of TEEV in production and value added of the country for more investment and saving forests is advised.

TABLES

Table 1. Basis for determination and value of firewood in Iran's forests Sr. No

Particulars Unit of Measurement Magnitude

1. Number of foresters ---------- 230000 2. Average demand for firewood per forester (Cum/year) 20 3. Annual demand for firewood by foresters (Cum/year) 4600000 4. Price of firewood (Dollars/ton) 22.4 5. Value of firewood (Dollars /ton) 1.03×10 8

Source: Jalayeri's Findings

Table 2. The value of received owner's interest for NTFPs Magnitude

(Dollar) Amount of Measurement (kg) Name of Product Sr.

No 89824.37 129128 Turpentine 1. 13857.73 25945 Forest Pistachio 2.

198.66 376 Forest Walnut 3. 103880.76 155449 Total Value 4.

Source: Forest & Rangeland Organization Statistics

Table 3. The value of export NTFPs Export Value

(Dollar)

Value Per Ton

(Dollar)

Number of Authorization

Amount of Measurement

(kg)

Name of Product Sr. No

92678 600 41 154463 Turpentine 1. 9360 3000 1 3120 Forest Walnut 2.

102038 3600 42 157583 Total 3. Source: Forest & Rangeland Organization Statistics

Table 4. Recreation values of Iran's forests

Magnitude Unit of Measurement Particulars Sr. No

66 Dollar/ hectare Recreation value of forests based on Costanza Approach

1.

0.558 ------------------ Costanza adjustment index for Iran 2. 12.4 Million hectare Iran's forest area 3.

4.57×108 Dollar Recreation value (1×2×3) 4. Source: Finding Results

Table 5. Total Direct Use Value of Iran's forests

Value (Dollar)

Product/ Provision

Sr. No

3.75×108 Timber 1. 1.03×108 Firewood 2. 2.06×105 NTFPs 3. 4.57×108 Recreation 4. 9.35×108 Total 5.

Source: Based on Tables 1 to 4.


Table 6. Basis for estimation recycling of water and controlling humidity Sr. No

Particulars Unit of Measurement

Value (Dollars)

1. Amount of storage by Iran's forests Cum 1.045×109 2. Value of store Dollar/Cum 0.0148 3. Value of recycling water and controlling humidity Dollar 1.55×107

Source: Jalayeri's Findings (row 1), Ministry of Power Information (row2) and Finding Results (row 3).

Table 7. Controlling of soil erosion and soil fertility values of Iran's forests Magnitude Unit of Measurement Particulars Sr.

No 96 Dollar/ hectare Controlling of soil erosion and soil

fertility 1.

0.558 ------------------ Costanza adjustment index for Iran 2. 12.4 Million hectare Iran's forest area 3.

6.64×108 Dollar Total value (1×2×3) 4. Source: Finding Results

Table 8. Production of oxygen values of Iran's forests

Magnitude Unit of Measurement Particulars Sr. No

88 Dollar/ hectare Production of oxygen 1. 0.558 ------------------ Costanza adjustment index for Iran 2. 12.4 Million hectare Iran's forest area 3.

6.1×108 Dollar Total value (1×2×3) 4. Source: Finding Results

Table 9. Basis for estimation of IUV of Iran’s forests

Sr. No

Environmental service Value (Dollars)

1. Production of oxygen 6.1×108 2. Controlling of soil erosion and soil fertility 6.64×108 3. Recycling of water and controlling humidity 1.55×107 4. Sheltering of birds, squirrels, insects and plants 6.79×105 5. Total IUV 1.29×109

Source: Based on Table 6-8 and Central Bank's Information.

Table 10. The 'OV' of Iran's forests at various discount rate (r) scenarios. Discount rate (r) Interpretation Value (Dollars)

5 0.22 (DUV+IUV) 4.895×108 10 0.052(DUV+IUV) 1.157×108 15 0.013(DUV+IUV) 2.89×107 20 0.0035(DUV+IUV) 7.79×106 25 0.001(DUV+IUV) 2.225×106

Source: Finding Results.



Table 11. Sensitivity analysis of Iran's forests 'EV' at various discount rate (r) and adjustment rate (γ) scenarios.

(Unit: $ billion) r γ

5 10 15 20 25

85 2.3 1.99 1.91 1.89 1.893 91 2.47 2.13 2.05 2.03 2.026 95 2.58 2.22 2.14 2.12 2.116

100 2.71 2.34 2.25 2.23 2.227 Source: Finding Results.

Table 12. Sensitivity analysis of 'TEEV' of Iran's forests at various discount rates (r) and adjustment rate (γ) scenarios.

(Unit: $ billion) r γ

5 10 15 20 25

85 5.01 4.33 4.16 4.12 4.12 91 5.18 4.47 4.30 4.26 4.25 95 5.29 4.56 4.39 4.35 4.34


Table 13. Iran's TEEV to GNP ratio per cent at various discount rate (r) and adjustment rate (γ) scenarios

r γ

5 10 15 20 25

85 6.49 5.59 5.38 5.33 5.328 91 6.69 5.78 5.57 5.51 5.50 95 6.83 5.9 5.68 5.63 5.61


REFERENCES Anderson, L., 1997. A cost-benefit analysis of deforestation in the Brazilian Amazon. Department of Economics, University of Aarhus, Denmark, (Discussion Paper). Anderson, P., 2005. Potential monopoly rents from international wildlife tourism: An example from Uganda's gorilla tourism. Eastern Africa Social Science Research Review, 21(1): 1-18. Barndill, T., 1999. Our green is our gold: The economic benefits of national forests for southern Appalachian communities. A forest link report of the southern Appalachian forest coalition. Southern Application forest coalition, place? Bello, L., Cistulli, V., 1997. Economic valuation of forest recreation facilities in the Liguria region (Italy ). Working Paper GEC 97-08, Centre for Social and Economic Research on the Global Environment, University of East Anglia and University College London. Bogahawatte, C., 1999. Non – timber forest products and the rural economy in the wet zone forests. Forestry policy, EEPSEA Research Report Series, Singapore: Economy and environment program for South-East Asia. Chopra, K., 1993. The value of non-timber forest products: Estimation for tropical forests in India. Economic Botany, 47(3): 251-257. Clayton, C., Mendelson, R., 1993. The value of watchable wildlife: A case study of Mc Neil River. Journal of Environmental Management, (39): 101-106. Costanza, R., et.al 1997. The value of the world's ecosystem services and natural capital. Nature, (387): 253-260.

