ED 045 166

TITLE

TNSTITUTIONSPONS AGENCYPUB DATENOTE'

EDRS PRICEDESCRIPTORS

IDENTIFIERS

ABSTRACT

DOCUMENT RESUME

LI 002 385

Information Systems in the State of Minnesota1970-1980.Sheppard and Enoch Pratt Hospital , Baltimore Md.Minnesota State Pept. of Administration, St. Paul.Sep 70209p.

EDPS Price ME-T1.00 FC-.T10.9Automation, City Government, *Computer Programs,*Higher Education, *Information Systems, Planning,*State Government, *State Programs, Systems AnalysisMI DS, *Minnesota, Minnesota Information andDecisions Systems

In July, 1967, Governor Harold IeVandPr establishedthe Governor's Ccmmittee on State Information Systems. Thiscommittee, chaired by Gilbert A. Holmes, has had considerable impacton recent developments in information systems and computerization inthe State government and in higher education. In February, 1970, thecommittee recommended that the Governor and the Commissioner ofAdministration proceed with a study to develop a comprehensive Stateplan for information systems for prese-i:ation to the 1971Legislature. 7he study is presented in five parts: (1) Introduction,(2) Summary (also available as LI 002 3E6), (3) State Government, (LI)

Higher Education (also available as IT 002 387) and (F) LocalGovernment. Except for part I, a detailed table of contents appearsat the frcnt of each part of the report. (NH)

I

I

IIIIII

II

I

I

I

1

I 11 44:.

1 ti,

I

t.

U.S. DEPARTMENT OF HEALTH. EDUCATION& WELFARE

OFFICE OF EDUCATIONTHIS DOCUMENT HAS BEEN REPRODUCEDEXACTLY AS RECEIVED FROM THE PERSON ORORGANIZATION ORIGINATING IT POINTS OFVIEW OR OPINIONS STATED DO NOT NECES-SARILY REPRESENT OFFICIAL OFFICE OF EDU-CATION POSITION OR POLICY

INFORMATIONSYSTEMS

IN THE STATE OFMINNESOTA

1970-1980

STATE GOVERNMENT

HIGHER EDUCATION LOCAL GOVERNMENT

ACKNOWLEDGMENTS

Many persons were of material assistance in the study. In the Stategovernment portion of the study, the heads of the fifteen agencies ofthe government included in the study and legislative staff memberswere most cooperative and helpful. Particular assistance was given bythe EDP Coordinators of several of the large departments.

In the higher education part of the study, heavy assistance wasobtained from many administrative and academic personnel.Acknowledgement of that assistance is in the prefatory materialcontained in the higher education part of the report. In addition, Dr.Peter G. Roll of the University of Minnesota was granted a leave ofabsence to work for the Minnesota Higher Education CoordinatingCommission on the project. His work was crucial to the successfulcompletion of the task.

In the local government part of the study, assistance of the staffs ofthe League of Minnesota Municipalities and the Association ofMinnesota Counties was helpful. Commissioner Howard B. Casmey,State Department of Education, and members of his staff were ofparticular assistance in assuring understanding of school districtneeds and plans.

Valuable assistance was obtained from four consultants: Dr. WilliamH. Mitchel in certain aspects of local government informationsystems; Dr. Ronald W. Brady in higher education computer needs;Dr. Russell W. Burris in computer aided instruction; and Dr. RobertM. Hayes in reviewing computer needs of libraries in the State.

In addition to those mentioned earlier, particular thanks are due tothe following persons for their significant assistance (apologizing inadvance for possible omissions): G. A. Holmes, Chairman of theGovernor's Committee on State Information Systems; Richard L.Brubacher, Commissioner of Minnesota Department ofAdministration; Daniel B. Magraw, Assistant Commissioner,Minnesota Department of Administration; George L. Kieffer,Assistant Director for Planning and Coordination, InformationSystems Division, Minnesota Department of Administration; andGloria J. Gold, EDP Major Systems Manager, Information SystemsDivision, Minnesota Department of Administration.

Analysts International Corporation personnel assisting with the studyin addition to Dr. Peter C. Patton, the project director, were: John E.Hackley, David W. Kohr, John D. Lonergan, Janet M. Mathiowetz,William H. Osmer, John H. Pemberton, and Edward A. Schneider.

ii

I

[

1I

1

1

11111.01

PREFACE

In recent years the substantial implications of the development andimplementation of information systems and other computer basedservices have become quite clear to governmental jurisdictionsthroughout the' State of Minnesota. Public officials andadministrators in the State administration, in higher education, andin counties, municipalities, and school districts are taking initiative inbringing the tools of systems analysis and computers to bear on theproblems of government.

Both Governor Harold Le Vander and the Legislature have providedleadership in efforts to assure the orderly and efficient use of thesetools in the State government and in higher education. Moreover,they have indicated interest in the local government problems inthese areas.

In July, 1967, the Governor established the Governor's Committeeon State Information Systems. This Committee, chaired by GilbertA. Holmes, has had considerable impact on recent developments ininformation systems and computerization in the State governmentand higher education. Since its inception, the Committee has felt theneed fcr a comprehensive State plan for information systems and inFebruary, 1970, recommended that the Governor and theCommissioner of Administration proceed with a study to developsuch a plan for presentation to the 1971 Legislature.

The study was to consist of three parts: State government, highereducation, and local government. The time and funds available forthe study limited its scope and the specificity of its results.Moreover, precise future projections in this dynamic field are notonly difficult but probably impossible. For these reasons, the planspresented must be viewed as general guidelines rather than fixedrecommendations. Analysts International Corporation, Minneapolis,was selected to conduct the study. Dr. Peter C. Patton of that firmwas project director. The study began in March, 1970, and wascompleted in September. It was financed by the specialappropriation for computer acceleration and planning provided bythe 1969 Legislature.

Analysts International CorporationMinneapolis, Minnesota

September 1970

Description

PART I

PART II

TABLE OF CONTENTS

INTRODUCTION

SUMMARY

Page

I

IntroductionState GovernmentHigher EducationLocal Government

2-1

2-3

2-132-25

PART III STATE GOVERNMENT

Introduction 3-1

Current Status and Projected Requirements 3-7

Cost/Benefit Considerations 3-23Minnesota Systems Plan 3-29

PART IV HIGHER EDUCATION

Introduction 4-1

A Summary of Present Computing Facilitiesand Programs in Minnesota Higher Education 4-11

Goals and the Computer Resources Needed to Meet Them . 4-17Facilities and Costs to Meet the Goals 4-29Recommendations and Development Schedules 4-51

PART V LOCAL GOVERNMENT

Introduction 5-1

State of tha Art in Local Government 5-5Approac'n to Local Government Information Systems

and Computerization 5-23 7

Recommendations 5-27

NOTE: A detailed Table of Contents appears at the front of each part of the report exceptPart I.

iv

c

PART IINTRODUCTION

TABLE OF CONTENTS

Description

QUALITY OF LIFETHE CHALLENGE TO GOVERNMENT

Population and Wealth

Dedication of Resources

GOALS OF SYSTEMS ANALYSIS AND COMPUTERIZATIONDirect Cost/BenefitIndirect Cost/ BenefitInformation Systems

ECONOMIC PRESSURES

COMPUTER SYSTEMS IN GOVERNMENTSUMMARY COMMENTS

LIST OF FIGURES

Page

Figure Page

1.1 Population Trends of Minnesota and Metropolitan Area Projected Through 1-21985

1.2 Minnesota Per Capita Personal Income 1-3

1.3 Electric Utility Industry Sales in Minnesota 1-3

1.4 Value of EDP Systems Installed by Year of Installation (Estimated Purchase 1-6Prices) Scale AGovernment Purchases of Goods and Services (Current Prices ) -- Scale B

Table

1.1

LIST OF TABLES

Minnesota State and Local Public Employees, 1965-1969, Full-TimeEquivalent Basis

Page

1-7

1-i

INTRODUCTION

In this introductory part of the report, thegeneral nature of the current challenge togovernment is stated, with some indications of itsdimensions in Minnesota. Then a delineation ismade of the role to be played by systems analysisand computerization by information systemsin meeting that challenge. Finally, the economicpressures requiring systems analysis andcomputerization are discussed. Projections as tothe general increase in governmental expendituresfor computer hardware are reviewed.

The growth in the complexity and volumes ofrequirements for service placed upon governmentat all levels and the resultant costs are among thephenomena of our time. The growing awarenessof the interrelationships between and among theforces leading to requests for service, and a desireto look at the problems jointly rather thanseverally are, fortunately, further phenomena. InMinnesota, the State government, the highereducation complex, and local governments arecaught up in the growth, the complexity, theinterrelationships, the costs, and all of the otherproblems related to the almost overwhelming rateand nature of change in society.

In many ways Minnesota has been fortunate inits natural resources, in its leadership, in itsoolitical climate. A summarization of the resultsof the conditions and circumstances of life inMinnesota and a subjective ranking of the Staterelative to all others on a number of sociological,political, and economic factors is to be found in aMidwest Research Institute st u d y. 1 I n thewidely reported 1967 study, Minnesota wasranked second on overall quality of life. Certainlythe decisions made by the Legislature and localgoverning boards and the execution of thosetJecisions by the Executive Branch of the Stateand local governments over the years have had

considerable to do with the high rankingMinnesota enjoys.

THE CHALLENGE TO GOVERNMENT

Nonetheless, improved performance by all levelsand agencies of government in the State is

necessary to assist in coping with the rapid paceof societal change and the acceleration indemands for governmental service. The growth ofknowledge in the physical and social sciences hasbrought clear evidence of the difficulties and ofincipient disasters resulting from 'failure tounderstand the ecological or sociological impactof many decisions: the relationships of housing tocrime, of health to education, of airports to watertables, of taxation to economic development, andof the above and many more upon each other.

Population and WealthAll of this is compounded by the populationpressure in the State and the greater affluence ofits people. Since 1940, Minnesota's populationhas increased by nearly a million people,increasing from 2.79 million to 3.77 million, orabout 35 percent, as shown in Figure 1.1. Thegeographical shift of the population to urbanareas is at least as significant as its growth and is aprimary reason for the changing emphasis ingovernmental programs.

Indicators of the economic changes are many. Forexample, the Minnesota per capita personalincome, as shown in Figure 1.2, increased from$2,436 in 1964 to $3,608 in 1969, or 48 percent.Minnesota production of electrical energyincreased at approximately 12 percent a yearfrom 1964 (11 billion kilowatt hours) through1969 (19 billion kilowatt hours), or 73 percentsee Figure 1.3.

1. Wilson, John 0., Regional Differences in Social Welfare, Midwest Research Institute, 1967. 1-1

1940 1950 1960 1970 1975 1980 1985

Figure 1.1. Population Trends of Minnesota and Metropolitan Area Projected Through 1985(Metro Area Includes: Hennepin, Ramsey, Dakota, Scott and Anoka Counties)

Dedication of Resources

Heavy resources have, of course, been allocated toattack the problems facing society. For the fiscalyear 1968, total Federal expenditures were about$179 billion. Estimates for the 1971 fiscal yearare at the $200 billion level. Various projectionsindicate a total annual Federal budget of $230billion to $236 billion by 1975.

On the State and local government levels, totalannual combined expenditures were $86.5 billionin 1965, $94.8 billion in 1966, $105.9 billion in1967, $107.4 billion in 1968, $118.9 billion in1969, and $128.8 billion (estimated) in 1970, anaverage annual increase of 7 percent. Within this

1-2

period State government expenditures increasedan average of 12.6 percent per year for the1 9 65 -67 period. For Minnesota Stateexpenditures, the average annual rate of increasefor 1966 through 1971 is 11.9 percent. (See alsoTables 2.2 and 2.4 in Section 2 of Part III of thisreport.)

GOALS OF SYSTEMS ANALYSIS ANDCOMPUTERIZATION

Direct Cost/Benefit

Decision making in a dynamic society and thecost effective execution of the decisions may

I

I

I

I

1

I

I

I

Ii

'64 '65 '66 '67 '6I8

CALENDAR YEAR

SOURCE: SURVEY OF CURRENT BUSINESS, APRIL 1970, U.S. DEPT. OFCOMMERCE; OFFICE OF BUSINESS ECONOMICS

'69

Figure 1.2. Minnesota Per Capita Personal Income

both be materially aided by systems analysis andcomputerization. The assistance is provided intwo general ways. One may be characterized as_direct cost/benefit, that is, the activity provides aclearly measurable benefit that exceeds the costof attaining the benefit, such as minimizing theunit costs of processing income tax returns, driverlicense applications, and motor vehicle licenseapplications; preparing the payroll; manipulatinghighway engineering calculations; improving thequality and quantity of education by usingcomputers in instruction,. etc. Typically, these are_ operational activities as distinguished from the

SOURCE:ELECTRIC POWERSTATISTICSFEDERAL POWERCOMMISSION

2

'66 '67 '68 '69

OTHER

INDUSTRIAL

COMMERCIAL

RESIDENTIAL

Figure 1.3. Electric Utility Industry Sales In Minnesota

managerial activities of planning, research, andcontrol.

Indirect Cost/Benefit

A second type of action may be characterized ashaving indirect cost/benefit: for example, promptretrieval of data on a suspect before attemptingan arrest; instantaneous, comparative budgetaryanalysis for a legislative committee; analysis ofprobable impact of a change in tax law as an aidin determining tax policy. These provide a benefitthat is not clearly measurable but which is

1-3

estimated to exceed the cost of attaining it Thereis obviously real value in safeguarding a

policeman's life; in giving legislators promptlymore accurate and related information on whichto base appropriations; and in providing a sounderbasis of information on which to debate taxpolicy even though in the short run no directcost/benefits result. These are often managerialtype activities, although the first example is

operational.

Information Systems

The term information systems reflects thegrowing understanding of and interest in theapplication of systems analysis to theadministrative functions of society. The term andits increasingly frequent companion, decisionsystems,are in the midst of a semantical stormbecause of the relative newness of the formalizedconcepts and the rapidly changing state ofsystems analysis and computer technology.Generally, however, they reflect the assumptionsthat: (1) all rational decision making is based oninformation; (2) the better the information, thebetter the decision; (3) data gathering,information producing, and decision makingshould be systematized and then, to the extentfeasible, computerized.

It is emphasized here and should be recalledfrequently when reviewing this report that theterm "information systems" embraces both directcost/benefit and indirect cost/benefit activities. Itshould also be noted that operational applicationssuc!. as processing income tax returns, can aisogenerally provide for the data base needed tosatisfy the administrative and planning needsrelated to that function of State government ifthe systems analysis and design properly considerboth types of requirements.

Further, the term systems analysis is used tomean an all-encompassing study of all factorsrelating to a function or an activity including, butnot limited to, objectives, organization structure,policies, programs, and the required proceduresand methods, both manual and mechanical. Thecomputer itself is simply one of many managerial

1 -4

or mechanical tools that systems analysis maydesignate as being appropriate to accomplishcertain objectives. To put this another way, thecomputer is a servant of systems analysis, notvice-versa.

Because of its centrality to the whole question ofinformation systems, further comment is offeredon the nature of a data base. A data base is a setof data upon which an information system, orsystems, are based. Often data bank is usedinterchangeably with data base. Quite obviously,a system can produce for informational ordecision purposes only what is in the data base orwhat can be produced as a result of manipulatingtwo or more pieces of data in the data base. Inother words, to the extent that data are in thebase, they can be retrieved "as is" or reformattedor manipulated as desired for operational ormanagement purposes.

Although a simple concept, it is crucial to anysystems analysis. For example, historically thedata used for analytical and planning purposeshave been collected in an ad hoc fashion. Most ofthe time of local government employees engagedin such efforts lies in collecting the data, leavingrelatively little time for analysis. And even thenthe data are often outdated or incomplete. At thesame time, however, much of the data is generallyin the up-to-date operating files of thegovernmental agencies. In fact, most of the dataneeded for analysis and planning are inoperational files.