Heafele, M., Kramer, R., Holmes, T., 1992. Estimating the total economic value of forest quality in high elevation spruce-fir forests, in C. Payne, J- Bowker and P. Reed (eds) Economic Value of Wilderness, Athens, Georgia: USDA Forest Service. Jalayeri, M., 2004. Investigation of renewable natural resources effects on GNP. Thesis for MS.c degree at agricultural economics, University of Sistan & Baluchestan, Zahedan, Iran. Khalilian, S., 1996. Analysis of the place of natural resources at country's economic development. Thesis for PhD degree at agricultural economics, University of Tarbiat Modares, Tehran, Iran. Khazaei, A., 1997. Investigation of government's policies on deduction of natural resources demolition, Journal of Eghtesad Keshavarzi & Tawsea, (20): 16-38. Kling, C., 1993. An assessment of the empirical magnitude of option values for environmental goods. Environmental and Resource Economics, (3): 471– 485. Lerner, S., Poole, W., 1999. The economic benefits of parks and open space: How land conservation helps communities grow smart and protect bottom line, S. Ives (ed). The Trust for Public Land, San Francisco, California. Loomis, J., Richardson, R., 2000. Economic values of protecting areas in the United States, The Wilderness Society, Washington, DC. Mesdaghi, M., 1998. Natural Resources Economics, 2nd edition, Mashhad, Iran. Nayak, B., 2001. Economic-ecologic values of an Indian forest. Indian Journal of Agricultural Economics, (56): 326 – 334. Pearce, D. W., Moran, D., 1994. The economic value of biodiversity, International union for the conservation of nature. The world conservation union, Earthscan, London.


RECLAIMING INSTITUTIONS AS A FORM OF CAPITAL

Bénédique PAUL

Laboratory of Economic Science of Richter (LASER) University of Montpellier I

Avenue de la Mer, CS 79606 34960 MONTPELLIER Cedex 2, France

ABSTRACT

Economists have recognized that “institutions matter” and the renewal of Institutional Economics has gained a large scientific authority, but institutions, as unit of analysis, stay a concept not understood on all its dimensions. Recently, the notion of “institutional capital” appeared in the literature, without neither satisfied definition nor demonstration. In this article, we adopt the “Resourced-Based View” approach to show that some kinds of institutions can theoretically be considered as form of capital, namely “institutional capital”. The main conclusions of this article are explicated in the promises of this new approach for future researches on growth theories, economic development theories, organizational theories, and overall on New Institutional Economics. We underline the importance to test the empirical strength of this economic resource. Meanwhile, economists may accept that “institutional capital matters”.

INTRODUCTION Economists have recognized that “institutions matter” since few decades (Sachs, 2003), and have decided to complete the institution’s description engaged in the Old Institutional Economics (OIE). The concept of institution has gained a large scientific authority, becoming a unit of analysis. But the concept stays a research object that is not understood on all its dimensions yet. At the time when the institutional economics is renewed with the publications of Coase (1937, 1984, 1992), Williamson (1975, 1985, 2000) and North (1990, 1995, 2005), the concept of institutional capital appeared. In 1969, André Micallef published in the “Revue Economique” a treating article of the collective institutional capital. By this concept, the author understood the capital represented by “the stock of goods necessary for a real production or a whole of services” (Micallef, 1969, p. 4). He defined this stock as a source for the products of the national capital’s specific part, and also distinguished the administration’s general capital and the sociocultural capital. But this managerial vision made it possible to study the field of the public services. Thus, Micallef’s theoretical design did not take into account the insights of the new institutional economics (NIE). His point of view on collective institutional capital cannot be useful for research

in NIE as being interpreted by economists like Douglass North, Dani Rodrik, Geoffrey Hodgson, etc. Since this first occurrence in the literature, the concept of institutional capital has appeared more than twenty years after. In 1996, Michael Trebilcock inaugurated the resumption of this concept in “What Makes Poor Countries Poor? The Role of Capital Institutional in Economic Development”. Thereafter, this inaugural article opened the way to various publications: Picciotto, 1996; Palley, 2001; Khakee, 2002; Bresser and Millonig 2003; Ashan, 2003; Bauder, 2005; Fedderke and Luiz, 2008. These authors approached very differently the institutional capital. The cases in which a definition is proposed for the notion are rare. In addition, it lacks a theoretical junction with new institutionalism. Moreover, it lacks a demonstration of the scientific validity of this notion. Whereas the characteristics of the capital are known, it is not the case for institutional capital. There is thus a methodological insufficiency that this article wishes to contribute to fill. The fundament of this article is to question the theoretical and scientific bases of the institutional capital. For that, we explore the economic literature relative to the concept of institutions, which is the institutional capital constituent. After having pointed out the more accepted definition of the institutions, we will propose a restriction for some kind of institutions, using the resource-based view approach. Then, we will demonstrate that such economic assets have the capital characteristics. A new definition for the institutional capital will be given and we will underline three main implications for future researches on economics.

WHY TO RECLAIM INSTITUTIONS AS A FORM OF CAPITAL?

Here is the main question that needs a scientific answer before talking about “institutional capital”. Any valid answer should start with a useful definition of institutions. Defining institutions, North’s point of view Institutions are the first type of resources to be accumulated in any society. However, the ambition of Emile Durkheim to make sociology the science of the institutions did not arrived to impose those like a central object of study in the disciplines of the social sciences. It was necessary to wait


until the return of the institutions in the economic researches, with authors like Oliver Williamson, Douglas Cecil North, to highlight the importance of these resources as conditions of the economic performance (North, 1990). In addition, the meaning of institutions was for a long time so perverted that the concept had been confused with organizations, in particular the officially registered organizations. However North, in his theory on institutions, did not underestimate the urgency to make distinction between the two concepts. He defines organizations as “the groups of individuals bounded together by some common purpose” (North, p. 361, 1994 cited in Hodgson, 2006). Whereas the institutions are “the human devised constraints that shape human interaction” (North, 1990). The organizations (such as associations, groups, communities, companies, trade unions, etc.) are not institutions and vice versa. Even if they participate both in structuring the economic agents’ interactions, the institutions must be distinguished from the organizations (Arrous, 1999). The concept of institution according to North is central in the analysis of the economic development process. Institutions are guides of the human interaction. They define and limit the whole of choice of the individuals (North, 1990, p. 4). More lately, North established the link with the definition of the economic growth. “The institutions provide the structure of incentives of an economy; at the same time as this structure evolves/moves, it works the direction of the economic change towards the growth, the stagnation or the decline” (North, 1991, p. 97). For this reason, one cannot claim that an economic policy can function without the institutions. Institutions are not only constraints. In the direction indicated by North, we can go further and stress that “institutions are all (rules, standards, constraints, mechanism of incentive/constraint) that codifies the interactions between the economic agents”. Such a definition makes it possible to exceed the formal/informal duality that is confused with the official/unofficial duality. This progressive codification of the rules leads us to prefer the distinction between “written/not written institution” to the “formal/informal institution”, when it is a question of global typology of institutions. From the preceding posture, we adopt North’s approach: institutions are thus rules in force in a given social space. In a general way, we define them as “all that codifies the relations between the individuals”. Institutions are operational if they are more or less collectively accepted. They can be integrated in the economic and social actors’ behaviour. They can even become rules of behaviour. The concept of institution gathers at the same time the rules known as formal and/or informal, as well as the social norms.