Thus, the key consideration in systems design isto assure that the data elements necessary for allpurposes are in fact in the data base. A few shortyears ago, it was almost heretical to think in suchterms because of the innovative concepts and thedata storage costs involved. In the near future itwill be commonplace; the new "heresies" willarise with combinations of the new technologiesin computing, microfilming, lasers, and chemicalmemories.

ECONOMIC PRESSURES COMPUTER SYSTEMS IN GOVERNMENT

The number of services rendered by governmenthas been increasing; at the same time the workloads in virtually all services have been expanding.Intergovernmental requests for data likewise havebeen increasing and in fact in some cases are acondition of funding. Governmental operatingcosts, too, have been increasing generally, andmore recently inflation has sharply accentuatedthis trend.

The costs of computer hardware per unit ofoutput, however, have been dropping ever since

,.

the first computers were introduced. Radical- decreases have been experienced and are still

occurring. Recently, a major computer companyhas announced a new line of computers withestimates of hardware cost reduction per unit ofoutput ranging around thirty percent.

The combination of generally increasing costs anddecreasing computer hardware costs is inevitablyforcing greater emphasis on systems analysis andcomputerization for direct cost/benefits. Thevirtually limitless ability of computers to relateand analyze the vast and increasinglycomprehensive data bases and to produceinformation to aid in decision making areproducing significant indirect cost/benefits. Thesituation is well summarized in the following:

In just a few short years, questions aboutwhether or not a company should acquire acomputer system have virtually disappeared.Today's questions are more nearly "How big acomputer do I need?" or "How do I make mycomputer system work better?" In a hundredpercent of the top five hundred companies, andvirtually all the rest of American business ofany size at all, the systems function has alreadybecome as necessary, as much an irrevocablefactor of daily business life as marketing,distribution, or production. This instrument(and the extraordinary new notions about therunning of a business which come with it) is nolonger a luxury for those few who can afford it.It has become a necessity for competitivesurvival. 2

Figure 1.4 shows two sets of data from whichsome important observations can be made. Twoof the curves show the annual expenditures forgoods and services by the Federal government andby state and local governments. Spending by stateand local governments for goods and servicessurpassed that of the Federal government in1965; and in 1970, state and local governmentsspent an estimated $120 billion while the Federalgovernment spent only $100 billion. For the1971 through 1980 period, spending by state andlocal governments will increase more rapidly thanFederal spending in terms of both the annual rateand the annual amount of increase. By 1980 stateand local government expenditures of this typeare projected at $278 billion annually, whileFederal spending is projected at $155 billion.

The estimated net annual value, at purchase price,of EDP systems installed in the Federalgovernment and in state and local governments isalso shown in Figure 1.4. Obviously, governmenthas become a large user of computers and willincrease its use substantially. Federal governmentuse far exceeds that of all state and localgovernments combined. In 1969, for example, thevalue of Federal government computer hardwareis estimated at between $800-850 million whereasthe comparable figure for state and localgovernment is $250-300 million. Although thereare no published projections of overallgovernmental computer expenditures, it appearsreasonable to assume that such expenditures willincrease at a rate at least equal to the increase inspending for goods and services. In fact, for manyof the reasons mentioned in this report and withthe knowledge that information systems andcomputerization are in their relative infancy, it ishighly probable that the slopes of the computerexpenditure curves will be significantly steeperthan the slopes of the total goods and servicesexpenditures curves.

It is also indicated that the state-local computerexpenditure curve will increase its slope relative

2. Ditri, Arnold E. and Wood, Donald R., The End of the Beginning, July 1969. 1 -5

DOLLARSIN MILLIONS

DOLLARSIN BILLIONS

1,000 250

800 200

600 150

400 100 -

200 50 -

0 0

FEDERALGOVERNMENT(SCALE A)

STATE AND LOCALGOVERNMENT(SCALE B)

STATE AND LOCALGOVERNMENT(SCALE A)

120

FEDERALGOVERNMENT(SCALE B)

35 e

'64r

'66

73 80100

LOCAL GOVERNMENT ONLY

'68 '70 '72 '74

CALENDAR YEARS FOR "A" SCALE

FISCAL YEARS FOR "B" SCALE

'76 '78 '80

Figure 1.4. Value of EDP Systems Installed by Year of Installation (Estimated Purchase

Prices) - Scale A

Government Purchases of Goods and Services (Current Prices) - Scale B

to the Federal curve as the state and localgovernments move to close the computerizationgap between the levels of government.

One measure of the relationship that exists amongthe three Minnesota entities covered in this report

State government, higher education, and localgovernment is the full time equivalentemployment shown in Table 1.1. Although thereis not necessarily a direct reiationship betweenthe number of employees and need forinformation systems and other computerization,there is indeed some relationship. It is clear thatlocal government is where the greatestopportunity lies, particularly in elementary andsecondary education, in information systems

1-6

needs and in computerization as an instructionalaid. The four year growth rates in all areas giveurgency to the problem.

SUMMARY COMMENTS

The development of systems analysis andcomputerization technology over the past fifteenyears has had a tremendous impact on all phasesof business, edlucation, government, industry, andscience. The State of Minnesota has been in themidst of these developments. Its computerindustry ranks high in the nation. Its Stategovernment is among the leading states in theeconomical development of information systems.Its higher education establishment has moved

Table 1.1. Minnesota State and Local Public Employees, 1965-1969, Full-Time Equivalent Basis

-ar

Education Non-Education Total

LocalState LocalState andLocal State Local

State andLocal State

State andLocal

1965 11,728 55,878 67,606 22,043 41,775 63,818 33,771 97,653 131,424

1966 13,257 58,994 72,251 22,621 42,560 65,181 35,878 101,554 137,432

1967 14,280 60,016 74,296 22,771 43,420 66,191 37,051 103,436 140,487

1968 15,483 64.185 79,668 24,316 44,165 68,481 39,799 108,350 148,149

1969 17,304 65,407 82,711 24,713 45,513 70,226 42,017 110,920 152,937

4-YearGrowth 47.5% 17.1% 22.3% :12.1% 8.9% 10.0% 24.4% 13.6% 16.4%

Source: Public Employment in the United States 1.965-1969U.S. Department of CommerceBureau of Census

ahead in implementing computer capability tobuttress its instructional, research, and publicservice activities and to aid in theiradministration. Its local governments have begundevelopment in this broad area and in one notableeffort have set up a coordinated program.

Building on this background, the broad goal ofthis study is to develop a plan for informationsystems development and computer utilization inthe State of Minnesota for State government,higher education, and local government. Keyelements within this plan include theidentification of major areas of application andassociated benefits, cost projections, andrecommendations for legislative andadministrative action required forimplementation.

The commentary and data presented in thisintroduction to "Information Systems in the

State of Minnesota 1970-1980" have reviewedsome of the forces surrounding the developmentof information systems and the increasing use ofcomputers nation-wide. The goals and concepts ofinformation systems, systems analysis, and databases were discussed, and the level ofgovernmental expenditures for computerhardware was outlined. Recommendations aresummarized in Part 2; Parts 3, 4, and 5 detail theresults of the state government, higher education,and local government studies, respectively.

A word of explanation is in order concerning theplacement of the area vocational technical schoolsin the report. The AVTS's are established underthe authority of school districts and the StateBoard of Education. They are also considered aspart of higher education of the State and as suchare of concern to the Minnesota Higher EducationCoordinating Commission. Discussion andrecommendations concerning them are in boththe higher education and local government partsof the report.

I

1 ^'

iN,

iI

)

PART IISUMMARY

TABLE OF CONTENTS

Description Page

INTRODUCTION 2-1

STATE GOVERNMENT 2-3

Current Status and Projected Requirements 2-4

Information Systems Costs 2-4

Cost/Benefit 2-5

The State Goveriiment Systems Plan 2-6

HIGHER EDUCATION 2-13

Computers and the Objectives of Higher Education 2-13

Existing Computing Facilities and Programs 2-14

Goals and Computer Resources Needed to Meet Them 2-15

Facilities and Costs to Meet the Goals 2-17

Recommendations and Development Schedules 2-18

Goals and General Recommendations 2-19

Recommendations for Regional and Statewide Facilities and Programs 2-19

Recommendations Related to Individual Systems 2-20

Recommendations Addressed to the Higher Education 2-21Coordinating Commission

Development Schedules 2-22

LOCAL GOVERNMENT 2-25

Status in Minnesota 2-26

Local Government Recommendations 2-27

General Recommendation 2-28

Organizational Requirements 2-28

Approaches to Cooperative Centers 2-30

Area Computer Centers 2-30

Orientation and Training 2-31

Funding 2-31

Implementation Suggestions 2-32

2-i

INTRODUCTION

This part of the report summarizes Part I

Introduction and the three substantive parts ofthe report: Ill State Government, IV HigherEducation, V Local Government. It is intendedto serve those who cannot devote time to readingthe entire report and to be available fordistribution to those not requiring the fullreport.

The growth in the complexity and volume ofrequirements for service placed upon governmentat all levels and the resultant costs are among thephenomena of our time. The growing awarenessof the interrelationships between and among theforces leading to requests for service, and a desireto look at the problems jointly rather thanseverally are, fortunately, further phenomena. InMinnesota, the State government, the highereducation complex, and local governments arecaught up in the growth, the complexity, theinterrelationships, the costs, and all of the otherproblems related to the almost overwhelming rateand nature of change in society.

Although Minnesota enjoys a high "quality oflife" rating among the states, in part attributableto its State and local governments, improvedperformance in all governmental operations isnecessary to cope with the rapid pace of societalchange and growing demands for service.Accentuating this are the significant populationgrowth and the increasing affluence of Minnesotacitizens. Heavy tax resources have, of course,been allocated to cope with these problems andtheir demands.

The tools of systems analysis andcomputerization hold out hope for improvinggovernmental operations, governmental decisionmaking, and the effective execution of thosedecisions. Realization of these hopes can comethrough direct and indirect cost/benefits affectingoperational and management performance.

The concepts of information systems and decisionsystems have developed from the growingunderstanding of the application of systemsanalysis to the administrative functions ofsociety. These concepts are based on theassumptions that rational decision making is

based on information, that the quality of thedecision is based in large measure on the qualityof the information, and that much decisionmaking is capable of being systematized.

Systems analysis by its nature must include allfacets of an activity or function, not only itseasily computerizable aspects. Of particularimportance is analysis of the informationrequirements of the data base that is necessaryto provide that information. Computers havemade possible the development and managementof comprehensive data bases. Moreover, theeconomics of generally increasing costs on onehand and decreasing computer costs on the otherhand are forcing the rapid development ofinformation systems and computerization fordirect and indirect cost/benefit. This trend isevident in the present expenditures for computerhardware in government; it can be expected thatsuch expenditures will increase more rapidly thanmost other types of expenditures and that stateand local expenditures for these purposes willincrease at a faster rate than in the Federalgovernment.

Each of the three segments covered in this reportstate government, higher education, and local

government has considerable significance to thefuture of the State of Minnesota. It would befutile to suggest that one is qualitatively moreimportant that the others. In a quantitative sense,the full-time equivalent employment datasummarized below is one measure of the relativesignificance.

2-1

Although there is not necessarily a directrelationship between the number of employeesand need for information systems and othercomputerization, there is indeed somerelationship. On that basis, local government iswhere the greatest opportunity appears to lie,particularly in elementary and secondaryeducation where both administration andinstruction requires computerization.

Viewed on a broad basis, however, the real need isfor comprehensive information systems to aid (a)the legislative bodies in setting priorities andfunding programs and (b) the administrative

officials in executing those programs. The fouryear growth rates in all areas are evidence of thegrowing urgency to make optimum use of thesenew technologies in administration andmanagement.

It should be noted that discussions andrecommendations concerning the area vocationaltechnical schools (AVTS) are found in both thehigher education and local government parts ofthe report, reflecting the AVTS's relationshipswith both the Minnesota Higher EducationCoordinating Commission and the State Board ofEducation.

MINNESOTA STATE AND LOCAL PUBLIC EMPLOYEES 1969 FULL-TIME EQUIVALENT BASIS

Year

Education Non-Education Total

State LocalState andLocal State I Local

State andILocal State I Local

I S tilt ec aal and

1969

4-YearGrowth

,

17,304

47.5%

I

65,407

17.1%

I

82,711

22.3%

I

24,713

12.1°A

I

45,513

8.9%.

I

70,226

10.0%

42,017

24.4%

.

110,920

13.6%

152,937

16.4%.

Source: Public Employment in the United States 1965-1969U.S. Department of CommerceBureau of Census

2-2

t

1

1

1

i

T

4.111 ;

I

I

STATE GOVERNMENT

In 1957, the Minnesota Legislature laid the basisfor a gradual centralization of the State's dataprocessing facilities by setting up a CentralServices Division in the State Department ofAdministration. In 1960, most of the State'stabulating installations were merged into thatdivision. A principal result of that beginning andthe subsequent developments i s that Minnesotahas avoided much of the extensive and expensiveproliferation of computer equipment, personnel,and space that inevitably accompanies separatecomputer operations.

Like other states, however, the importance of thesystems analysis function has not been clearlyrecognized until recently. Although computerapplications were designed, insufficient attentionwas given to management information; andsystems analysis was generally limited to theeasily "computerizable" characteristics of a

function.

I n 1966, the State Planning Agency drewattention to the growing need for informationsystems for each segment of the Stateadministration. In 1967, the Governor and theLegislature agreed to a $500,000 specialappropriation for computer acceleration and setup a Computer Services Division (CSD) in theDepartment of Administration. The Governoralso established the Governor's Committee onState Information Systems, a committee whichhas played a major role in subsequentdevelopments including calling for this report.

The Governor, by Executive Order in 1968, gavethe Computer Services Division broad authorityand responsibility in the State's computer efforts.Significant action took place on many fronts inthe next two years including another specialappropriation by the 1969 Legislature, this time

in the amount of $900,000 with. $300,000earmarked, for higher education. in 1970, theGovernor issued another Executive Order on thesubject of information systems, further clarifyingthe State's direction. Shortly thereafter theCommissioner of Administration reorganized theComputer Services Division, equipping thedivision to provide better service to its users andto set up a Local Government Unit to assist inlocal government coordinating efforts. Shortlythereafter, the computer functions of theDepartment of Highways were merged into theComputer Services "Division, leaving only theDepartment of Manpower Services with a separatecomputer installation.

Since 1967, significant progress has been made inbuilding the Computer Services Division. Thedivision was strengthened in structure, personnel,and equipment and converted to a heavilymulti-programmed operation under full operatingsystems. Further, extensive progress has beenrealized in coordinating systems and computerdevelopment in State government and highereducation.

This part of the study included the Legislature,twelve agencies of the State government, and thethree principal retirement agencies. Thecombination of agencies accounts for over ninetypercent of the budget, employees, and computerexpenditures of the State, exclusive of highereducation. Much of the study effort was devotedto interviews with management personnel of theagencies tt, determine information needs, bothpresent and prospectively. The results of theinterviews were summarized for each department.Inventories of personnel, equipment, andapplications were also obtained and analyzed.Departmental informational interrelationshipswere identified.

2-3

CURRENT STATUS AND PROJECTEDREQUIREMENTS

From its first IBM 650 computer in 1959, theState computer capacity now consists of two IBM360-50's, an IBM 360-40, an IBM 360-25, an IBM1401 (at Manpower Services to be replaced soonby an IBM 360-40), and a UNIVAC 418-11. Thefirst three of these are multi-programmed underfull operating systems. One IBM 360-50 and theUNIVAC 418-11 operate in communication mode,the 418-11 serving as the basis for MINCIS(Minnesota Crime Information System) andinterfacing with the 360-40 on which both thedriver I icense and motor vehicle files are

maintained.

Four of CSD's largest applications are heavilyinvolved in on-line input, updating, and/or dataretrieval: M INC IS, driver's license, motor vehicles,and income tax. The Division's planning assumesthat in time most of the State applications will beavailable for on-line update and retrieval. A widevariety of applications is handled by the Divisionwhich services most agencies of the Stategovernment.

Effective management of the systems andcomputer function, of considerable interest toLegislators and top administrators alike, is a

complex task. Principal factors are the quality ofsystems analysis, personnel performance, andefficiency of equipment and its use.