However, in the definition of the notion of institutional capital, we adopt a restrictive posture. The institutions that compose the institutional capital are those considered as a resource for the economic agents. Institutions as resources for economic agents, the RBV Since the publications of Douglass North, establishing institutions as constraints and/or incentives for human actions, the perception of the institutions evolved. The actual trend perceives less and less institutions as constraints. The analytical framework called “Resource-Based View” provided by the theorists of management appears very usefully. This approach makes it possible to apprehend the institutions like resources for the individuals, in organizations. The theorists of the Resource-Bases View (RBV) propose to answer a basic question “how the organizations (firms) obtain and maintain comparative advantages?” They supports that the answer to this question is in the fact of the possession of certain key resources, like values, barriers to duplication and appropriability (Fahy and Alan, 1999). It is Christine Oliver who, in “Sustainable competitive advantage: Combining institutional and resource-based views” (1997), integrated explicitly the institutions in this vision. Her analysis has been continued by Bresser and Millonig (2003). But one fundamental question is “what is necessary to understand by resource?” Caves (1980) cited by Birger Wernerfelt (p. 172, 1984).provided us an interesting definition of the concept of resource. His definition was taken reproduced by Wernerfelt (1984). According to Caves, in the case of the organization, “a firm’s resources at a given time could be defined as those (tangible or intangible) assets which are tied semipermanently to the firm” (p. 172, 1984). Based on this conception, Wernerfelt admits that elements such as “trade contacts”, “efficient procedures”, “capital”, etc., are resources for the organization (ibid.). His observations, as developed after by the continuators of the RBV, enabled him to note that “in some cases, a holder of a resource is able to maintain a position relative vis-a-vis other holders and third persons, as long as these act rationally” (ibid p. 173). From there, Bresser and Millonig (2003) develop the idea of comparative advantages. Generally, a resource is defined as “Something that can provide satisfaction to a need, what can improve a situation” (Le Robert, 1st edition, 1973). In the Dictionary of contemporary economics and the principal political and social facts, Lakehal (2002) puts this: “a resource is a means of subsistence for a person, a family or a group of people”. The concept of resource is related to a utilitarian approach of the institutions. Thus, institutions can make it possible to improve the production process, consumption, interactions, exchanges, etc. Consequently, we can consider its accumulation process (already presented by North


(2005, p. 20). The institutional dis-accumulation carried out within the framework of the institutional change can also considered. This conception of institutions enables us to consider the economic utility of the resource. In our case, we define a resource as a factor allowing an economic agent to satisfy a need or to achieve an objective. It is for this reason an institution can be regarded as a resource. And when this need or this objective is of economic order (like consumption, production, investment, exchange or trade…), the institution in question can be considered as an economic resource. It is in this order of idea that we will take institutions as economic resources, with characteristics of capital. In an economic sense, this statement seems to be relevant. Because, institutions that are productive are being considered. It stays to demonstrate and illustrate this statement. Indeed, as announced by Loury (1977, 1987) and quoted by James Coleman (1990, p. 300), the factors making it possible for the actors obtain their objectives are a resource for them. They are for example the institutions making it possible to reduce the costs of transaction within the context of the economic exchanges. Indeed, by shaping the institutions structuring their interactions, the actors - under the assumption of their rationality - seek the order, unit, simplify their relations. The demonstration of Michael Lounsbury and Mary Ann Glynn (2001) for the contractors also fits in this line. On a broader level, North (1990) showed that the institutions have a particular importance in the economic development of the nations. It is for this reason that the institutions were arranged among the assets that are required for a nation’s economic development. What justifies the name of “institutional capital”, this concept deserves to be defined and demonstrated.

THE SCIENTIFIC BASIS OF INSTITUTIONAL

CAPITAL In Cent and sociability: Household income and social capital in rural Tanzania, Narayan and Pritchett (1999) underline clearly several six types of capital which are to be taken in account by actors and analysts of economic development strategies. They state: “Beyond apparently now old fashioned “physical” capital, human capital, natural capital, institutional capital and social capital all clamor for attention” (Narayan and Prichett, p. 1999). If the literature is largely abundant for the other forms, it is quite different for the institutional capital. Defining institutional capital One of the first efforts to connect this notion to the NEI was recently initiated by Rudi K.F. Bresser and Klemens Millonig (2003). They propose a very general definition of the institutional capital. For the two theorists of management, the institutional capital is defined as “the specific conditions in an organization’s internal and

external institutional context that allow the formation of competitive advantage” (ibid., p. 229). For these authors, the institution can be defined as “behavioral expectations that can be sanctioned if violated” (ibid., p. 221). Knowing that, for them, the institution has three components in interaction: cognitive, normative and regulative (ibid., p. 226). They govern economic agents’ interactions. This point of view is acceptable, but we consider that it is too restricted (to competitive advantage) and it is only functional. We will show that institutional capital allows status (thus specific or competitive advantage) to its holder. But it is more than this. Michel Garrabé (2007) proposed a more descriptive definition of the concept, in a contribution to MED-TEMPUS training program implemented by the International Centre of the High Mediterranean Agronomic Studies. In this contribution, the term institutional capital is understood as “the whole of the formal and abstract institutions which constitute the inciting structure organizing the relations between individuals or organizations, within the process of economic and social production” (Garrabé, p. 127, 2007). This definition is closer to the term of our apprehension because it seems to be useful within a framework of an empirical study. Despite of that, Garrabé’s definition is larger than the precedent, even if he presented the institutional capital as a kind of equipment the production of which would be largely generated by the organizations of the social economy. More recently, Joost Platje (2008) defined institutional capital as “institutions, institutional governance and governance structures that reduce uncertainty, stimulate adaptive efficiency (i.e. the ability of a system to adapt to changing conditions) and stimulates the functioning of the allocation system and sustainable production and consumption patterns” (Platje, p 145, 2008). But we can denote confusion in this definition. Because Platje’s conception of “institutional governance” concerns the judge of the game. He states that “institutional governance” concerns “organizations that interpret and enforce the rules of the game such as the judiciary, police, government and government agencies” (ibid.). This conception of institutional capital is in opposition to our statement. It is outside the Northian perspective of institutions. For institutional and organizational structure (Ahrens and Jünemann, 2009) are obviously different. The institutional capital, as we conceive it, is in the prolongation of the neo-institutionalism. We define institutional capital as the asset composed by the written and unwritten institutions that affect economic activities. It concerns the institutions that are directly or indirectly productive. Ahrens and Jünemann (2009) talk about these productive institutions in their work on “adaptive efficiency” of institutions. It can usefully provide economic agents (individuals or organizations) with economic