A comprehensive systems approach is necessary indesigning cost effective information systems. Toooften in the past such an approach has beenmissing, not only in state government but nearlyeverywhere. An entire system must be reviewed,not just a portion that someone wants to "put onthe computer." This is an extremely difficult taskbecause the management of any significantactivity is complex and the systems necessary toassist with that management are also complex.The task obviously cannot be done effectivelyunless each manager assures himself that thesystem is in detail what he needs as an example,not just to account for tax collections, but to givethe manager and his principal subordinates what

2-4

they need for the research, planning, and controlaspects of tax administration.

The ratio of hardware cost to totalcomputerization costs in administrative dataprocessing approximated 2 to 1 a few years ago.Now it is typically about 1 to 2. Thus, themanagement of computer personnel, always amajor task, has taken on added relativeimportance. Controlling the personnel turnover ofcomputer professionals with proven competenceand comprehensive knowledge of large andcomplex systems became critical.

Particular interest has been expressed in theefficiency of computer equipment and itsoperation. It requires an evaluation of manyfactors including computer speed and memorysize, multi-programming, utilization, scheduling,peripheral equipment, and productive use.

Utilization is a particularly difficult evaluationsince there is no one definition of utilizationwhich f its all situations. Minnesota Stategovernment computers have been generallyutilized heavily by any standard in recent yearsso heavily that customer service has beenoccasionally impaired. Recent additions to theState's computer hardvvare have provided a slightbut temporary cushion for additionalapplications. Completion of all aspects of therecent reorganization of the State computerfacilities is necessary to maximize efficiency ofoperation.

Information Systems Costs

The approximate costs of the State systems andcomputer operations have increased from$2,000,000 in 1966 to $5,100,000 in 1970. Thelast three years have shown an increase of about32 percent a year. The figures are incompletebecause costs have not been accumulatedstatewide, a deficiency to be corrected by a newState accounting system now under development.

A number of factors are expected to lead toconstantly increasing computerization spendingover the coming years. These include economies

of computerization versus manual methods;increases in the volume of State business;additional State services; data requirements ofFederal agencies; increasing sophistication insystems design and comprehensiveness ofcomputer services; more productivity fromsystems and programming personnel; and increasein the State's service to local governments. On theo.:her hand, the growth in comprehensiveir formation systems will put a downwardpressure on other costs, both through direct andindirect cost/benefits. To obtain data on which toproject future budgets, several factors werere ii ewed.

A large number of applications wereidentified in the course of the study. It wasestimated that 497 applications would bein operation by 1975, compared with 332at present. To reflect increasing complexityand comprehensiveness of newapplications, a weighting factor of 2 wasassigned to the 265 expected newapplications to permit comparison with thepresent 332. The net results was that the1975 application load on a basis equivalentto 1970 would be 762.

The trend in the State's expendituresshown previously amounted over the pastthree years to an average annual increase of32 percent.

A comparison was made of the growthtrend in the overall State expenditures andthe expenditures for State computeroperations. The latter as a percent of theformer shows a significant increase over thepast five years, a trend almost certain tocontinue.

Computer hardware costs among severalstates (Illinois, Indiana, Iowa, Michigan,Minnesota, Ohio, and Wisconsin) show thatMinnesota is (a) in a middle group on costper resident although low until 1968 and(b) lowest in cost per state employee in1967 and 1968 but in 1969 higher thanOhio and equal to Michigan. Such figuresare merely indicative since qualitative andperformance considerations are omitted.( I n fiscal 1970, Minnesota Stategovernment information systems costs

approximated $1.35 per resident and$207.00 per State employee.)

Although no specific budget recommendationsare made, it is suggested that a curve reflectingthe estimated increases in applications would bethe most likely base for realistic budget planningwhich, if used, would result in an annual increaseover the next five years of about 24 percent.Thus, the 1975 fiscal year would show a total of$14,900,000 compared with the 1970 total of$5,100,000. Budget projection beyond thatperiod was thought to be totally infeasible.

COST /BENEFIT

As a generality, information systems and othercomputerization efforts will not survive unlesscost effective. The measurement of the benefit,however, has often not been made at all or insome cases made incompletely. Among thereasons for this are: costs of existing systems arenot known; follow-up on new systems to assurecost/benefit realization often is ignored; newsystems bring benefits not previously available;impact of higher workload may be overlooked;inflation's effects may be ignored. Because of thegeneral lack of detailed feasibility studies prior tocomputerization and the deficiencies in the Stateaccounting system it has not been possible topresent an authoritative statement ofcost/benefits in Minnesota.

In one example, some facets of income taxprocessing, it is conservatively estimated thatdirect personnel cost avoidance since 1964amounted to the annual amount in 1969 of$230,000. There is substantial evidence that theunit costs of income tax return processing,excluding effects of salary schedule increases,decreased by about 13 percent in that period andthis in spite of more extensive and complexprocessing requirements. A well-documentedstudy of certain Department of Highwaysengineering tasks showed that pre-automationcosts of $2,748,875 were reduced bycomputerization to $1,129,971, a reduction of$1,618,904 or nearly 59 percent of the originalcost. The indirect cost/benefits, such as are to be

2-5

found in MINCIS and many other systems, willprobably far exceed direct cost/benefits.

Minnesota's decision to centralize its computercentral processors produces direct cost/benefit. Ahypothetical analysis generated a theoreticalpotential savings of about one-third. Even a halfof this potential would provide a large savingwhen measured against the probable $15,000,000budget in 1975. But effective administration isnecessary to obtain the advantages of suchcentralization while avoiding its potentialdisadvantages.

THE STATE GOVERNMENT SYSTEMS PLAN

A plan outlined here for Minnesota Stategovernment and designated as MIDS (MinnesotaI nformation and Decision Systems) aims toachieve two primary objectives:

i

To make available to the Legislature and toofficials of the Executive Branchcomprehensive and accurate information ina form and at a time to be of maximumvalue in decision making and research.

Provide to all State personnel the mosteffective framework of systems,procedures, and methods within which tocarry out the operations of the State.I

1

I

The implications of implementing MIDS areenormous. In its broadest terms, MIDS meansdefining, designing, and installing comprehensivesystems for each area of government to serve asthe framework for attaining optimumeffectiveness of State operations. A majorrequirement is the ability to associate andinterrelate the data bases of the informationsystems serving major functional areas ofgovernment. Corollary to this is the installation ofprocedures to assure the integrity andconfidentiality of data.

Action necessary to implement MIDS is

recommended under five headings:

Comprehensive systems approach

Organization structure

2-6

Guidelines for development

Security and confidentiality of data

Intergovernmental coordination.

A more comprehensive systems approach to theproblem of the State. The recommendations are:

1

Assignment of overall responsibility fordesign, coordination, and implementationof MIDS to one agency of the government.Executive Order No. 56 should beinterpreted to include all systems analysis(not just the computerizable aspects) andincorporated in the State statutes.

More effective execution by the topexecutive in each agency of Stategovernment of his responsibility forclarifying the objectives of each functionof his agency and for specifically approvingthe general systems required to reach thoseobjectives.

Attendance by top management personnelof all agencies of State government atintensive courses in systems analysis,information and decision systems concepts,and computer familiarization. Legislatorsand legislative staff members should beinvited to participate.

An organization structure effective incoordinating the design of the many majorsystems, building efficient operational systems forall areas of State government, attaining maximumproductivity from the State's systems andcomputerization efforts, and providing assistanceto Local government. The recommendations are:

The State information systems functionshould be strengthened by improvedorganization and by statutorycentralization in that organization ofauthority over development and operationof information systems andcomputerization. The growing importanceof information systems to the operationand management of the State governmentcalls for improving the organization of thatfunction either by further strengtheningthe structure within the Department ofAdministration or by establishing a new

t

rt

Ij

Department of Information Systems. Forthe organizational alternative selected, themost important requirement is thatappropriate responsibilities and authoritybe given to the function. The acronym DISof this report refers to a Division ofInformation Systems, and itsadministrative officer is referred to as

Director of Information Systems.

All systems and computer functions of theState administration should beadministered or controlled by NS. Thisrecommendation is consistent with thebasic organizational concepts beingimplemented in Minnesota. Internalorganization of the division should be builton the Computer Services Division andshould consist of these six sections:

1. Systems and Programming withpresent responsibilities broadened toinclude all systems analysis and design,not just computerizable segments ofsystems.

2. Planning and Coordination withresponsibilities for research, standards,staff development, and highereducation coordination as presentlybut adding two units. One is a

Methods Unit to carry out certainsystems functions such as formscontrol, records management, andwork measurement and to providecertain systems specialization such asin microfilming. The second is a

Management Science Unit to maintaina comprehensive index of all Statefiles, advising and assisting Statepersonnel in how best to obtain andanalyze data from those files, and

introducing the use of operationsresearch techniques in the analysis ofState problems.

4. Telecommunications Sectionadministering all communicationsfunctions of the State.

5. Local Government Section withprimary responsibility to serve as staff

to the Intergovernmental Commissionon Information Systems recommendedin the local government part of thisreport.

6. Administrative Services Sectioncarrying out the tasks generallyassigned to such divisions in the State.Of particular importance in DIS is thecontrollership function because of theneed for sound data in feasibilityanalysis and in the cost accountingnecessary to the Division's operation.

Expenditure Control. Legislation should bepassed to continue in DIS the authority inExecutive Order No. 56 to review andapprove all proposed expenditures by anyagency of the State government forsystems development or computerization(equipment, personnel, programs, space,systems design, consultants, service

bureaus, etc.) and extension of theauthority to cover all systems activity.

Governor's Committee. Legislative action isrecommended to formalize the presentGovernor's Committee on StateInformation Systems or a similar group toadvise the Governor, the Legislature, andDIS on MIDS plans and progress and torecommend on the expenditure ofdevelopmental funds.

Agency Systems Coordination. Thedevelopment and implementation ofmanagement systems is a complex matterrequiring the attention, coordination, andauthority of top management of the useragency. Accordingly, a systems coordinatorshould be appointed by each major agencyof State government, reporting directly tothe agency head or to his chief deputy orassistant.

Communications Advisory Committee.Proliferation of communications systems,leading to duplication and overlapping, willresult unless coordination is effected. As avehicle for assuring the neededcoordination, it is recommended that aCommunications Advisory Committee be

2-7

established, chaired by the DISTelecommunications Section head.Membership should include representationfrom higher education and localgovernment; major State users such as theDepartment of Highways; and Statecomputer, radio, and telephonetechnicians.

Planned, controlled development of systems,procedures, and methods computer-based orotherwise to optimize the value of data basesand information obtainable therefrom fordecision making and research, and to provide themost efficient tools for State employees to carryout the activities of State government.

First and foremost is the need for DIS andits user agencies to work togetherefficiently. It is axiomatic in this field thatonly truly cooperative efforts aresuccessful. Specific action that should betaken includes:

2-8

1.

2.

3.

Establishment of a MIDS User'sCommittee to serve as a primarycommunication device in theimplementation of MIDS. Thecommittee should consist of theSystems Coordinator from each of themajor agencies of government, the DISDirector, and the appropriate sectionheads of DIS.

Agency Systems and ComputerPlanning accomplished jointly by theagency Systems Coordinator and theDIS Major Systems Manager for thatagency. It is strongly emphasized thatthe Systems Coordinator, as agencyrepresentative, must assume the keyrole in planning so the plan will reflectthe agency's real needs, not the DISview of those needs.

Formalized and continuing trainingprograms for agency personnel at theoperating level. A spectrum of subjectsshould be covered including systems;controls; systems conversiontechniques; input alternatives;computer techniques for data storage,management, and retrieval; general

computer capability; etc.

4. Formalized methods ofcommunication between DIS and itsusers. These should include proceduresfor requisitioning service, for advisingon progress, and for reporting onproblems.

5. DIS Newsletter. A simple newsletter,prepared monthly, could be helpful inthe continuing orientation of agencypersonnel to the accomplishments ofDIS users and to DIS capabilities.

The following guidelines should be adoptedfor MIDS:

1. A systems philosophy aimed atassisting the decision making processas contrasted with simply mechanizingor otherwise improving a productionsystem.

2. A systems design to permit integratedsystems. This requires thedevelopment of standard identificationsystems for persons, organizations,physical assets, etc.; standard datadefinitions; and standard computertechniques for file management.

3. Cost/benefit analysis to establish

system feasibility.

4. Computerization developed in an

environment based on centralizedcentral processing units and whateverremote terminals are necessary to serveusers.

5. Operational objectives of on-line filesfor real-time retrieval of informationana on-line entry and update of dataoriginating where feasible in the useragency.

6. Attaining maximum output from bothequipment and personnel.

The centralization of administration andcontrol of systems development andcomputerization should be completed.Specific action necessary to complete thecentralization includes:

1. Transfer of the computer functions ofthe Department of Manpower Servicesto DIS. Appropriate systems personnelshould remain assigned to MSD, as

i1

with other major departments, toassure systems design responsive to theneeds of that department.

2. Developing a schedule for phasing outtabulating machine installations infavor of computerization, includingprovision for remote terminals whereappropriate.

O Systems activities must be economicallyjustified. Each systems design which isrecommended for implementationwhether to be on a computer or otherwise

must be analyzed, and the feasibility ofimplementation determined. If feasible,and if implemented, then a plannedfollow-up must be completed to assure thatthe "pay-off" was realized or to discoverwhy not.

Adequate physical facilities must beprovided DIS. At present the spaceallocated falls short of minimumrequirements. The two computerinstallations are on opposite sides of theCapitol Mall, preventing optimum use ofthe computer equipment. Both computerrooms are jammed so full of equipment asto impair the efficiency of computeroperating personnel. Neither room is builtso that acceptable security standards canbe assured. Space assigned to theCentennial Building keypunch section is

extremely crowded and generally notconducive to maximum productivity.Space for computer professional personnelis at the crisis point.

Personnel costs are constantly growing as apercent of the overall costs of systemsdevelopment and computerization, withthe result that increasing emphasis must beplaced on proper management of the staff.Guidelines in personnel managementshould include: (a) establishment andenforcement of systems design andprogramming standards; (b) scheduling andcontrolling all systems and programmingprojects; (c) conducting formalizedperiodic performance reviews; (d) installingtraining and development programs; (e)assuring competitive posture inclassification and salary structure; (f) useof existing systems when practicable; (g)

selectively engaging consultants asspecialists or to provide overloadassistance; (h) assuring comprehensivedocumentation to maintain continuity.These guidelines in the main have beenfollowed by CSD. The quality of resultswill be in direct proportion to thesupervisory attention given to theirenforcement.

Computer capabilities of the State willrequire expansion as the applicationsincrease. Centralization provides theframework within which that expansioncan be done most efficiently.Recommendations related to equipmentrequirements are: (a) the "Model StateContract" for leasing or purchasingcomputer hardware should continue to beused by the State; (b) the requirements forcompatibility must be given appropriateweight in obtaining additional orreplacement equipment; (c)communications networks should becentered on one computer with strongcommunications handling capabilities; (d)appropriate back-up equipment should beavailable; (e) to attain the substantialsavings available through purchasing ratherthan leasing, computers and relatedequipment should be purchased outright orunder lease purchase plans when there isreasonable expectation that equipmentobsolescence will not occur prior to thebreakeven point; (f) peripheral equipmentfrom other than the vendor of the centralprocessing unit should be obtained wherecost effective; (g) computer utilizationrecords should be kept on all equipmentand should reflect each of the methods ofmeasuring utilization; (h) special emphasisshould be placed on evaluation of methodsand equipment for input preparationbecause of its large potential for costreduction.

A number of further actions are necessaryin the implementation of MIDS.

Data Dictionary and Standardizationof Data Elements. In the recentreorganization, CSD set up the newposition of Manager-Standards andQuality with the major objective of

2.

3.

establishing a data dictionary and,corollary thereto, standardizing thestructure of data elements andidentifiers. Because there is probablyno single function more important tothe long-range objectives of MIDS, thisprogram must be given top priorityand top level support.