advantages. Generally, its role is to structure economic relations between individuals or organizations through its inciting or constraining influences. Functionally, it is a potential for economic development. It is considered as a resource whose detention provides economic advantages. Those are called “competitive advantages” (Bresser and Millonig, 2003), “position barriers” (Wernerfelt, 1984). This gives it an important implication in economics and organizational theory. We are agreeing with Garrabé (opus cit.) when he says “the institutional capital represents the essence of the inciting equipment making possible the accumulation of other forms of capital”. We illustrate in Table 1 the main partition of the institutional environment highlighting the components of the institutional capital. The restrictions expressed in Table 1 will allow us to better determine the properties of the institutional capital. For example, the rules or social norms defining the hierarchical system compared to the age within the families in certain societies are thus excluded to our definition of institutional capital. Of course, social institutions have indirect relation to economic interactions. For example, social rules between families influence economic repartition between the families’ members. Even though, this is not what we are calling institutional capital in a strict economic view. These restrictions permit us to go beyond the formal/informal duality usually taken as basic in the institutional analysis while remaining within framework of the economic assets. Moreover, frequently formal/informal duality is used to refer to official/unofficial duality, which is very close to our written/unwritten distinction. This distinction better reflect the progressive institutions codification between economic agents. According to the table 1, we consider that the institutional capital is an element (stock and flows) of the environment or institutional framework. But overall, this new concept must be analysed in the light of the properties of any types of capital. Theoretical Justification of the Scientific Validity of Institutional Capital What characteristics do confer to a resource the properties of a capital? This is to such basic but fundamental question we intend to carry an answer for the institutional capital. To answer this question, we adopt the approach used by James Coleman (1988), to show that the social capital was a particular form of capital. In Coleman’s approach, a resource that presents the properties of any stock of capital is capital. These properties are mainly: properties of profitability, accumulation, fungibility and depreciation. We will analyze these properties for the case of institutional capital, since they had never been refuted. Methodologically, to be capital, only the properties of profitability, accumulation and durability could be regarded as necessary and sufficient conditions. With these last, we can add the fact of being a factor of production. The

properties like obsolescence, fungibility, productivity, the capacity to confer a social status to the holder, are necessary only for the economic analysis carried out under a very specific view. As for properties like the transferability, tangibility or intangibility, they could only be additional. The whole of the current properties of the capital could be summarized in the Table 2. We will analyze here only the most important properties for the demonstration. These are the necessary and sufficient properties quoted in Table 2. The first property we analyze is « factor of production ». The conception of the capital as factor of production is in conformity with the neo-classic view of capital. Economic institutions can be considered as factors of production by integrating the Labour factor or by assimilation in production process. As we explained, Labour is now established as factor of production. But everyone knows that Labour is education, skills and health. Then Labour is human capital. But part of the human capital is constituted by institutions that have been integrated in habit. When we accept Labour as a factor of production, we are implicitly accepting some specific institutions that allow workers achieve production process. Let’s take by example the case for the institutions that concern the production of exchangeable goods. We position in a context where the demand determines the supply and not the reverse, and where consumers benefit perfect information and have the capacity to check the authenticity of the good put on the market. In other words, the production will be regarded as such and will have a commercial value if and only if it is carried out according to rules/institutions defined and known in advance. Ceteris paribus, if the producer of the goods in question (a very good example is the case of the organic products) does not take account of the whole of the institutions that concern his production process, the output of its activity could not be regarded as an exchangeable production. Then institutions form part of the production process. Their absence or the fact that they are not taken into account cancels all the production’s value. The production while being material arises then as being an incorporation of specific institutions. These last can then be assimilated as factors of production, and consequently as capital. Moreover, they are not substitutable by any other factor, which authorizes to regard them as a form of differentiated capital. Their taking in account implies some costs and justifies a higher price for de products. Second property to be analyzed here is the profitability. It is the first one in James Colman adopted properties. Profitability is the relationship between a result obtained and the means in capital implemented to obtain it. We are


using this term here to signify the possibility for a factor to generate a surplus or an advantage. We take the case of two organizations, located in a fluid context of circulation of information at low costs, maintaining between them important economic exchanges. They have the choice to define in advance the rules of the exchange or on the contrary to engage there without preliminary negotiation fixing the rights and the duties of each one. In the last case, the possible costs being able to be caused by litigations can be very high. However a few hours of negotiation would be enough to establish and be appropriate of the institutions governing the exchanges. If one considers the opportunity cost of the development of the institutional framework of exchange, the option consisting in negotiating beforehand is more profitable. It is it more than one possible recourse to progressive negotiations or a third in case with litigation. The difference in costs with the first case is due to the institutions. Even when its production is regarded as intentional and is justified by a certain interest (that related to the application of the norms/sanctions), the comparison between the investment costs and the advantages provided make it possible to consider a profit. The existence of the institutional capital in the interaction context justifies the superiority of the economic advantages provided by this one over the implementation costs (Kaji, 1998). The advantages provided by the institutional capital accumulated in the preceding framework will largely exceed the effort agreed [cost in time and the value of this one] to constitute it and put it in action. It is this hopeful profitability which justifies the creation of institutions within a framework of interactions in a democratic atmosphere.

The institutional capital thus makes it possible to reduce the costs of information and uncertainty. Thus, it brings profit to the economic agents in interaction. It also allows more effective economic exchanges (except possibly from the point of view of the opportunist actor). With this property, the institutional capital is presented in the form of an input reducing the production costs. In this case, the profitability of capital institutional seems to be the least refutable condition. For example, like in the case of the legislative rules analyzed by Michel Garrabé (2007), one can show that the installation costs of certain rules are quite lower than the costs associated with the risks of errors or litigations which can involve of the obsolete rules.

When we return to the economic activity leading to the biological production, the incorporation of specific institutions defined in advance allows and justifies a higher selling price for the products put on the market. This profitability seems to be one of the elements (being added to the awakening of the climatic risks) which justify the expansion of the current organic production sector.

Fourth is the durability. The concept of durability, in the case of an economic asset, can be understood as its aptitude to persist in time. It is used here to mean the capacity of a factor of production to survive the production process. It doesn’t disappear or consumed during the production process. This is the nature of institutions. They are created to be durable while ensuring a time-saver and procedures. Of course, as we will see it thereafter, they are called to evolve/move. From here comes the idea of North’s “path-dependency”, identifying institutional permanence inside the change, consequently institutional capital accumulation is possible because this accumulation will continue as long as a social crisis did not come to oblige to change actual institutions. If we restrain our context at the dimension of the production process or exchanges, the institutional capital preserves its durability. Indeed, all things being equal, the institutions defined in advance to govern the process are not modified at the exit of this. In the case of the biological production, it is the stability of the preliminary institutions which ensure the authenticity and consequently the quality of the products to be exchanged. Third is the accumulation of the capital. As stated by Marx in 1867, accumulation of capital is the permanent reintroduction of the added-value in the circuit of production in order to form new capital. But the reproduction of the system requires its widening, and the accumulative tendency makes possible the overproduction crises. It is not different for the production process of the institutional capital, which involves an accumulation in time. In their interactions, the economic actors devise new institutions. If the new institutions do not enter in contradiction with the old ones, there is accumulation. Garrabé (2007) analyzed the accumulation of the legislative rules and arrived at the census of four forms of accumulation: institutional imitation, convergence or institutional harmonization, institutional innovation, and transformation of informal into formal. This last form, usually progressive, is what we describe as “progressive institutions codification”. It represents a very important form of accumulation of institutional capital. In much the same way, a good institutional reform participates in institutional accumulation. The institutionalization process can sometimes contain the institutionalization-destruction side whose most radical forms, according to Hodgson (2006), can be observed during invasions and occupations of a society. Accumulation or the destruction can come from an individual as well as an institutional convergence. It can be voluntary or negotiated (for example within the space of interactions) or imposed (it is the case in a dictatorship). Even in period of stability, accelerate institutional capital accumulation can turn into institutional inflation. In such case, too many institutions without operational link between them can become contradictory.


Institutionalization as process which can be heard as the accumulation of the institutional capital is discussed by several sociologists. Their contributions make it possible to distinguish some in several phases. Rene Loureau, in a publication in 1970 on the Institutional Analysis, distinguished three moments or three phases which we can use to study the institutional capital. Initially, it distinguishes the “instituted” who is pre-established institution integrated by the people finally seem normal to them. The “instituted” becomes “unconscious” and model what Pierre Bourdieu (1972) will call the “habitus” or “habit” for Geoffrey Hodgson (2006). With the appearance of social strains, or crisis in other words, a social change is announced and with time, the individuals can manage to create new institutions, then comes the moment for “instituting”. Therefore during this challenge, if the instituting movement manages to win the bet, it will have there a certain stabilization of new norms, rules, in manner of acting and of thinking which, while crystallizing, makes it possible to reach a new stage of stability. This last moment is “institutionalization” process itself. Institutionalization in general is thus a periodic process with more or less long run. For this reason the speed of accumulation can appear stronger in short-term. The process contains change and continuity; one does not set out again to zero. The present system is the result of a past. The institutional capital accumulates slowly in time, except in institutional crisis situation. The evolution of institutional stock is done by successive contributions (incrementally) North (1991). Whereas, the evolution even of the institutions supposes a mobilization of a surplus generated by their mobilization. One could also approach accumulation under stock point of view (Garrabé, 2008). This means an increase in the number of the institutions that would correspond to an accumulation of institutional capital. This moment, through a process of organisational training, the actors work out more and more institutions to govern their interactions without necessarily giving up the former institutions. Lastly, if we conceive the accumulation of institutional capital in terms of effectiveness, a useful criterion would be the value of stock. Consequently, an accumulation would not correspond inevitably to a modification constitutive of the stock of the institutions. The adaptation and the improvement of the institutions with the needs for the exchange contribute also to institutional capital accumulation. If we refer to the biological production in a general way, since the appearance of this sector, the institutional elements regularly accumulated, without notable contradiction. These four conditions or properties are satisfied and are enough to show that the institutions structuring economic relations between agents or providing advantages to them can be called capital institutional. The previous justification

joined perfectly the definition of Lakehal (2006) in the dictionary of contemporary economics remembering that the capital is “an economic asset having at least three characteristics: it survives a cycle of production, it provides a regular flow of incomes to its holder, and enables him to sit a social status through the economic capacity which it represents” (ibid, p. 43). Other properties could be discussed, like the obsolescence of the institutional capital caused by development of new more relevant institutions. We could consider its localization in the space and the time i.e. it is the product of the social innovation of the individuals actors of social space considered and is not irremovable from where limitations of the institutional imitation (Bajenaru, 2004). The institutions are worked out for the needs for the current economic processes and nothing guarantees their presence in the future, because certain new rules make disappear from others. Through their appropriation by individuals who integrate them in their habit of though and their habit of behaviour, institutions are more or less fungible in human capital. For more explanations on both types of habit, we refer to Hodgson’s article Reclaiming habit for institutional economics (2004). Certain determining characteristics of the nature of the institutional capital are summarized in Table 3. Relations between Institutional Capital and the others Forms of the Capital We consider here six forms of the capital, as fundamental for economic development and economic theory. In his 1998 publication, Carney reported by Katherine Warner in a FAO’s publication coined five forms of capital. When he was analyzing the forms of capital required for sustainable livelihoods, Carney retained theses forms of capital: natural, physical, financial, human and social. He didn’t recognize institutional form. Although one year later, in his Lessons from early experience, a publication with Ashby we can read: “Sustainability of livelihoods rests on several dimensions - environmental, economic, social and institutional” (Ashby & Carney, 1999). We then are in agreement with the authors addressing capital in its six forms: physical capital, natural capital, financial capital (in a Keynesian meaning), human capital, social capital and institutional capital. Several links exist between them and require deep studies. This is not the purpose of this paper. But we underline rapidly some of them. In order to simplify our standings, we consider together physical capital and natural capital, for they are both the most tangible forms of the capital. Institutional capital is a basis for social capital accumulation. It structures relations in which individuals work out their social capital. Institutional capital participates in limitations of what Portes and Landolt call « the downside of social capital » (Portes and Landolt, 1996). It provides conditions for repeatability in exchanges


and social relations. Conversely, social capital makes possible institutional capital accumulation, by providing a social framework for this. Social capital is ingredient to create organizations. Thus, institutional and organizational structures are close concept. There are strong links between human and institutional capital. Ahrens and Jünemann (2009) put it “For it the incentives resulting from the overall institutional structure that guide learning processes and the emergence of tacit knowledge”. Institutions pass in the individual’s habits and make part of human knowledge. That is why individuals are “instituted”. “But, human capital does not stand alone either” declare Fedderke & Luiz (2008). Mutual relations exist between human and institutional capital. Ahrens and Jünemann (2009) argue “the underlying process of acquiring knowledge will direct individuals and organizations gradually to create new institutional arrangements”. It involves institutional capital accumulation. It determines quality of institutions. In the other hand, institutional reform and institutional change are driven by people (Acemoglu and Robinson, 2008; Ahrens and Jünemann, 2009). Institutional capital is important for financial capital stocks and flows. Globally, institutional capital is basic for economic capital (financial and physical or natural) transaction and creation. It is a key to access financial resources. Market institutions, like price, are determinant for transactions. Likewise, financial capital is needed for investment in institutional capital drawing. In this time of financial crisis, more strong links are called up between financial capital and institutional capital. Either for Marx (1867) or Hilferding (1981), financial or not, capital comports an abstract dimension. For whatever other form of the capital, institutions are important because capital is relation. In Marx’s view of the capital (1867), “instead of being a thing, the capital is a social relationship between the people” (the capital, op. cit, chapter XXXII, volume 3, p. 207). And relations imply rules. That is why institutional and organizational structures are so closed variable to understand economic growth and development (Ahrens and Jünemann, 2009). Once the theoretical and scientific validity of the notion of institutional capital is demonstrated, it is important to analyze what is its theoretical and empirical fruitfulness. The next part is dedicated to some core implications of this notion and asset. Then, institutional capital is presented as a heuristic concept for future research in several domains of economic theories.