Controls. No principle incomputerization is so basic as the needfor controls. GIGO (garbage ingarbage out) has scuttled morecomputer installations and damagedmore user relations than all otherproblems. It is essential that adequatecontrols be established; continuedfailure or unwillingness by any useragency to provide control informationor by DIS to establish and administeraccuracy controls is a matter forprompt discussion at theCommissioner level.

Standards for Systems Analysis andProgramming. Standards should beestablished for all elements of workperformed by systems andprogramming personnel forexample, documentation of existingsystems, feasibility analyses,programming conventions, andprogram documentation.

4. Unified Operational Envircnment.Many of the remaining secondgeneration programs should bereprogrammed to the third generationcomputers. In addition, a single set ofprogramming conventions, similarityof operating systems, andstandardization of teleprocessing andgraphics network and interface controlprograms must be accomplished.

5. Customer Billing. Development of anew billing system should be givenhigh priority as a major aid to the DISinternal management informationsystem and as an aid in user relations.The new system must, of course, beconsistent with both the new Stateaccounting system and the Stateprogram budget requirements.

Adequate financing is essential. It is

recommended that future funding for allnew systems development follow theprecedent set in the 1969 specialappropriation. That is, all funds forproposed new applications orresystematizing of major existingapplications should be included in a specialappropriation to DIS. Assuming thecontinuance of the Governor's Committeeon State Information Systems, thatcommittee should advise the DISDirector on the expenditure of thosefunds. It is also recommended that anamount in addition to that necessary tofund the known new applications beincluded in the special appropriation forapplications where the time table should beacceleratedl.

A possible level of funding for MIDS isidentified in Section 2 of this part.Additional funding recommended is

$3,000,000 for adequate space for DIS,$1,900,000 to assist local government withdevelopment efforts see Part V of thisreport, and an addition of $250,000 to thecomputer revolving fund for purchase,rather than lease, of certain equipment.

Performance evaluation is desirable for allfunctions, but it is mandatory for DISbecause of the growing impact it has onState government functions and decisions.An evaluation is already being performedto a limited extent by the State PublicExaminer's Office and should continue tobe the responsibility of that office.

The development, installation, and policing ofsubsystems for assuring the integrity ofcomputer-based and all other data and to assureconfidentiality of such data by limiting itsaccessibility only to properly authorized persons.

In this recommendation, attention is focused onthe need for assuring integrity of data files bypreventing willful tampering with data and byproviding detailed procedures for the correctionof erroneous data or elimination of obsolete data.Further, and most important, is action necessaryto safeguard private rights by establishing

procedures to guarantee confidentiality of data. Itis emphasized that all data are included, not justcomputer-based data.

Accordingly, it is recommended that theLegislature enact legislation establishing State andI ocal government policy on integrity andconfidentiality of all data and direct theDirector of Information Systems and theIntergovernmental Commission on InformationSystems (see Part V of this report) to give priorityto implementing the policies.

A leadership position in the development andcoordination of intergovernmental informationsystems and of intergovernmental cooperativeefforts in systems analysis and computerization.In every way possible, DIS should encourage andparticipate in intergovernmental efforts. Forexample:

e Encourage Federal agencies to fund thedevelopment and implementation of majorinformation systems for State and localgovernments.

e Corollary to the above, encourage directFederal and State participation along withlocal' government in the development and

i

implementation of local governmentinformation systems.

e To the extent feasible, promote MinnesotaState or Local government agencies servingas prototypes in systems development assuggested above.

Encourage joint local governmentdevelopment of information systems,assisting where possible with Statepersonnel and funding.

e Promote the utilization of existing,successful applications including individualprograms as well as broader groupings ofapplications.

Support with personnel and funding anyprograms of NASIS (The NationalAssociation for State Information Systems)to promote common systems,intergovernmental data interchanges, dataand operating standards, and any otherNASIS efforts furthering State and Localinformation systems.

Coordinate and assist with thedevelopment of data bases of statewide orregional value, such as MAPS at theUniversity of Minnesota and RAFTdeveloped by the Citizen's League and theUpper Midwest Research and DevelopmentCouncil.

'OA

yip

HIGHER EDUCATION

In the State of Minnesota there are five "systems"of higher education, each comprised of numerousseparate institutions or campuses distributedthroughout the state: the University ofMinnesota, the State College System, the StateJunior College System, the AreaVocational-Technical Schools, and the PrivateColleges. Each of these five systems has as itsmission one or more of the three traditionalobjectives of higher education in the UnitedStates: Instruction, including vocational,professional, and general education; Research,defined as the expansion of knowledge and thedevelopment of techniques, methods, and devicesfor solving problems related to the natural worldand to man and his social institutions; and PublicServices to government and society. The entirethrust of this study has been to determine thecomputing facility and service needs of Minnesotapost-secondary institutions in terms of theirparticular and diverse objectives, and to find andsuggest ways of satisfying these needs which havethe greatest likelihood of being implemented. Tothis end, educators and administrators fromnumerous institutions in all five systemsthroughout the state, as well as from other states,were consulted extensively before and duringpreparation of the report.

The objectives of each of the five major sectionsof this part of the report are as follows:

To identify the specific objectives of eachof the five systems and the importance ofcomputers in meeting these objectives,both directly as a part of the process ofinstruction, research, and public service,and indirectly through support of learningresources, student services, and generaladministrative operation of institutions(Section 1);

To summarize the existing (1970-71)computing facilities and applications inMinnesota higher education, and their costs(Section 2);

To analyze the needs of Minnesota highereducation for computing facilities andservices over the next five and ten years(Section 3);

To determine the costs of various ways ofmeeting these needs, identifying the leastexpensive ways (Section 4); and

To recommend specific actions by theinstitutions and systems of highereducation and the legislature to meet theidentified needs and lay a sound basis forfuture development (Section 5).

COMPUTERS AND THE OBJECTIVES OFHIGHER EDUCATION

Without going into the detail and citing some ofthe examples to be found in the complete report,there are few areas of higher education in whichcomputers or information processing systemshave not already been applied in some way.Computers are both an object of instruction (forstudents in computer science, data processing,and other engineering fields) and a tool ofinstruction in fields as diverse as the basicsciences, social sciences and art and music. One ofthe few ways in sight to control the rapidly risingcosts of higher education, and at the same timeperhaps improve its quality, may be theas-yet-unproven direct application of informationsystems to the process of instructionCompu ter-A ssisted Instruction andComputer-Managed Instruction. In research,computers are similarly an object of study anddevelopment in the areas of engineering and

... 2-13

information sciences, and an indispensible tool inmost research fields of science, engineering, socialscience, medicine, and many of the humanities.Use of computers in both the instructional andresearch missions of higher education hasoverlapped considerably into the public servicearea already in-service training for personnelfrom schools, public agencies, and industry, aswell as development of applications to problemsof planning and operations in public and privateorganizations. The potentials for further publicservice applications of this kind can besubstantially increased through suitablecommunication and coordination between highereducation and other segments of society. Finally,higher education is one of the largest "industries"in the State of Minnesota. Its administrativeprocesses are similar to those of any other"industry" group. The economic savings resultingfrom a coordinated use of computer andinformation systems in higher education areconsiderable, just as they are in commerce anuindustry.

In 1967, a Committee led by John R. Pierce ofthe Bell Telephone Laboratories issued a reportComputers In Higher Education,1 under theauspices of the President's Science AdvisoryCommittee. In addition to presenting a mostcompelling justification for the substantial use ofcomputer systems in higher education, the PierceReport established a norm for the amount ofcomputer capacity needed namely, that allinstitutions of higher education should be at leastas well equipped in facilities and programs as werethe best-equipped colleges and universities in1967. Because of the wide disparity between thebest and the average in 1967, and because of thegenerally slow pace of funding since then, thesegoals are still reasonable ones for Minnesotahigher education in the coming decade. ThePierce Reportl is referred to extensively in thisstudy of Minnesota's needs.

The University of Minnesota is the State's largestsystem of higher education as well as its primary

resource for graduate and professional training,research, and public service. It is also the leadingstate institution in computer facilities and relatedresearch and training programs. The University isequipped with a major administrative dataprocessing system which is used to capacity, plusa large system for instruction and research whichincludes a major computer (the CDC 6600) andseveral small satellite computers or input/outputstations. With the exception of a medium-sizedcomputer serving the Duluth campus, othercomputing facilities at the University arespecial-purpose devices built into laboratoryapparatus or otherwise dedicated to specialapplications, and funded almost exclusively byfederal or private grants.

With its recent acquisition of a Univac 1106computer, Mankato State College is surely one ofthe best-equipped non-PhD-granting colleges inthe nation. However, this computer is intended to .

serve the needs of the other state colleges andoutside users as well. Until the magnitude ofoutside use is established, the adequacy of the1106 will be undetermined. The remaining fivestate colleges each have small, obsolete computerswith capacities considerably below their presentneeds for instruction and administrativeprocessing.

The State Junior College System has a small,second-generation computer which providesadequate administrative service for the eighteenjunior colleges at present, but with littlepossibility of expanding the kinds or amount ofservice. Beginning this year, each of the eighteencolleges is also connected to a commercialtime-shared computer service in Minneapolis forinstructional computing the first access theyhave had for this purpose. These tie-ins weremade possible by a $60,000 grant from a specialappropriation by the 1969 Legislature to theHigher Education Coordinating Commission.

Data processing curricula are alive and prosperingin the three Area Vocational-Technical Schoolswhich have small computers for this purpose.They exist in less prosperous condition at severalother schools which make use of small school

A..,

1.4.42:2

II.

board computers, or facilities made available onundependable terms by nearby business concerns.

Finally, computing in the private colleges rangesfrom generously adequate by present standards tozero. Partly by necessity, the private collegesseem to be more involved in cooperative activitiesamong themselves and with public institutions;the Twin Cities College Computer Committeeinvolving seven Twin Cities area colleges and theUniversity of Minnesota is the largest such group.

The table below summarizes order-of-magnitudeequivalent purchase costs of existing computerfacilities in Minnesota higher education, extractedfrom Tables 2.1 and 5.1, and from Appendix H.5of the complete report. The numbers have beenrounded to the nearest $100,000 and representequivalent purchase cost.

University of Minnesota

State College System

State Junior College System

Area Vocational-Technical Schools

Private Four-Year Colleges (Est.)

I $ 6,400,000

2,200,000

200,000

800,000

850,000

TOTAL $10,450,000

Computer industry marketing data is available ona confidential basis to compare Minnesota withother states. Although this information is not anaccurate source of total expenditures forequipment, it does provide a reasonably uniformbasis for comparison. Among the seven Big-10midwestern states, Minnesota ranks fifth in publicspending for computer facilities in highereducation, and fourth in spending among privatecolleges (Figures 2.1 and 2.2). Compared with thestandards established by the Pierce Reportlwhich recommended a $414 million nationalexpenditure for computing in higher education,Minnesota ought to be spending about 2% of thisamount, or $8 million. Actual equivalent annuallease plus staff, maintenance and operatingsupport is estimated at about $4.2 million for1970-71 (Table 5.1 plus a generous contributionof $500,000 for all of the private colleges).

GOALS AND COMPUTER RESOURCESNEEDED TO MEET THEM

The goals toward which the analysis of this andthe next section are directed are two:

Achievement of the Pierce Reportlobjectives for educational computing in theState of Minnesota, i.e., to provideeducational computing services for all ofhigher education in Minnesota equivalentto that available at the leading universitiesof the nation in 1967-68.

Establishment of machine-readable databases and an administrative data processingcapacity which is sufficient to support themanagement information needs ofinstitutions and systems of highereducation and lay the basis for programplanning and budgeting.

In this section the computer capacity necessary tomeet these goals will be estimated, in terms of therequired number of input/output terminals ofvarious kinds, computing power required instatements compiled per second, and the amountof mass data storage needed. The followingsection will convert these capacity estimates intocost estimates.

The capacity estimates for educational computingare based on the Pierce Report breakdown ofacademic areas into those requiring substantialuse of computers, limited use, and casual (verylittle) use. Following a procedure used in a

feasibility study by the General LearningCorporation2, it was estimated that, averagedover all kinds of students in Minnesota collegesand universities, the computers would need tohandle 11.3 problems per year for each full-timeequivalent student. Each problem would containabout 70 compute statements to compile. Ifcarried out on a time-shared computer system,between 1.1 and 1.8 hours of terminal timewould be required for each problem, dependingon the skill of the student. On a batch processingsystem using medium-speed remote terminals,between three and six runs would be required tocomplete each problem, depending on the

2 -15

comp lexity of programming language used.Although such numbers may not be typical ofany particular computer used by students, theyhave been accepted as reasonable estimates ofoverall use by most of the educators andcomputer specialists consulted during the courseof the study. Further details of the analysis, alongwith modifications included for junior collegesand area vocational-technical schools, may befound in the complete report and in AppendixH.6. Results are summarized in Table 3.1.

It proved to be more difficult to estimate thecomputer capacity required to support theadministrative data processing needs of Minnesotahigher education, because no quantitativenational norms or feasibility studies were foundto guide the analysis. Therefore, the University ofMinnesota, which has the most completeadministrative system in the state, was used as amodel. Most of the possible administrativeapplications are in either the operational orplanning stages at the University, and the isolatedexceptions (a computer-based student registrationsystem, is the major omission) were added on thebasis of a crude but reasonable design study. Theadministrative computing resources required by

the other collegiate institutions were estimatedthen by scaling according to enrollment, staff sizeor other appropriate application-dependentquantities. (Area vocational-technical schoolswere omitted from this analysis, since theiradministrative processing is handled by the schooldistricts to which they are attached units oflocal government.) Details of this analysis areexplained in Appendix H.6 and tabulated in TableH.6.8, while the results appear in Table 3.2.

The small table below provides a brief summaryof the estimated computer capacities required in1975-76 for all of the Minnesota collegiateinstitutions University, State Colleges, StateJunior Colleges, and Private Colleges.

Research and public service computing needs havenot been included in this analysis. It should bekept in mind, however, that access to sufficientcomputer capacity is needed by those institutionswith missions in these areas, but that the cost ofproviding such services is normally covered by theoutside sources which fund or request them.There were three special computer applications inhigher education which merit specific discussion:

TOTAL ENROLLMENT

EDUCATIONAL COMPUTING

Batch Processing

No. Medium-Speed Terminals

No. Statements Compiled Per Second

Time-Shared Computing

No. Terminals

No. Statements Compiled Per Second

ADMINISTRATIVE DATA PROCESSING

Mass Storage (millions of characters)

Required Peak Output Rate (lines per minute)

Required Peak Computer Processing Rate (transactions per minute)

A Statewide Automated Library System,which has the potential of substantiallyincreasing the quality of library servicesavailable throughout the state, and at thesame time controlling an otherwiseinevitable con escalation and servicedeterioration associated with the explosionin volume of printed information.

Computer-Assisted and Computer-ManagedInstruction, which hold promises ofsubstantial increases in the effectivenessand efficiency (cost savings) of highereducation if the techniques and technologyare carefully and properly developed andapplied.

Information Services, building partly uponseveral research projects at the Universityand providing ready access and analysis ofplanning information required by manypublic agencies and private organizationsthroughout the State.

The first two special areas listed are sufficientlyimportant to higher education that experts wereconsulted during the course of this study, andasked to prepare position papers which appear asAppendices H.2 and H.3. The third area is ofsufficient interest to other public agencies that itis discussed in the State and Local Governmentparts of this report.

FACILITIES AND COSTS TO MEET THEGOALS

Having estimated the amount of raw computerpower, input/output capacity, and mass datastorage needed by Minnesota higher education,the next task is to convert these estimates intoreasonable configurations of computer equipmentto which costs can be assigned. The largestcomponents of the total cost are the computersthemselves. Using benchmark test data from theAuerbach Reports, plus the personal experienceof several staff members of Analysts InternationalCorporation, a scale was constructed relatingcomputer cost to computing capacity, asmeasured by the number of FORTRAN- languagestatements compiled per minute or data

processing transactions executed per minute. (Thescale is given in Figure 4.1 and in Table H.7.1 ofAppendix 7.) Then this scale was used to convertthe previous computer need estimates forMinnesota higher education into cost estimatesfor three kinds of computer systems:

Systems supporting only batch processingvia local and remote terminals (except forthe junior colleges, for which a small,central time-shared computer is less

expensive).