CONCLUSION In this article, we intended to demonstrate the scientific validity of the “institutional capital” concept which appears more and more frequently in the economic literature. We

define the concept through a restricted view of institutions, using the Resource-Based View findings. Thus, only the institutions providing economic advantages to economic agents (individuals or organizations) are considered as part of the institutional capital. This restriction allows new option in the definitions of institutional capital. We describe institutional capital that the asset composed by the written and unwritten institutions that provides economic advantages to the agents (individuals or organizations, in interactions). This definition as shown brings interesting perspectives on future researches in economic theory. After a theoretical justification that the institutions are a resource for the economic agents, we answered the question that is: as of when one can say that institutions are forms of capital. The known properties of the capital (accumulation, durability, profitability, factor of production) being checked for the categories of institutions retained in the justification, we arrived at the statement that some set of institutions that structure economic interactions constitutes a form of the capital (as Frankel’s view) with whole share, name institutional capital. This study does not answer all the questions about the subject, but it poses strong foundations for research even more fertile and of greater range in the field of the economic theories. This paper comes to fill a methodological vacuum insofar as it is important to define and show the validity of a concept before its empirical mobilization. From a theoretical point of view, it is however necessary that other studies come to consolidate the foundations of the debate on this new concept. Research on institutional capital would not be interesting if this concept was not helpful in applied economics. But the utility of the mobilization of such a notion is to be sought in its capacity to give an account of another socio-economic reality. We could underline several core implications of the concept of institutional capital for future researches. First, it can help economists to deepen specification of growth models, especially in endogenous growth models. Considering institutional capital as namely an economic asset would conduct economists to include an I (for institutional capital) factor in the model Nicolas Sirven developed in 2004, following Frankel (1962). Then, the function Y = Â.Sγ.Hβ.Kα.L1-α (Sirven, 2004) would be Y = Â.IιSγ.Hβ.Kα.L1-α, where Iι represents the part of institutional capital affecting the growth process in which Â is the residual, I the institutional capital, S the social capital, H the human capital, K the technical capital and L the Labour. Second, theorists of economic development can find an analytical and explanatory framework in this concept. Early in the twentieth century, economists have recognized that development is more than growth. “A great deal of economic research in recent years suggest that institutions are vital for economic development and growth” (Rodrik and Subramanian, 2003; Edison, 2003; Acemoglu, 2003;



Acemoglu and Robinson, 2008). The accumulation of physical assets is not enough to durably feed the growth, the socio-economic context matters. For us, we agree with Platje when he says: “institutional capital is a fundament of sustainable development, and the lack of such a capital is likely to cause of a unsustainable development” (Platje, 2008). Indeed, the institutional capital represents the essential component of the social and economic order necessary to sustainable development (ibid). And, high levels of institutional capacity are conditions to sustainable development policy achievements (Evans et al., 2006). According to the writings of Michael Trebilcock (1996), Kaji (1998) and Ahsan (2003), inter alia, Nations would benefit better understanding on institutional capital. The authors underline well the determining role of the institutional capital in the economic development and the reduction of poverty. For this reason, the politicians of the developing countries could draw advantage from more a great attention paid to this asset, particularly by its definition and its public management. Recent publications show institutions as determinants for Foreign Direct Investment attraction (Bénassy-Quéré, Coupet & Mayer, 2005). More especially, efforts are being done in order to establish “institutional country profile” (Berthelier, Desdoigts & Ould Aoudia, 2004). It may correspond to the quantity and the quality of institutional capital the stock and the flows. Third, in the same way, organizations’ theorists and practitioners could have great interest to deepen research in this concept of institutional capital because managing organization is managing institutions. It is besides of this reason that institutional capital is addressed as an element detention and management of which is strategic in the organization (Oliver, 1991; Bresser and Millonig, 2003). In fact, institutional capital, by structuring relations between economic agents, improving good economic relations and reducing transaction costs, represents a potential for economic development. As development is use of several assets, institutional capital is required for it. Fourth, after several Nobel’s Prize obtained by New Institutional Economists (like Coase, North), research on institutional capital would be a great step for New Institutional Economics. Comprehension on institutional capital can not only confirm but consolidate the NEI approach. It appears as a core object and the premise of a solid paradigm for NEI researches. Recently, in a collective publication, Robert Solow (2001) has associated institutions to a potential growth in Europe. Solow puts it:

“I will define institutions as regular and codified modes of behaviour which, according to the cases, can appear in an endogenous way or come from social norms” (p. 405, 2001). Solow’s definition of institutions is very near to our standing here. In a prophetical statement, Solow explains that future of economic theory on institutions and growth […] must develop a quantitative part or it must be at least connected to the growth models worked out by the standard economic theory (ibid.). However, institutional capital is one more time a very good way to obtain such theoretical objectives. We can finally insist on these main implications to underline the importance of reclaiming institutions as form of capital and to continue research on institutional capital. In practice, the institutional capital appears as being an analytical framework for analysts of actions and development strategies. More especially it is presented as one of the assets whose use is necessary for sustainable economic development (Platje, 2008). In organizations, it is a strategic resource that allows competitive advantages (Bresser and Millonig, 2008). Equally, at the level of a nation, institutional capital can strongly contribute to attract foreign direct investment. It appears to be crucial for specific sector expansion, like Microfinance (Paul, 2009). Then, institutional development can be analyzed as fundament for economic development. In addition, it maintains complex interrelations with the other forms of the capital (physical, financial, human and social). While waiting for an empirical evaluation of this concept, the theoretical debates on its scientific basis, as initiated in this paper, show interesting perspectives. Its heuristics is available either for economists or management theorists. Table 1. Components and Delimitations of Institutional Capital

Institutional Environment Other Institutional

Resources Institutional Capital

Written Institutions

Unwritten Institutions

Written Institutions

Unwritten Institutions

Rules or institutions with any direct relation

economic interactions

Rules or institutions with direct relation economic interactions (market exchanges, production relations)

Source: the Author.

Table 2. The Properties of the Capital by Authors and by Degree of their Importance Authors Properties

Leon WALRAS

James COLEMAN

Adam SMITH

Other (Encyclopedia)

necessary and sufficient properties

Durability Accumulation

C

apita

l = w

hat p

erm

its

to g

ain

prof

it.

Resource Profitability Profitability

Factor of Production

— Factor of Production

Sufficient properties Social Richness Depreciation Productivity

Fungibility Confer Status

Materiality Intangibility Social relation Neither necessary nor sufficient properties

Transferability Divisibility

Source: the Author.

Table 3: The Properties of Institutional Capital

Properties typology Institutional capital

Essential properties Resource collective

Confer a status to its holder Factor of Production/of Development

Is productive and durable Space-time localization

Other properties

Limited but Crucial ability to be seized by Individuals Slow and long process of accumulation

Negative effects if excessive accumulation: “too many rules kill the rule“ It is linked with the other forms of the capital, and improves and facilitates their

accumulation Source: the Author.


ACKNOWLEDGEMENTS We are grateful to Michel Garrabé for pertinent comments and Sopin Jirakiattikul for corrections of the first draft of this article.

REFERENCES Acemoglu, D., and J. Robinson. 2008. The Role of Institutions in Growth and Development, Working Paper, n° 10, Washington: The International Bank for Reconstruction and Development/World Bank, Working Paper Series. Acemoglu, D., and J. Robinson. 2000. “Political Losers as a Barrier to Economic Development,” American Economic Review, 90: 126–44. Acemoglu, Daron. 2003. Root causes: A historical approach to assessing the role of institutions in economic development, Finance & Development, 40(2): 26-30. Aoki, Masahiko. 2007. Endogenizing institutions and institutional changes, Journal of Institutional Economics, 3(1): 1-31. Ahrens, J. and P. Jünemann. 2009. “Adaptive efficiency and pragmatic flexibility: characteristics of institutional change in capitalism, Chinese-style”, Unpublished Paper, [online], available at http://www.centralasiaproject.de/index.php?option=com_docman&task=doc_download&gid=32&&Itemid=7, 19 p. Ahsan, S. M. 2003. Institutional Capital and Poverty: A Transition Perspective, in Perspectives on Poverty and Growth, van der Hoeven, R. and Shorrocks, A. (edit.), United Nations University Press. Arrous, Jean. 1999. Les théories de la croissance, Paris: Seuil. Ashby, C. and D. Carney. 1999. Sustainable livelihoods: lessons from early experience. London, UK, Department for International Development. Cited by Warner 2000. Bajenaru, Violeta. 2007. Complexité et Diversité des Comportements Individuels dans le Contexte du Changement Institutionnel Post-Communiste. Paul Cézanne University: PhD dissertation. Bajenaru, Violeta. 2004. Le mimétisme institutionnel et la croissance économique en Roumanie : une transition à deux vitesses, Bruxelles, XLème Colloque de l’Association de Science Régionale De Langue Française (ASRDLF), Unpublished Paper,

http://www.ulb.ac.be/soco/asrdlf/documents/Bajenaru1_001.pdf. Bauder, Harald. 2005. Institutional Capital and Labour Devaluation: The Non-Recognition of Foreign Credentials in Germany, INTERVENTION: Journal of Economics, 2(1): 75-93. Bénassy-Quéré, A., M. Coupet, and T. Mayer, 2005. Institutional Determinants of Foreign Direct Investment, Centre d’Etudes Prospectives et d’Informations Internationales (CEPII), Working Paper (May): 2005. Berthelier, P., A. Desdoigts, and J. Ould Aoudia, 2004. Profils Institutionnels : Une base de données sur les caractéristiques institutionnelles de pays en développement, en transition et développés, Revue Française d'Economie, 19(1): 121-196. Bourdieu, Pierre. 1972. Esquisse d'une théorie de la pratique, Genève : Droz. Bresser, R. K. F. and K. Millonig. 2003. Institutional Capital : Competitive Advantage in light of the New Institutionalism in Organization Theory, Schmalenbach Business Review, 55: 220-241. Carney, David. 1998. Implementing the sustainable livelihoods approach. In D. Carney, ed. Sustainable rural livelihoods: what contribution can we make? London, UK, Department for International Development. Coase, Ronald. 1992. The Institutional Structure of Production. The American Economic Review, 82(4): 713-719. Coase, Ronald. 1937, The nature of the firm. Economica, 4(16): 386-405. Coleman, James. 1990. Foundations of social theory, Cambridge, Massachusetts: The Belknap Press of Harvard University Press. Edison, Hali. 2003. Testing the Links: How strong are the links between institutional quality and economic performance? Finance & Development, 40(2): 35 – 37. Eigen-Zucchi, C., G. S. Eskeland, and Z. Shalizi. 2003. Institutions needed for more than growth: by facilitating the management of environmental and social assets, institutions underpin sustainable development, Finance & Development, 40(2): 42 – 43.


http://www.centralasiaproject.de/index.php?option=com_docman&task=doc_download&gid=32&&Itemid=7

http://www.centralasiaproject.de/index.php?option=com_docman&task=doc_download&gid=32&&Itemid=7

http://www.ulb.ac.be/soco/asrdlf/documents/Bajenaru1_001.pdf

http://www.ulb.ac.be/soco/asrdlf/documents/Bajenaru1_001.pdf

Evans, B. et al. 2006. Governing local sustainability, Journal of Environmental Planning and Management, 49(6): 849-867. Fahy, J., S. Alan. 1999. Strategic Marketing and the Resource Based View of the Firm, Academy of Marketing Science Review, 3(4): http://www.amsreview.org/articles/fahy10-1999.pdf. Fedderke, J. W., and J. M. Luiz. 2008. Does human capital generate social and institutional capital? Exploring evidence from South African time series data, Oxford Economic Papers, 60(4): 649-682. Feeny, David. 1988. “The Demand for and Supply of Institutional Arrangements”, in Vincent Ostrom, David Feeny, and Hartmut Picht, eds., Rethinking Institutional Analysis and Development. San Francisco: Institute for Contemporary Studies Press. Frankel, Marvin. 1962. The Production Function in Allocation and Growth. American Economic Review, 52: 995-1022. Garrabé, Michel. 2007. Economie sociale et développement, MED-TEMPUS Program, http://www.formder.iamm.fr. Garrabé, Michel. 2008. Note sur l’existence du capital institutionnel, Unpublished Paper, Workshop University of Montpellier 1 and Sherbrook University, (June). http://www.michel-garrabe.com. Glaeser, E., et al. (2004b), “Explaining Growth: Institutions, Human Capital, and Leaders,” (Preliminary Draft) http://www.brook.edu/es/commentary/journals/bpea_macro/papers/20040325_glaeser.pdf. Greenwood, D. T., and R. P. F. Holt. 2008. Institutional and Ecological Economics: The Role of Technology and Institutions in Economic Development, Journal of Economic Issues, 42(2): 445-452. Hilferding, Rudolf. 1981. Finance Capital, A Study of the Latest Phase of Capitalist Development, Tom Bottomore (Ed.), London: Routledge, (first published in 1910: Das Finanzkapital. Eine Studie über die jüngste Entwicklung des Kapitalismus, Vienna, Wiener Volksbuchhandlung). http://www.marxists.org/archive/hilferding/1910/finkap/index.htm. Hodgson, Geoffrey. 2006. What are institutions? Journal of Economic Issues, 40(1): 1-25.