Systems supporting time-shared and batchcomputing on the same general-purposecomputers.

Separate computer systems iD providebatch processing and time-sharedcomputing, using for the latter smallcomputer configurations especiallydesigned to process the short studentprograms and interactive applicationscharacteristic of much instructionalcomputing.

One observation resulting from this analysis isthat instructional time-shared computing can becarried out most economically on a smallcomputer system dedicated to this use. The costis not much greater than that of the equivalentinstructional computing carried out on a largebatch-processing computer.

In a similar fashion, a scale was constructedrelating the mass data storage capacity of variousmagnetic disk and drum units to their cost.

A more difficult task was the configuration ofremote terminal systems to connect variousinstitutions to central or regional computersystems. The components required for suchsystems include the computer terminalsthemselves (input/output devices); modems ordata sets to interface these to communicationlines; communications controllers to efficientlymatch the computer to communication lines;multiplexors to concentrate data from or toseveral terminals over one communication line;and a variety of communication line servicesavailable for data transmission. Costs were

2 -17

estimated for a spectrum of each of thesecomponents, and then the costs associated withseveral different geographical deployments ofcomputer and terminal equipment were analyzedfor each of the five systems of higher education.Included with equipment costs in these analyseswer 3 operating support costs staff, salaries,supplies and auxiliary equipment, andmai itenance without which no computersystem can function. Adequate operating supportwas estimated from published data and pastexperience to equal the equivalent lease cost ofequipment for educational computing, and to betwice the equivalent lease cost for administrativedata processing.

Complete details of this analysis may be found inSec: on 4 of the report, supplemented byAppendices H.7, H.8, and H.9. Those readerswho are not familiar with remote terminalcom niter configurations and the functions of theman y components they contain may find ithelpful to refer to Figure 4.2 and itsaccompanying text. In Table 4.5 are summarizedthe resulting monthly costs of the variousconfigurations analyzed. The brief table belowabstracts the total costs of the least expensivealternatives for the four public systems of higherecluc3tion. (It was not feasible to include theprivate colleges in a similar analysis; however, thecomputer capacity implied here for the publicsystems is conveniently distributed for access byprivate colleges.) The costs quoted includeequivalent monthly lease of equipment plusmonthly operating support.

University of Minnesota System (U-2)

State College System with TwoCentral Computers (S-2)

State Junior College System:Central Administrative Computerphis Time-Shared InstructionalComputer (JA-1 plus JI-1)

Area Vocational- Technical Schoolsus ng Regional Centers(instruction only) (V-2)

TOTAL

Fuliy Regionalized Computer Systems(R1), plus Central AdministrativeCo nputer for Junior Colleges

$321,000/month

172,000

69,000

65,000

$627, 000/month

$636, 000/month

2-18

One of the interesting conclusions from thisanalysis is that regionalization of computerfacilities within higher education does not result,at present, in significant cost savings. The savingsin communication line costs are small because ofthe way institutions are distributedgeographically. With a few notable exceptions,these savings are partly offset by increased costsof services from regional computers, whichprovide many kinds of services rather thanspecializing in one application for which thesystem can be optimized (e.g., instructionaltime-shared computing).

RECOMMENDATIONS AND DEVELOPMENTSCHEDULES

This final section presents:

a series of recommendations for actions orpolicies in general, for each of the fivesystems of higher education, and for theHigher Education CoordinatingCommission;

a rather specific five-year schedule fordeveloping computing capability inMinnesota higher education, complete withorder-of-magnitude cost estimates;

a more general indication of developmentsduring the second five -year period,1975-1980;

some comparative information on othercost figures in higher education to placethe cost estimates in perspective; and

suggestions for priorities and planning touse the funds which may become available.

In a certain sense, the section could be considereda "Master Plan" for computers in Minnesotahigher education. The recommendations andfunding estimates, however, must be thought ofas guidelines; they must not be used as rigid,specific plans. The scope and time available tocomplete the study have not been sufficient todevelop specific, detailed implementation plans,which in any case are the responsibility of theinstitutions and systems themselves.

As presented briefly in this summary, therecommendations are organized somewhatdifferently than in the complete report, toemphasize their statewide implications ratherthan their relation to the institutions with whichthey may be most directly associated.

Goals and General Recommendations

Minnesota higher education should attemptto achieve by 1975 a computer capacityfor education and administrationequivalent to that enjoyed by thebest-equipped colleges and universities in1967. That is, the Pierce Report goalsshould be achieved throughout Minnesotahigher education by 1975, and anadministrative computer capacity shouldbe established capable of supportingcollege management information systemsand program planning and budgeting.

It is neither necessary nor desirable toestablish full-scale educational programs inComputer Science or Data Processing at allpublic institutions of higher education.

Computer installations in higher educationmust be operated on a service-orientedbasis, with instruction and trainingapplications normally scheduled duringhours convenient to students. A serviceorientation, including a meaningful voicefor all major users in setting priorities andpolicy, is indispensible for the successfulfunctioning of regional or statewidecomputer services.

The capacity for research computingshould be available to those institutionswith missions including research. Such

computing is supported appropriately inthe same way as other research activities,and should not require special funding.

Institutions which train teachers andeducational administrators, and the StateDepartment of Education which certifiesschool personnel, should be requested toestablish degree and certificationrequirements which include appropriatetraining in computer and informationtechnology and data processing. Futureschool administrators and teachers in

certain fields (e.g., science, mathematics,business education) cannot be consideredas adequately prepared for their jobwithout some competence in managing andusing information systems and computers.

Recommendations for Regional and StatewideFacilities and Programs

Existing regional computer centers withinhigher education should be supported at alevel which will permit them to provideadequate regional services. In particular,funds should be provided to expand thecommunications capability of theUniversity's CDC 6600 computer so that itcan service more outside users from remoteterminals; and adequate operating supportshould be provided to the new RegionalComputer Center at Mankato StateCollege, so that this experiment inproviding computer service to a widevariety of institutions within and outsideof higher education has a realistic chanceto succeed.

A Regional Computer Center at St. CloudState College should begin with amedium-sized satellite computer to theMankato Regional Center, for installationin 1973. If the Regional Center concept issuccessful, this satellite would be expandedin 1975 to a full Regional Center.However, this and other expansion in thesize or numbers of Regional ComputerCenters should proceed only after a carefulevaluation of the experience from existingCenters.

To meet the instructional computing needsof the State Junior Colleges and manyprivate colleges, and the needs of theUniversity of Minnesota for time-sharedand other interactive computing, a

statewide time-shared computing facilityshould be established. Initially, this facilitywould be under the management of one ofthe systems of higher educationprobably the University of Minnesotabut with operating policies established byall users. As the system grows and multiplecomputers are added, it may be desirableto decentralize the locations and managinginstitutions, still retaining some statewide

2-19

coordination of policies and systemsdevelopment (hardware and software).

Planning should commence immediatelyfor a Statewide Automated LibrarySystem, based on preliminary work doneby the University of Minnesota LibrarySystems Staff, other libraries, and librarygroups in the State, and evaluation byBecker and Hayes, Inc. (Appendix H.2).The mechanism to begin this effort shouldbe a Planning Board established by theHigher Education CoordinatingCommission, with the mission ofdeveloping for the 1971-73 bienniumeither a complete plan, or interim measuresto continue present activities which willcontribute eventually to a StatewideAutomated Library System.

To support investigation of the potentialeffectiveness and economies ofComputer-Assisted and Computer-ManagedInstruction, a statewide facility should beestablished to provide computer servicesand other technical support todevelopment activities at the University'sCenter for Research in Human Learningand at similar centers throughout the Stateof Minnesota. This statewide supportfacility would also coordinate developmentprograms in Computer-AssistedInstruction, and aid in their transition fromdevelopment to operational status.

Research and development activitiesassociated with information services

depending on accurate,continuously-updated, computerized databases should be supported. It isrecommended that the Department ofAdministration, the University ofMinnesota, and other state agencies jointlydevelop the projects, funding requests, andcontracts necessary to implement suchdevelopment programs which are ofpotential value to the people of Minnesota.

Recommendations Related to Individual Systems

University of Minnesota

Special laboratory computers are requiredfor certain professional programs at the

2-20

University, especially in computer scienceand engineering. If these programs are tomaintain their quality, such laboratorycomputers must be funded.

The Coordinate Campuses of theUniversity can most effectively be suppliedwith computer capacity as follows.

Duluth a medium-sized computer,somewhat expanded from the presentconfiguration. If the Regional ComputerCenter concept prevails, then Duluthwould expand into a Regional Center fornortheastern Minnesota during the last halfof this decade.

Morris remote terminals to theUniversity's CDC 6600 and to thestatewide time-shared computer facility.

Crookston and Waseca terminals to thestatewide time-shared computer facility.

State College System

As soon as possible, each of the statecolleges (except Mankato) should beprovided with a remote terminal capable ofinterfacing with both the MankatoRegional Computer Center and theUniversity's CDC 6600. This will enablethe colleges to begin providing neededservice to their students andadministration, and to acquire experiencein the remote use of large computers.

State Junior College System

In addition to instructional computing viathe statewide time-shared computerfacility, the Junior Colleges require a

somewhat larger administrative dataprocessing capacity than they now possess.It is recommended that they acquire asmall third-generation computer withadequate magnetic disk storage and an

efficient COBOL compiler. Although mailand parcel service has been and willcontinue to be adequate for most of theirdata communications, the Junior CollegeData Processing Staff should, within a fewyears, begin to develop a remote querysystem for more rapid access by theindividual colleges to certain kinds ofinformation.

f

.1-

.

i

Area Vocational-Technical Schools

Over the next five years, AreaVocational-Technical Schools withprograms in Data Processing andAccounting should gather experience inusing remote terminals to large computers,and in using business minicomputers whichcan operate either by themselves for smallprocessing jobs, or as remote terminals to alarge computer. It is particularly importantfor one or two of these terminals to beginextensive use of the Mankato RegionalCenter as soon as possible, so thatexperience in working with a center of thiskind can be acquired. After a few years ofexperience with this variety of terminalsand computers, it will be possible to makean informed decision on the most effectiveand economical future direction ofcomputing facilities for vocational training.

Private Colleges

Technical assistance and advice should beprovided by the staffs of the largestate-supported computer installations tohelp the private colleges developcommunications interfaces between theirterminal and computer facilities and thoseof the public institutions. In addition,computer systems in the public institutions

most especially the regional andstatewide facilities should provideservices to private colleges at cost.

Within constitutional constraints, the Stateshould aid the development of computingcapacity in the private colleges bysupporting regional and statewidecomputer facilities and datacommunications networks, and byproviding grants to the private colleges tosubsidize their use of available computerservices. It is recommended that thesesubsidy grants build up to coverapproximately half the cost of thecomputer services (exclusive of terminaland communications costs) required tomeet the educational needs of the privatecolleges.

Recommendations Addressed to the HigherEducation Coordinating Commission

To increase the efficiency and effectivenesswith which its needs for informationprocessing can be met, the HigherEducation Coordinating Commissionshould establish a full-time staff positionfor an analyst to analyze and advise onprogram and budget requirements forinformation and data processing, and to actas a consultant and service manager forapplications programming. In addition,remote access to computing facilities anddata bases should be made available toCommission staff through the State CollegeBoard terminal in the Capitol Square

Building.

It is recommended that the HigherEducation Coordinating Commissionestablish a mechanism to develop standardsfor the coordination, compatibility, andsecurity of system data bases andapplications programs.

The Higher Education CoordinatingCommission should establish a technicalcommittee to develop the specifications,implementation standards, andcoordination mechanisms to create andmaintain compatible communicationsinterfaces between all terminal and

computer facilities in higher education.This effort should be coordinated throughthe Telecommunications Section of theState Department of Administration withsimilar standards for other agencies of stateand local government.

The Minnesota Higher EducationCoordinating Commission should explorewith its counterpart in North Dakota thepossibility of sharing computer resources.A reasonable first step in this directionwould be to invite a North Dakotarepresentative to participate in the work ofthe previously-mentioned TechnicalCommittee on Communications Standards.

An ad hoc committee should be establishedto open channels of communication and

2-21

coordination among systems andinstitutions offering in-service andpre-service training in computing and dataprocessing, and to promote an increase inthe amount and geographical distributionof such training.

The special appropriation of $300,000 bythe 1969 Legislature has been mosteffective in stimulating the development ofmany useful computer applications andservices in Minnesota higher education. It isrecommended that sur,h appropriations becontinued at a similar or slightly expandedlevel, that they be made to and disbursedby the Higher Education CoordinatingCommission, and that the Commissionprepare a report to each session of theLegislature summarizing and evaluatingwhat has been accomplished with thedevelopment funds.

The interrelationships between institutions andsystems implied by the precedingrecommendations will require a large degree ofcoordination in developing the facilities andorganizations required to achieve an effective andeconomical computing and informationprocessing capacity in Minnesota highereducation. In approving this report, the HigherEducation Coordinating Commission has adopteda specific procedural policy to insure thecoordination of plans and budgets for computingfacilities and services in the various systems andinstitutions of higher education. A statement ofthis policy appears following theRecommendations in the complete report.

Development Schedules

Converting the recommendations above intospecific facilities and costs for the five years1970-75 yields the contents of Tables 5.1 through5.5 in the complete report. The total annualcosts estimated for facilities and operatingsupport over this period are summarized in thetable below, where the 1970-71 expenditures arethose actually planned and committed for thatyear,' and the facilities costs are in terms ofequivalent annual lease, regardless of whether the

2-22

Fiscal or

AcademicYear Facilities

ONmairigSupport Total

1970-71 $2,360,000 $1,864,000 $4,224,000

1971-72 2,949,000 2,864,000 5,813,000

1972-73 3,463,000 3,164,000 6,627,000

1973-74 3997,000 3,754,000 7,751,000

1974-75 4,672,000 4,338,000 9,010,000

facilities are actually being leased or have beenpurchased.

The specific facilities recommended for thisfive-year period are illustrated schematically inFigure 5.1 of the complete report, along with ageneral indication of the expansion which will berequired to maintain an adequate computingcapacity during the next five-year period1975-80. The small graph below is a reducedversion of Figure 5.2, presenting the costssummarized in the above table, bracketed on theleft by the estimated expenditures for computingin Minnesota public higher education from 1964to 1970, and on the right by the estimated costsfor 1975 to 1980 scaled according to theprojected enrollment increase from 1975.

0

2 9..

10

8.3.1;.

W

2g z 6

5

4

5 3 -LL

2 . .

5:ex 1-54

0 ESIIMATED

RE( DMMENDED

A FACILITIES EXPENDITURESFACILITIES ANDOPERATING SUPPORT

FACILITIES

FIRST FIVE YEARS OF PLAN

"RECOMMENDED EXPEND-" ITURES FOR SECOND FIVE

YEAR PERIOD ARE SCALEDTO PRDJECTED ENROLL-MENTS

SECOND FIVE YEARS OF PLAN

1964 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 1980 81

Although the total expenditures recommendedare certainly formidable, they appear morereasonable on a per-student basis. If the goals ofadequate computing are achieved in 1975, thenthe cost per student per year, according to theestimates here, would be about $56, comparedwith $60 estimated by the Pierce Report in 1967.The Pierce Report also points out that the cost oflibrary services in colleges and universities wasbetween $50 and $200 per student per year in1966, and the cost of chemistry laboratoryinstruction was about $95 per chemistry studentper year. The $8.8 million for computingrecommended for 1974-75 is about 2.5% of thetotal operating costs estimated for public highereducation in Minnesota for that year, comparedwith 4% of operating costs recommendednationally for computing in higher education bythe Pierce Report.

Despite these indications that the expendituresrecommended here are reasonable, it is clear that,when balanced against other demands on the tax

dollar, it may not be possible to achieve thedesired computing capacity in higher education asearly as 1975. If support cannot be provided atthe level recommended, then it is the opinion ofthe authors of this report that priorities forspending available funds should include supportfor the existing facilities; establishment of alimited-capacity statewide time-shared computerfacility to serve existing needs in the JuniorColleges, the private colleges, and the University;and establ ishment of a center to supportdevelopment work on Computer-Assisted andComputer-Managed Instruction throughout thestate. However, it is properly and ultimately theresponsibility of the individual institutions andsystems to higher education to plan the facilitiesa2d programs in detail and to set the prioritiesjointly. The policy for planning and budgetcoordination adopted by the Minnesota HigherEducation Coordinating Commission as a result ofthis study establishes a sound mechanism toaccomplish these tasks.

LOCAL GOVERNMENT

In large measure, local government is where the"action" is. It is here that many of the domesticissues are centered, and it is here that much of thegovernmental service to the public is rendered. Itis also at the local government level that thescience and art of systems analysis andcomputerization can be of most assistance topublic administrators in effecting both betterservice and economies.

Local governments have generally lagged behindthe Federal and State governments in suchefforts. But there is a growing interest andconcern among local governments; the question ishow best to approach the overall problem. Withfew exceptions, "go it alone", uncoordinatedefforts by local governments will not get the jobdone. The extent of the job in Minnesota is 87counties, 139 municipalities of 2,500 populationor greater, and 440 elementary-secondary schooldistricts.

In Minnesota, the problem has been recognizedby the Legislature, the Governor, the Governor'sCommittee on State Information Systems, theState Commissioners of Administration andEducation, and others. The Legislature consideredlegislation relating to local governmentcomputerization at the 1969 session and hascontinued its interest since the session. TheGovernor in Executive Orders No. 15 and 56 hasassigned responsibility to the Computer ServicesDivision in developing local governmentinformation systems including staff support forsystems design and programming and pilot projectcomputerization. The Commissioner ofAdministration, in a recent reorganization of theComputer Services Division, established a LocalGovernment Unit in the Planning andCoordination Section of that division. TheGovernor's Committee on State InformationSystems was involved in these actions by theGovernor and the Commissioner of

Administration, and its chairman believes thelocal government area offers the greatest potentialpayoff from information systemsimplementation. The Commissioner of Educationand the State Board have made clear their interestin efficient development and use of computers.

Local government interest has been graduallygrowing. A State-Local Data Systems Group hasbeen set up to begin work on coordinated effortsin information systems development.

The business of local government is particularlysusceptible to assistance from informationsystems and computerization. The evidence ofservice such as a birth certificate, building permit,or welfare payment may be produced by acomputer. The computer can serve as therepository of the raw material to produce suchproducts; and based on the same data, thecomputer can provide management informationto control program execution and evaluateprogram performance.

With some exceptions, however, localgovernments have not been in a position to takeadvantage of the new technology and arestruggling with masses of data in manual fileswhich cannot be focused on current problems in atimely fashion. Further, most efforts to improvethe situation have emphasized how to dosomething better rather than looking to theobjectives and determing what is to beaccomplished.

What is needed? Information systems are requiredto serve operational needs and at the same timegenerate the data necessary for research andplanning and permit interrelationship amongsystems for both operational and planningpurposes. In local government, four data bases, orfiles, are needed: natural persons, legal persons,real property, and personal property. The data

a - vir . 2-25

elements in the individual agency files shouldserve the day-to-day operating needs and planningand research requirements and should be relatablethrough a master index file, under extremelycarefully prescribed controls, to data in otherfiles. To the extent possible, routine decisionmaking should be assigned to the system.

Development and implementation of suchsystems are sizable tasks and require extensivecoordination to assure that intergovernmentalinformation needs are met and to avoid the costof redundancy. In April 1988, theIntergovernmental Task Force on InformationSystems emphasized the need (or suchcoordination in its imaginative and authoritativereport entitled "The Dynamics of InformationFlow." The Task Force, composed of personsrepresenting Federal, state, and localgovernments, addressed the needs forintergovernmental data systems and exchange andthe means for meeting those needs.

Although there is

information on thegovernment, mostapplications injurisdictions. Inredundancy, present

a substantial amount ofuse of computers in localof it relates to specificindividual governmentfact, a measure of theand potential, of systems

design and development cost might be obtainedfrom a review of the literature. There is someevidence in the present trend, however, to makeone optimistic about significant improvements inlocal government information systems andcomputerization.

Among the hopeful signs is the USAC (UrbanSystems Advisory Committee) projectadministered by the Department of Housing andUrban Development. This project has as itsobjective the building of municipal informationsystems that will be of value to jurisdictions otherthan those in which they are being developed. St.Paul, Minnesota, is one of six participating citiesand is developing a human resources informationsystem. Another example is the State-financedMINCIS (Minnesota Crime Information System)and its federally aided adjunct project SEARCH

2-26

which serve local law enforcement agenciesthroughout Minnesota.

Several states are responding to the localgovernment information systems problem in wayssimilar to those recommended herein forMinnesota. For example, both California andPennsylvania have set up intergovernmentalboards to establish guidelines forintergovernmental information systems andcomputer development. There are also a numberof examples of local government cooperativedevelopment of computer facilities or datainterchanges.

STATUS IN MINNESOTA

Questionnaires were used in an attempt to obtaina picture of county and municipal status inMinnesota. With few exceptions, county andmunicipal governments are in beginning stages ofmechanization or have done nothing. Littlecooperative development has occurred. Countyand municipal governments in Minnesota aregenerally at the threshold and at a point wherecoordination and cooperation can be most easilyrealized.

The situation in school districts differs in somedegree from that in counties and municipalities.This appears to be the result of three factors:State Board and Department of Educationsupport of efforts to tackle the problem; theTIES (Total Information for Education System)development; and the interest and action in manyschool districts on instructional uses ofcomputers. State department financed reportshave addressed the problem of coordinatingcomputer facilities for school districts, concludingthat centers based on state regions would beappropriate. In Juiy 1970, the State Board ofEducation adopted a policy statement endorsingthe regional concept and a close tie to highereducation.

TI ES was established in 1967 under theMinnesota Joint Powers Act to provideinformation systems development services andcomputer facilities for a group of school districts.

1

Now totaling 29 school districts with anenrollment of nearly 240,000 students, TIES isproviding both administrative and instructionalcomputer facilities. A recent TIES evaluationreport summarizes TIES' problems andaccomplishments, both of which have beensubstantial, and concludes that TIES is in theprocess of "restoring the confidence of its users"and that by "strict adherence to systemdevelopment discipline, TIES can continue thisupward trend and move to new levels of service."It appears that both in scope of service and size,TIES stands at or near the head of the classnationwide in coordinated local governmentcomputer efforts.

Local governments striking out on their own inthe information systems computerization mazeare destined, with few exceptions, to follow acourse that at best leads to marginal results and atworst to disaster. The process begins with apartial feasibility study on mechanizing a fewapplications which often results in obtaining asmall computer, often not suitable for the fewapplications and totally unsuitable for theinformation systems task that needs doing. Itcontinues with a staffing search and the firstround of implementation which is undertaken inan unrealistic time frame, resulting in badlymissed schedules, heavily overrun budgets, anddisenchanted users. Ultimately there is somethingto see, but it has no resemblance to aninformation system. At this point it becomesclear that the facility is not adequate for the job,leading at some later date to an upgrading to alarger computer resulting in costlyre-programming. The data base needed for theinformation system, however, is still non-existent.Cost/benefit evaluation then proves catastrophic.

In order for effective coordination to exist,however, there must be agreement on certainfundamental principles among the users; orsuccess is imperiled. These include: aninformation system must be implemented foreach service program; highly qualified systemsanalysts and medium-to-large scale centralcomputer are required; smaller governmentaljurisdictions can benefit proportionately to the

large; the computer's availability and capacity areimportant, not its ownership or location; thecomputer facility is in the nature of a utility; usermanagement must get deeply involved with thenecessary comprehensive systems design; commonsystems development to cover the needs of similargovernmental jurisdictions is feasible;programming is a professional matter notrequiring the attention of individual jurisdictions;computer management is best left toprofessionals; priority should be given to systemswith largest payoffs; conversion planning is

essential; intergovernmental informationrequirements demand consideration and priority.

Direct and indirect cost/benefits of localgovernment information systems andcomputerization in Minnesota are not clear andwould be in retrospect virtually impossible todetermine. It can be said with certainty, however,that a coordinated approach will cost far less thana series of "go-it-alone" approaches. It can be saidwith virtual certainty that cost trends (costsgenerally higher but computer hardware costslower) will make computerization increasinglymore attractive economically in the future. It canbe said with high probability that the payoff frominformation systems-based decisions will farexceed other cost/benefits.

A series of recommendations are made to providea framework within which Minnesota localgovernment can move in an orderly, cost effectiveway, without delay, toward the development andimplementation of information systems and othercomputerization needs. The organizationalrecommendations are believed to be mostimportant recommendations in this report.

LOCAL GOVERNMENT RECOMMENDATIONS

Information systems and computerization placesubstantial demands on legislative and executivegroups in local government. The decision to moveinto these areas is a major one which, takenseriously, requires such officials to defineobjectives, and clearly delineate programs andpolicies in all functions of government.

2-27

Recommendations and guidelines to serve as aplan for accelerating the efficient development ofMinnesota local government information systemsand computerization are under these headings:

General recommendation for State policy.

Organizational requirements.

Approaches to coperative centers.

Area Computer Centers.

Orientation and training.

Funding.

Implementation suggestions.

General Recommendation

In summary, it is the strongest recommendationin this report that in 1971 the Legislatureestablish an Intergovernmental Commission onInformation Systems with coordinative authorityover all cooperative efforts in local governmentinformation systems and computerization. Theauthorities and responsibilities of the Commissionare limited to those necessary to attaining acoordinated, statewide approach to the buildingof local government information systems and toestablish the framework for the orderly andefficient development of cooperative computerfacilities. Implementation of thisrecommendation would call for amending theMinnesota Joint Powers statute (a) to require as afinal step, insofar as information systems andcomputerization are concerned, approval by /theIntergovernmental Commission and (b) for anysuch joint efforts involving school districts theapproval of the State Board of Education prior toCommission action. In addition, developmentalfunding by the State on a matching basis withlocal government is recommended as is

Comm ission control over Federal, State, orfoundation funding for all local governmentinformation systems and computerizationdevelopmental efforts. It should be noted that therecommendations are limited to cooperative andcoordinated activities; "go-it-alone" efforts bylocal government are not recommended to be inthe purview of the Commission.

2-28

Organizational Requirements

A formalized organization structure must beestablished in order to provide the framework andthe machinery for orderly development of localgovernment information systems andcomputerization.

An Intergovernmental Commission onInformation Systems should be establishedwith coordinative authority overcooperative efforts in information systemsand computerization in local government.Only in this way will it be possible to lookat the entire problem in a manner that canmaximize the cost effectiveness of suchefforts. State Board of Education approvalshould be required for any joint effort ofthis nature on the part of school districts,prior to submission of the plan to theIntergovernmental Commission. It is

believed that through these two steps thestatutory responsibilities of the StateBoard may be met and the necessarycoordination be obtained through theCommission. The authorities andresponsibilities recommended for theCommission must include only thosenecessary to permit statewide coordinatedefforts toward compatible informationsystems and to assure carefully planneddevelopment of computer facilitiesthroughout the state. In addition, therecommended representation on theCommission is designed so as to assure aneffective voice for all groups. Therecommended responsibilities andauthorities of the Commission are:

To establish goals and objectives forintergovernmental informationsystems in Minnesota as well as generalpolicies governing coordination,cooperation, joint efforts, andpriorities.

2. To authorize establishment of AreaComputer Centers; to promulgatepolicies and procedures assuring thecoordination of the efforts of AreaComputer Center Boads, so that anintegrated statewide system may

ultimately be developed; to establish(a) criteria for determining economicfeasibility for the Centers and theseveral information systems, and (b)procedures for measuring actual versusexpected cost/benefits.

3. To coordinate the development andimplementation of standards ofcompatibility of procedures,programming languages, codes, andmedia to facilitate the exchange ofinformation within and amongsystems.

4. To review and approve or disapproveall applications requesting Federal,State, or foundation funding forplanning and development of programswith potential impact on local orintergovernmental informationsystems.

5. To review and comment upon policiesaffecting intergovernmentalinformation exchange established bythe Governor's Committee on StateInformation Systems and/or theDirector of Information Systems andprovide advice to the State onlong-range policies, programs, andplans relative to intergovernmentalinformation systems.

6. To develop policies and designprocedures for the security of data inlocal government information systemsand for safeguarding theconfidentiality of such data consistentwith state statutes, local ordinances,and pertinent administrative rules andregulations (also see recommendationin Part III on security and privacy).

It is recommended that the MinnesotaJoint Powers statute be revised to requireapproval of the IntergovernmentalCommission on Information Systems priorto establishing any future cooperativeeffort in information systems andcomputerization. The reason for thisrecommendation is that the coordinatedand efficient development of a statewideinformation system has to be accomplishedwithin some general framework. Only the

Intergovernmental Commission will havethis viewpoint. If this is not done, it will bevirtually impossible (a) to avoid theoverwhelming cost of redundancy and (b)to develop standardized and compatibleinformation systems.

1 In order to attain a true intergovernmentalbase, members of the Commission shouldrepresent all governmental echelonsinvolved with intergovernmentalinformation requirements. Majorityrepresentation, however, should be given tolocal governments as a group. Accordingly,it is recommended that the Commissionmembership should consist of two personsrepresenting counties, two representingmunicipalities, three representing schooldistricts, three representing the StateGovernment, and one each from theGovernor's Committee on StateInformation Systems and Minnesotaregional government. The State membersshould be the Director of InformationSystems, a representative of the StateBoard of Education, and a representativeof higher education. A Federal governmentobserver should be obtained from adepartment such as HEW or HUD. It is

further recommended that theappointments be made by the Governor, toserve at his pleasure, and that localgovernment representation be made fromnominations submitted by the Associationof Minnesota Counties, by the League ofMinnesota Municipalities, and jointly bythe Minnesota School Board Associationand the Minnesota Association of SchoolAdministrators. Finally, it is recommendedthat the State Director of InformationSystems be chairman.

It is recommended that the LocalGovernment Section of the Division ofInformation Systems serve as staff for theIntergovernmental Commission. Theadvantages of this over establishing a newgroup are that (a) the necessary interfaceswith state agencies will be more readilyattainable, (b) staff costs will be

minimized, (c) staff specialists from theDivision of Information Systems will bemore readily available and more effectively

2-29

used, and (d) the relationship between therecommended chairman (Director ofInformation Systems) and the staff shouldbe most effective. The disadvantage is thepossibility that the State's interests may begiven undue weight, a possibility that theCommission itself should be able tocontrol.

It is recommended that Area ComputerCenter be authorized by the Commissionand operated by the Area Computer CenterBoard under broad policies established bythe Commission to assure integrated,consistent, and economical systemsanalysis, programming, andcomputerization efforts. The AreaComputer Center Board, limited to a

workable number such as seven, should beelected by the local governmental bodiesinvolved with the Center.

In those cases where departments of theState government have supervisory orregulatory authority over localgovernmental functions, it is recommendedthat requests for Federal, State, orfoundation funding for development ofintergovernmental information systems orcomputerization require approval by theState department involved prior tosubmission to the IntergovernmentalCommission for final approval.

Approaches to Cooperative Centers

The only cost effective computer cooperation formost local governments in Minnesota is thecommon use of systems and computers. Thealternative methods are essentially of three types:geographically oriented, such as a county, a groupof counties, or a State region; functionally basedserving school districts, for example; or acombination such as a State region wherein oneArea Computer Center serves all school districtsand another serves all counties and municipalities.

Although the recommendation of specificlocations is beyond the scope of this study, it isrecommended that the presently designated Stateregions be used as a basis for preliminary planningof Centers. A Center should cover one or more

2-30

complete regions or a part of a heavily populatedregion, but not parts of two or more regions.