Kaji, Gautam. 1998. Institutions in Development: the Country, Research, and Operational Challenges. In R. Picciotto and E. Weisner (eds.), Evaluation and Development: The Institutional Dimension. World Bank, Washington, D.C.: 3-6. Khakee, Abdul. 2002. Assessing Institutional Capital Building in a Local Agenda 21 Process in Göteborg, Planning Theory and Practice, 3(1): 53-68. Lakehal, Mokhtar. 2002. Dictionnaire d'économie contemporaine et principaux faits politiques et sociaux, Paris : Vuibert. Lounsbury, M., and M. A. Glynn. 2001. Cultural entrepreneurship: Stories, legitimacy, and the acquisitions of resource, Strategic Management Journal, 22( 6/7): 545 – 564. Loureau, René. 1972. L'Analyse Institutionnelle, Paris, Les Editions de Minuit, (Arguments), 298 pages. Lucas, Robert. 1988. On the Mechanics of Economic Development, Journal of Monetary Economics, 22: 3-42. Marx, Karl. 1867. The Capital, The process of production of capital, volume one, English edition, 1887, (first edition in German, 1867), Moscow: Progress Publishers. Micallef, André. 1969. « Essai sur une théorie du capital institutionnel collectif », Revue Economique, 20(1): 117-140. Narayan, D., and L. Pritchett. 1999. Cent and sociability: Household income and social capital in rural Tanzania, Economic Development and Cultural, 47(4): 871-897. North, Douglass. 1990. Institutions, Institutional Change and Economic Performance, Cambridge University Press. North, Douglass. 1991. « Institutions », Journal of Economic Perspectives, 5(1): 97-112. North, Douglass. 2005. Le processus du développement économique, Paris, Editions d’Organisations, (Traduction de Understanding the Process of Economic Change, Princeton University Press [2005] par Michel Le Séac’h). Oliver, Christine. 1997. Sustainable competitive advantage: Combining institutional and resource-based views. Strategic Management Journal, 18: 697-713. Palley, Thomas. 2001. Labour Standards and Governance as Public Institutional Capital: Cross Country Evidence from the 1980s and the 1990s, Unpublished Paper, AFL-CIO, Public Policy Department, Washington, and D.C.


http://www.amsreview.org/articles/fahy10-1999.pdf

http://www.amsreview.org/articles/fahy10-1999.pdf

http://www.formder.iamm.fr/

http://www.michel-garrabe.com/

http://www.brook.edu/es/commentary/journals/bpea_macro/papers/20040325_glaeser.pdf

http://www.brook.edu/es/commentary/journals/bpea_macro/papers/20040325_glaeser.pdf

http://www.marxists.org/archive/hilferding/1910/finkap/index.htm

http://www.marxists.org/archive/hilferding/1910/finkap/index.htm


Paul, Bénédique. 2006. Des organisations aux institutions : une introduction au concept de « capital institutionnel », University Paul Valéry – University Montpellier III, Master Degree dissertation (October). Paul, Bénédique. 2009. Institutional capital: A new analytical framework on theory and actions for economic development, Unpublished Paper, Communication at 7th Development Dialogue, at the International Institute of Social Studies (ISS), June, 2-3, The Hague, The Netherlands. Picciotto, Robert. 1996. What is education worth? From production function to institutional capital, World Bank: Human capital development working paper. Platje, Joost. 2008. An institutional capital approach to sustainable development, Management of Environmental Quality, Bradford, 19(2): 222. Portes, A., and P. Landolt 1996. The downside of social capital. The American Prospect Online, 7th Retrieved, from www.prospect.org/print-friendly/print/v7/26/26-cnt2.html. Rodrik, D., and A. Subramanian. 2003. The primacy of institutions (and what this does and does not mean), Finance and Development, 40(2): 31 – 34. Rodrik, Dani. 1999. Institutions for high-quality growth: what they are and how to acquire them, Harvard University Press. Romer, Paul Michael. 1986. Increasing Returns and Long Run Growth, Journal of Political Economy, 94(5): 1002-1037. Searle, John. 2005. What is an institution? Journal of Institutional Economics, 1(1): 1-22. Sirven, Nicolas. 2004. Capital Social et Développement, Bordeaux 4 University, PhD Dissertation in Economics. Solow, Robert Merton. et al. [Centre Saint-Gobain pour la recherche en économie]. 2001. Institutions et Croissance, Paris: Albin Michel. Trebilcock, Michael. 1996. What Makes Poor Countries Poor? The Role of Institutional Capital in Economic Development, Berkeley Olin Program in Law & Economics, Working Paper Series. Warner, Katherine. 2000. Forestry and sustainable livelihoods, in Forests, food security and sustainable livelihoods, Unasylva, Rome: FAO, 202: 1-71.

Wernerfelt, Birger. 1984. A resource-based view of the firm, Strategic Management Journal, 5: 171-180. Williamson, Oliver Eaton., 1975. Markets and Hierarchies:Analysis and Antitrust Implications, New York: Free Press. Williamson, Oliver Eaton., 1985. The Economic Institutions of Capitalism. New York: Free Press.

http://www.children.smartlibrary.org/NewInterface/www.prospect.org/print-friendly/print/v7/26/26-cnt2.html

Author Index

Armstrong, Thomas 25

Balough Robert 46

Chiu, I-Ming 63, 105

Chu, Hung M. 55

Cotherman, Tyler 71

Dunkin, David 71

Frickel, Beverly 19

Imandoust, Sadegh B. 130

Kara, Orhan 63

Kaufman, Chelsea 116

Kotcherlakota, Vani V. 19

Larson, Nicholas 71

Linn, Johnnie 75

Lundgren, Christopher 71

McCollough, John 86

Nugent, David 127

Paul, Bénédique 137

Reaves, Natalie 79

Sanders, William 71

Schmaeder, Jared 71

Sissoko, Yaya 93

Smith, Lynn 92

Stegman, Scott 71

Tenkorang, Frank 19

Zhu, Lei 55


pennsylvania economic associationaux.edinboro.edu/pea/pub/proceed2009/pea proceedings 2009.pdf ·...

Documents