Major factors to be considered in establishing aCenter are: (a) a service base sufficient to justifythe center: total population and schoolenrollment are principal examples; (b) generalreadiness of the potential clientele to make use ofthe facilities; (c) adequate financing; (d)contractual agreement as to charges, services, andcoordination personnel; (e) geographic area anddistances; and (f) communications cost/benefit.These criteria are dynamic because of externalforces and because of their mutual interaction.

Area Computer Centers

Against the background of the precedingdiscussion, specifics of Area Computer Centersare now considered. There are three alternativesthat appear to be feasible in Minnesota.

A Center created anew under authority ofthe Intergovernmental Commission. Anexample of this type of Center is TIES, inwhich case twenty school districts set up anew computer center under the JointPowers Statute, and employed the systems,programming, and operating staff to carryout all phases of development andoperations.

A Center already existing in highereducation. Certain of the larger highereducation institutions have or will havecomputer configurations with the capacityto handle information systems andcomputerization requirements of localgovernments including school districts.Provision of such service would beconsistent with the institution's publicservice objectives. The development ofinstructional uses by such an institutioncould be most helpful to school districts.Depending upon developments in type andquantity of utilization over the next fewyears and the composition of the highereducation network as discussed in Part IVof this report, the institution may find thatit would be faced with a choice of servingonly higher education, limiting its service

i

to higher education in favor of serving localgo% ernment, or greatly increasing itscon iputer capacity.

A Center already existing in an AVTS.NoAVTS now has a computer configurationcapable of handling the kinds ofinfc rmation systems and computerizationcall ml for in this report. In fact, most ofthe present computer configurations in theAVTS's are less than adequate for theedu;:ational needs of those schools becauseof their size and generation. The largercor figuration necessary for an AreaComputer Center could fill the educationalneed. Further, the programming andoperational requirements of a Centerwot.ld offer substantial opportunity forexcollent student training. The AVTS's,und3r school district control, are a logicalcho ce to provide service to school districtsand their schools. In addition, the AVTS'sin many communities enjoy a closerelalonship with county and municipalgovcrnments.

Orientation and Training

Information systems, involved as they are withthe broad, complex, and largely innovativeconcepts of systems analysis andcomputerization, simply cannot be developedwithout a significant training effort. Curriculashould be Established for elected officials, for topand middle management, for supervisorypersonnel c irectly involved in services likely to becomputerized, and for local governmentpersonnel who will form or be the coordinativelink with, the information systems cadre.

Educatior al programs of va!ue to localgovernment information systems efforts shouldbe catalogued and periodically disseminated tolocal government through the existing channels ofthe Association of Minnesota Counties, theLeague of Minnesota Municipalities, and theMinnesota Association of School Administrators.

Of equal or perhaps greater importance is theneed for such emphasis in teacher trainingsequences. Recommendation No. 7 in Part IV of

I

this report states, "The major problem inimplementing modern information systems andcomputer instruction technology in local schooldistricts is a lack of people trained in thesefields."

Funding

It appears unrealistic to expect developmentalfunding by local government alone. Somecontribution, however, should be expected fromlocal government, perhaps on a matching basis.Because the need is so great and the time so short,it is recommended that the State make available asubstantial sum for developmental work. Thismoney would be available for expenditure by theDirector of Information Systems uponrecommendation of the IntergovernmentalCommission on Information Systems. The fundswould be available for developmental purposes. Inthe main, they should be used to develop orobtain systems and to fund pilot projects, butthey should not be available to subsidize anon-going service to a local government.

The amount recommended from the State budgetfor the next biennium is $1,000,000 for use indevelopment of county and municipalinformation systems, $700,000 for schooldistricts, and $200,000 as the initial budget forthe Local Government Section. Thecounty-municipality sum should be available foruse through one or more Area Computer Centerson a matching basis (for example, $3 of statefunds for $1 of local funds). The distributionbetween county and municipal systems should bedetermined by the IntergovernmentalCommission with a principal criterion being earlypay-off.

The school district distribution is suggested as a$1 augmentation of school aid per year (or $2 thefirst year) for all of the students in TIES,approximately 240,000, and all of the students inthe second TIES-like Area Computer Center,assuming the new center covers approximately100,000 students. The money would be madeavailable by the Director of Information Systemsto TIES and the new school district Area

2-31

Computer Center upon recommendation by theIntergovernmental Commission after reviewingplans for development to be financed by thefunds. The TIES suggestion assumes that all TIESsystems and programs will be available withoutcharge to other Minnesota jurisdictions.

The amount of money recommended could easilybe doubled and spent productively.. The odds arehigh that every dollar invested in this way willreturn an extremely rewarding pay-off to thetaxpayers in Minnesota. It would be appropriate,of course, to seek Federal and foundation fundingassistance for both of these efforts. Pooling ofresources and efforts among the larger units oflocal government and the state in specific subjectmatter areas should also be attempted. Inparticular the possibility of Federal fundingshould be aggressively pursued by theCommission, either as a Minnesota project or incombination with other states. Further, jointdevelopment action with other states should beexplored.

IMPLEMENTATION SUGGESTIONS

Specific action to implement the generalrecommendations above is outlined below. Theseactions assume legislative adoption and enactmentof the general recommendations made previously.An interim approach is also suggested so thatdevelopment can proceed prior to such action.

1. The Intergovernmental Commission onInformation Systems should authorize anArea Computer Center dedicated to thedevelopment and implementation ofinformation systems for county andmunicipal governments. It is recognizedthat accomplishing this task may bedifficult. However, the local governmentneeds, the availability of some Stateand/orFederal funding, and significant Statetechnical assistance should provideincentive to a group of progressivejurisdictions.

2. The Intergovernmental Commission shouldspell out its delegation of authority andassignment of responsibility to the Area

2-32

Computer Center Board and then conductan election for the Board.

3. The Area Computer Center Board shouldappoint a Standards Committee and aSystems Development Committee. TheStandards Committee, assisted by theLocal Government Section, Division ofInformation Systems, should undertake thesizable task of drafting a standards manualfor local government in Minnesota,including among other things, a datadictionary. Every opportunity should bemade available for other local governmentinput. Draft manuals should be circulatedto all counties, all municipalities (over2,500 population), TIES, the MetropolitanCouncil, and the State Department ofEducation for review and comment. Theinitial standards manual should ultimatelybe presented to the IntergovernmentalCommission for approval and would thenserve as the local government standardsmanual.

The objective of the Systems DevelopmentCommittee is to obtain or develop soundinformation systems. The systems selectedfor initial development should be thosewith greatest payoff. Further, they shouldbe capable of modular development so thatpayoff may be realized from part of thesystem. As with standards, the systemsdesign must be submitted to alljurisdictions for review. Local governmentswith particular interest, expertise, oraccomplishment in related systems shouldbe invited to participate in development oftransferable information systems. TheLocal Government Section should devotemuch of its effort to the work of thisCommittee.

4. In the initial stages, the Area ComputerCenter should make use of a computercapability other than its own. It is

expected that only a sizable computerconfiguration can handle the informationsystems that will be developed. The Centersimply could not utilize such equipmenteffectively at the outset.

...

I

)

I

iI

I

1

5. The implementation of all except the mostproven of existing systems should be

limited to one user. Until extensive testingis completed, no additional users should beallowed to use a new application.Comprehensive user's manuals must be inbeing prior to implementation.

6. The Area Computer Center Board shoulddevelop written policies, regulations, andprocedures governing the operation of theCenter including a model contract forservices; input and output controlrequirements; security, privacy, andconfidentiality of data; charging forservices rendered; etc. These should bereviewed and approved by theIntergovernmental Commission inasmuchas they should serve as a pattern for othercenters.

7. The Area Computer Center Board, in

consultation with the Intergovernmental

Commission, should establish an operatingplan for the first two years and a

corresponding budget.

8. Selection of a director for the center andan appropriate site should be made by theArea Computer Center Board.

9. The Intergovernmental Commission shouldauthorize a second Area Computer Centercomposed of school districts, preferably bebased on a region (MESA and conduct anelection of the Board. The Center shouldnot begin from scratch but should utilizesystems now operational such as TIES orthe Honeywell Newport-Mesa, EDINETcombination.

Prior to legislative action, the State-Local DataSystems Group and the Local Government Unitof the Computer Services Division should lay asmuch groundwork as possible for the initiation ofthe foregoing program.

2-33

I

I

f

I

1

t1

PART IIISTATE

TABLE OF CONTENTS

Section Page

1. INTRODUCTION 3-1

MINNESOTA BACKGROUND 3-1

Systems and Computer Acceleration 3-2Executive Order 15 3-2Executive Order No. 56 3-3

OBJECTIVES AND SCOPE OF THE STUDY 3-4Agencies Included 3-4Study Methods 3-4Agency Relationships 3-6

2. CURRENT STATUS AND PROJECTED REQUIREMENTS 3-7COMPUTER CONFIGURATION 3-7OPERATING PHILOSOPHIES 3-7APPLICATIONS 3-9MANAGEMENT EFFECTIVENESS 3-9

Systems Design Concepts 3-9Personnel 3-10Equipment 3-11Minnesota Computer Utilization 3-12

COMPUTERIZATION COSTS, WORKLOAD, AND FUNDING PROJECTIONS 3-14Past Expenditures 3-14Trends Which Will Affect Future Requirements 3-14Potential Budget Levels 3-15

3. COST/BENEFIT CONSIDERATIONS 3-23DOES COMPUTERIZATION REALLY PAY? 3-23

Income Tax Processing 3-23Engineering Tasks 3-24Another State 3-24Indirect Cost/Benefit 3-25

BENEFITS THROUGH CENTRALIZATION 3-25

4. THE MINNESOTA SYSTEMS PLAN 3-29IMPLEMENTATION OF MIDS 3-29

An Effective Organization Structure 3-30Planned, Controlled Development of MIDS 3-33

REFERENCES 3-41

I

I

I

LIST OF FIGURES t

FigurePage

2.1 Minnesota State Government Computer Configurations, 1959-1970 3-8

2.2 Projected Workload State Government Computer Operations 3-17

2.3 Seven State Cornrarison Estimated Value of Computer Hardware Per 3-19

State Resident (Based on Purchase Price)

2.4 Seven State Comparison Estimated Value of Computer Hardware Per 3-19

State Government Employee (Based on Purchase Price)

2.5 Expenditures and Projected Funding Trends State Government 3-20

Computer Operations

3-ii

1

I

r.,

I

LIST OF TABLES

Tab:e Page

1.1 Agencies Served in Fiscal Year 1970 by the Computer Services Division 3-5and Selected Agencies in Study

1.2 Information Interfaces Among the Sixteen Selected State Agencies HigherEducation, Local Government, and Federal Government

3-6

2.1 State Government Average Monthly Computer Utilization by Calendar 3-13Quarter

2.2 Total Costs Minnesota Systems and Computer Operations 3-14

2.3 State Government Estimate Major Applications by 1975 3-16

2.4 Total Expenditures and Annual Growth Rates for State of Minnesota 3-18Government and State Computer Operations (Dollars in Millions)

2.5 Minnesota Estimated Costs of Computer Hardware and Information 3-20Systems Per State Resident and State Employee 1970

2.6 Funding Projections for State Government Computer Operations (Dollars 3-21in Millions)

3.1 Minnesota Income Tax Processing, Seasonal Hours vs. Returns Filed 3-24(Figures in Thousands)

3.2 Estimated Cost of Administering Minnesota Income Tax Laws FiscalYears 1964-1969

3.3 Estimated Computing Power as a Function of Monthly System Leasing CostsIBM 360 Series

3.4 Estimate of Equipment Rental and Other Costs for Six Separate SiteConfigurations vs. A Centralized Configuration

3-24

3-26

3-26

3-iii

1. INTRODUCTION

State governments are playing an increasinglypivotal role n the affairs of the nation. Publicservice programs, such as natural resources andhighways, traditionally state administered, aretaking on new dimensions as are programs withheavy Federal overtones administered by orthrough the states, such as manpower services andwelfare. State-local relationships have taken ondifferent meaning as area considerations becomemore common and as the Congress and Federalagencies look more to the states as intermediateal locators of or advisers on Federal funds for localuse.

Information systems play an ever-increasing partin these activities. It has been suggested in theintroduction to this report that informationsystems can be the most important factor ingovernmental efforts toward the good life.

I n this introductory section of the Stategovernment part of the report, the background ofsystems and computerization in Minnesota is

reviewed and the approach to this study outlined.

Section 2 summarizes the present status of theState in terms of operating philosophies,personnel, equipment, applications, and costs.Estimates of potential State budgets for thosepurposes are presented together with historicalequipment cost comparisons for selected states.

In Section 3, cost/benefit data are presented.These data include limited, quantitative directcost/benefits and examples of indirectcost/benefits. Also included are data concerningthe cost/benefits of computer centralization.

Section 4 contains recommendations forimplementing MIDS, an acronym for MinnesotaInformation and Decision Systems. The necessarysystems approach, organization structure,guidelines for growth, data security and

confidentiality considerations, andintergovernmental coordination are included.

MINNESOTA BACKGROUND

I n 1957, the Minnesota Legislature enactedlegislation establishing a Central Services Divisionin the Department of Administration. Among thedivision's several functions was providingtabulating services to departments of the Stategovernment. When appropriate space was madeavailable in 1960, tabulating installations fromseven agencies of the State were merged andorganized within the Central Services Division.Installations in the Departments of EmploymentSecurity (now Manpower Services) and Highwaysand in the offices of elected offirjals (StateAuditor, State Treasurer, and Secretary of State)were not included. The Department of Highwaysobtained the State's first computer in 1959, anIBM 650. The Central Services Division obtainedits first computer, an IBM 1401 card system, in1962.

A principal result of the centralizationaccomplished ten years ago has been thatMinnesota has avoided much of the extensive andexpensive proliferation of computer equipment,personnel, and space that inevitably accompaniesseparate computer operations. In so doing,Minnesota accomplished in large measure whatmany states are now beginning to move towards

namely, a centralized, or at least a highlycoordinated, systems and computer operation.Although Minnesota was fortunate in its earlyapproach to the computer organization andoperation, it did not fare so weal in developmentof its systems philosophy and in laying out itslong range plans for systems and computer efforts

shortcomings it apparently shared with allother states.

3-1

The systems analysis functions of the State havenot been clearly recognized until recently. Thedesign of tabulating and computer systems wasgenerally limited to accomplishing someoperational objective. Little attention wasdirected to management informationrequirements. Further, the systems were generallylimited to the "computerizable" characteristics.Overall systems analyses of activities ofgovernment including a review and clarification ofobjectives and a comprehensive review of allmatters pertaining to the attainment of thoseobjectives such as organization structure,policies, all processes (not just machine relatedprocesses), personnel assignment, controls, etc.simply were not undertaken.

The Central Services Division computer activitieswere financed and operated under a revolvingfund. The division had no funds or personnelavailable for planning and development effort. Itscapabilities and services gradually grew inresponse to departmental demands, but thedivision was in no position to give strongdirection. Conditions were different in theHighway Department computer operation. Solidmanagement support and adequate financing andstaffing resulted in excellent planning.

The growing need for state information systemswas recognized by the State Planning Agencyshortly after it was organized in June, 1966. In.i he fall of that year, a report was prepared undercontract with the State Planning Agency entitled"A Plan for an Information System for MinnesotaState Planning." 1 In the summary of that reportis the following paragraph.

IThe conclusions from these meetings were that(a) the state has the beginnings of a data base; (b)there remains a large effort to develop a

comprehensive data base; (c) computerpersonnel and facilities must be greatly expandedif the data base is to be fully computerized; (d)top state management is thoroughly receptive tothe development of a system, and not resistant toa strong central system; and (e) some statedepartments are already moving ahead withdevelopment of departmental informationsystems. In short, the state has a long way to go

3-2

1

in developing an information system, but theclimate in terms of management understandingand support is extremely promising. What is nowneeded are adequate personnel and computerfacilities.

Systems and Computer Acceleration

In 1967, the Executive and Legislative branchesteamed up on action aimed at strengthening thestate effort. Governor Harold LeVanderproposed, and the Legislature enacted, a specialappropriation of $500,000 to acceleratecomputerization of State systems. This legislationrequired that the computer function be set up asa separate division of the Department ofAdministration. The Governor then establishedthe Governor's Committee on State InformationSystems to advise on the development of theState's systems and computer efforts. Thecommittee, composed of persons from privateindustry with extensive experience in themanagement of large systems and computeroperations, has played an important role since itsappointment in July, 1967.

In January 1968, the Computer Services Divisionwas established in the Department ofAdministration. Shortly thereafter, the Governorissued Executive Order No. 15 in which hedefined the division's objectives as follows:

To plan the orderly and efficient developmentof information systems and other computerservices for all agencies of the state government;guide the implementation of the plans so thatoptimum systems and computer integrationmay be accomplished; administer or coordinatethe administration of systems, programming,management science, and computer operations;and provide for controls over accomplishmentof the objectives to the end that the State ofMinnesota will have a highly effective andefficient overall systems, management science,and computer capability.

The Executive Order also specified broadresponsibilities and authority to carry out thoseobjectives.

9,,

we

.....

Governor's Committee and the Commissioner ofAdministration, the Governor took a further stepto enhance computer development. He issuedExecutive Order No. 56, superseding the earlier

The results of the order were a gradual beginningtoward further coordination and integration ofthe State's computer development efforts and ofactivities leading to vastly improved coordinationand cooperation in higher educationcomputerization. The Governor's Committee wasdeeply involved in all of these efforts and waskey in their development. Both gubernatorial andlegislative support were evidenced in the 1969Session by a second special appropriation forcomputer development $600,000 for theComputer Services Division and $300,000 forhigher education, to be spent on authority of theCommissioner of Administration followingrecommendation of the Governor's Committee.

By the end of 1969, two years after itsestablishment as a separate division, theComputer Services Division was materiallystronger in its personnel and equipment resources.I t was moving aggressively and was deeplyinvolved in development, implementation, andoperation of advanced systems applications for anumber of state agencies. Notwithstanding thegeneral progress and the support and satisfactionof many users, a number of users have beendissatisfied with results of computerizationefforts. Dissatisfaction ranged from mild tosubstantial and resulted from a wide range ofproblems. Most often the reported problems werea "joint product" of CSD and the user. As agenerality, satisfaction appeared to be in directproportion to the extent of user top man?, ^-nentinvolvement and the systems sophistic on ofuser personnel. An example is the A/IINCIS(Minnesota Crime Information System)development and implementation which involvedand still receives heavy support and participationfrom Bureau of Criminal Apprehension topmanagement.

Executive Order No. 56

In January 1970, upon recommendation of the

order, making specific his approach in severalimportant areas and directing the Commissionerof Administration to take action to attain greatercoordination and integration of the State'scomputer effort in the State administration, inhigher education, and in local government.

Shortly thereafter, the Commissioner ofAdministration announced a major reorganizationof the Computer Services Division aimed atmeeting the division's needs for planning, forcontrol, for operations, for systems design andprogramming, for coordination with highereducation and local government, and forimprovement of services to the using agencies.This reorganization has recently been completed.(Parallel to this has been a reorganization of theentire Department of Administration in which theComputer Services Division was renamed theInformation Systems Division. In order to avoidconfusion with the future organization structurereferred to herein, the title of Computer ServicesDivision is used throughout the report.)

The Commissioner of Administration, underauthority of Chapter 1129, Laws of 1969, thenmerged the Department of Highways computercenter and staff into the Computer ServicesDivision. This was an extremely important step.Its accomplishment is a tribute to the Governor,the Commissioner of Administration, and theCommissioner of Highways. The latter official'swillingness to relinquish control over a large,successful computer operation typifies thecooperative spirit among top State officials thatwas encountered repeatedly in the course of thisstudy.

Thus, with the exception of the Department ofManpower Services and a few tabulating activitiesin agencies not under the Governor's authority,Minnesota has centralized responsibility forsystems design and computerization and for thedevelopment of an integrated information systemfor the state. The Department of ManpowerServices, although administering its owncomputer center and related activities, is underthe computer coordinating authority of the CSD.

3-3

It should also be noted that some majordepartments of government have one or morepersons comprising a systems staff in thatdepartment to assure systems design responsive tothe needs of the department.

Since its establishment as a separate division ofthe Department of Administration in early 1968,the Computer Services Division has been engagedin a broad variety of activities strengthening theState's systems and computerization capability.Principal among these are:

Establishing a foundation for moreeffective internal administration of thedivision and for improving userrelationships by functional reorganizationand strengthening staff.

Reviewing all proposed expenditures oncomputerization to assure consistency withexisting broad State plans and objectivesand elimination of duplication.

Converting virtually all of its operation to a"third generation environment" under fulloperating systems.

Attaining a high level ofmulti-programming (handling more thanone task on a computer at the same time)and attaining a high level of utilization ofthe central computers.

Implementing in conjunction with users anumber of major new systems.

With the assistance of the Civil ServiceDepartment, revising job classifications andpay structure to support the reorganizationand to enable the State to compete moresuccessfully for systems and computerpersonnel.

Uti I i zing more systematic and provenmethods in systems design and analysis.

During this period, the computer efforts of theDepartment of Highways continued to produceoutstanding results including the implementationof sophisticated computerized systems for thedriver license and motor vehicle operations. Insummary, substantial progress has been made inrecent years, and extensive efforts have been

3-4

made to coordinate systems and computerdevelopments in the State.

OBJECTIVES AND SCOPE OF THE STUDY

This study has as its primary objectiveestablishing a plan for state information systems.The plan, called MI DS, an acronym for MinnesotaInformation and Decision Systems, includes aseries of specific recommendations and a numberof suggested guidelines.

Agencies Included

Time and funds prevented including all Stateagencies in the study. The sixteen agenciesselected are shown in Table 1.1 and includedtwelve departments from the Executive Branch,the Legislature, and the three principal retirementagencies. Selection was based for the most part onsize of the agency in terms of both budget andpersonnel and on its present and estimatedpotential need for information systems. Theretirement systems were included because of theirclose relationship to the personnel informationsystem of the Departments of Administration andCivil Service. The twelve executive branchagencies account for over ninety percent of thetotal State budget, of the State employees, and ofState systems and computer expenditures,exclusive of higher education.

Study Methods

A work plan was developed to provide guidelinesfor the State agency portion of the study.Interviews were he,.; with departmentalmanagement to determine ,nformation needs. Theconsultants met each Commissioner or Directorof the sixteen selected agencies, the Deputy andAssistant Commissioners, and the Directors ofeach division in those agencies. EDPCoordinators, where such personnel existed, weremost helpful. Additional persons wereinterviewed as suggested by agency heads.

Following the interviews a position paper wasprepared for each of the sixteen agencies. Source

,-

Table 1.1. Agencies Served in Fiscal Year 1970 by the Computer Services Divisionand Selected Agencies in Study

Legislature*

Supreme Court (Revisor of Statutes)

STATE DEPARTMENTS:

Administration*

Aeronautics

Agriculture

Civil Service*

Commerce*

Conservation*

Corrections

Economic Development

Education*

Health*

Highways*

Iron Range Resources & Rehabilitation

Labor & Industry*

Manpower Services*

Military Affairs

Public Safety*

Public Welfare*

Taxation*

OTHER STATE AGENCIES:

Board of Hairdressing

Board of Nursing

Governor's Office

Mankato State College

Minnesota Higher EducationCoordinating Commission

Minnesota Historical Society

Minnesota State Retirement System*

Pollution Control Agency

Public Employees Retirement Association*

Southwest State College

State College Board

State Fair Board

State Junior College Board

State Planning Agency

State Police Officers Retirement Association

State Treasurer's Office

Teachers Retirement Association*

University of Minnesota

OTHER AGENCIES:

Hennepin County

Ramsey County

U.S. Bureau of Public Roads

U.S. Office of Economic Opportunity

Veterans Administration

U.S. Postal Service

*Sixteen selected agencies included in study.

material for these papers included notes trom theinterviews and publications. The papers describedthe agency organizational structures and theirfunctions, existing computer systems, systems inthe planning or implementation stage, andpotential applications that the departments wereconsidering. The interrelationships of eachdepartment with other agencies in the study werelisted.

The Computer Services Division provided a Stateinventory of systems and computer personnel and

equipment. The Division also obtained detaileddescriptions of all computer applications in theState agencies. Included in the equipmentinventory were any additional data processingequipment or replacements on order. Thepersonnel inventory included all employees in theEDP section in each department and allmanagement, systems, programming, operations,and data entry personnel in the computer centersof the State. The application inventory included adescription of the system, computer timerequirements, languages used, and number ofprograms for each application.

3-5

The above data were supplemented by referenceto the experience and plans of other states as wellas by the consultant staff background in Stategovernment.

Agency Relationships

One of the most important factors in thedevelopment of information systems is the

3-6

inter-agency commonality of data. Table 1.2illustrates the existing information relationshipsbetween the selected agencies. In some cases theserelationships may be strictly procedural, and inother cases they are highly substantive. In allcases, however, they indicate the requirements forsome type of information exchange. Theimplications of this in the development ofinformation systems are sizable.

Table 1.2. Information Interfaces Among the Sixteen Selected State Agencies Higher Education,Local Government, and Federal Government

111111,

<<,<

2. CURRENT STATUS ANDPROJECTED REQUIREMENTS

COMPUTER CONFIGURATION

Computer operations for State functions areorganizationally unified with the exception of theDepartment of Manpower Services. The personneland equipment of the Computer Services Division(CSD) are still housed in the same two areas(Highway Building and Centennial Building)occupied prior to the recent organizationalunification. The Manpower Services computerstaff and equipment are housed in thatdepartment's building in downtown St. Paul.

The Computer Services Division computerconfiguration presently consists of a UNIVAC418-11, two IBM 360-50's, one IBM 360-40, andone IBM 360-25. Two of the computers arehandling telecommunications systems theUNIVAC 418-11 serving MINCIS and an IBM360-50 handling certain engineering applications.Figure 2.1 shows the gradual development of thestate's computer resources since 1959.

The IBM 360-50's and the 40 are operated underfull operating systems twenty-four hours a day,seven days a week. One 360-50 has core memoryof 384,000 bytes; the other 360-50 and the360-40 have 256,000 bytes of memory. Over 2.1billion bytes of random access memory arepresently available on the systems. (For purposesof this discussiorl, one byte may be viewed as onealphabetic character for two numeric characters.)

The IBM 360-i ) handles the large driver licenseand motor venicle applications which are basedon random access files (about 2,300,000 recordsin each file). These files are interfaced with theUNIVAC 418-11 computer which handles MINCIS(Minnesota Crime Information System).Currently, MINCIS includes message switchingand data retrieval from 161 teletype machines ona state-wide network and is linked to the NationalCrime Information Center in Washington, D.C.,

and to the National Law Enforcement TeletypeSystems (N LETS) centered in Phoenix, Arizona.The I BM 360-50 at the Highway Buildinginstallation can serve as back-up to the 360-40 ifneeded.

The Manpower Services computer activities,presently based on a second generation computer(IBM 1401), will be converted in the near futureto a third generation computer (IBM 360-40).Three major applications are scheduled forimplementation on the new computer in additionto the applications already existing. All three ofthe new applications have been or are beingdeveloped centrally by the Federal governmentfor implementation in all states.

OPERATING PHILOSOPHIES

The Computer Services Division (CSD) is heavilyinvolved in on-line input and updating of on-linefiles available for data retrieval. CSD is committedto centralization of the central processing units,to multi-programming, to operations controlledby the most advanced operating systems, toremote terminals when approximate for on-lineentry and on-line retrieval, to whatever variety ofdata collection and input methods will attain themost efficient results, and to integration ofinformation systems and the standardizationimplicit therein.

These are extremely significant decisions thathave far-reaching implications for theadministration of the business of the State. Forexample, some of the major systems involve entryof data via CRT (Cathode Ray Tube) terminalsdirectly into the computer files, updatinginformation already in the files via CRTterminals, and retrieval of data via CRT terminalinquiry directly from the computer data files.CSD is developing its plans on the assumptionthat economics and operating improvements will

3-7

x

CO

MP

UT

ER

SE

RV

ICE

SD

IVIS

ION

12 K

65 K

81K

81K

'UN

IVA

C41

8-11

II

UN

IVA

C41

8-11

128K

384K

.. ,,,

...

...../

IBM

I14

01IB

M36

0/30

IBM

'60/

40IB

M36

0/50

.0'

,...

.0'

IBM

1401

IBM

1401

0....

c....

/IB

M14

01I

IBM

I14

01IB

M14

01IB

M36

0/25

.4K

4K16

K

DE

PA

RT

ME

NT

OF

HIG

HW

AY

S

1

IBM

1401

MO

TO

RV

EH

ICLE

RE

GIS

TR

AT

ION

IIB

M

I14

01

4K

131a

m60

/40

IIB

M36

0/40

128K

256K

IBM

650

I IB

M14

10I

IIB

M14

10I

I IB

M14

10I

IBM

1410

[IBM

I60

/50

IBM 0 5

.s

40K

40K

40K

60K

256K

256K

256K

DE

PA

RT

ME

NT

OF

MA

NP

OW

ER

SE

RV

ICE

S

-

81(

8K8K

12K

1

128K

[ IB

M24

011

1401

11

1401

I14

0111

401

I 1

I ID

Ai

1401

IBM

340/

40

1959

1960

1961

NO

TE

: DO

ES

NO

T IN

CLU

DE

HIG

HE

R E

DU

CA

TIO

N C

OM

PU

TE

R E

QU

IPM

EN

T

1962

1963

1964

1965

1966

1967

Fig

ure

2.1.

Min

neso

ta S

tate

Gov

ernm

ent C

ompu

ter

Con

figur

atio

ns, 1

959-

1970

1968

w

1969

11--

----

1R

--".

"^"1

1

w

1970

rons

......

,1

eventually require virtually all State applicationsand their associated files to be computer- basedand available on-line.

APPLICATIONS

Applications handled by the Computer ServicesDivision computers include a great variety rangingwidely in complexity and computer requirements.Four major on-line systems provide inquirycapability into large files the driver license andmotor vehicle applications, MINCIS, and incometax. Heavy use is made of CRT input equipmentfor on-line entry and update and for dataretrieval.

Additional major applications include sales tax,withholding tar, corporate income tax,institutional billing, and central payroll for allstate employees. A massive amount of work isbeing handled for many engineering applicationsfor the Department of Highways. Otherapplications include the accounting functions forDepartments of Administration and Education,Minnesota State Retirement Association andTeachers Retirement Association receipt anddisbursement processing, Division of VocationalRehabilitation client information, Department ofEducation teacher personnel, and vital statisticsfor the Departmem of Health. These are but afew of over three hundred applications now inproduction on the State's computer systems.

Applications now under development includeexpansion of the MINCIS system to a largecriminal justice name file with inquirycapabilities, a pupil information system forEducation, insurance verification and workmen'scompensation case control for Labor andIndustry, land information system, revision of thesales tax system, and many others. Enormouspressures will be placed on CSD and on the usersin implementing pending applications. The crucialactivity is the effort required in systems design toassure a highly productive end result an effortthat must be jointly accomplished by the user andCSD. Beyond that, the programming and testingschedules must be realistically set. Perhaps most

often underestimated, the conversion of data andimplementation completion must be carefullyplanned and soundly financed. Finally, theremust be resistance to the plea for expediency inviolation of long range plans and priorities.

The extensive systems design, programming, andoperating experience gained to date by Statepersonnel in teleprocessing, on-line entry andupdate, and "real-time" retrieval of data inshort, in the most sophisticated kinds ofcomputer applications will materially aid CSDin moving ahead with the extensive backlog nowfacing it.

MANAGEMENT EFFECTIVENESS

Obtaining the greatest value for dollars expendedin systems analysis and computerization is one ofthe toughest challenges facing management, bothpublic and private. There are many factors whichmust be carefully managed, all of which areinterrelated. Principal among them, however, arethe quality of systems analysis and design;personnel performance in systems, programming,and operations; and type and utilization ofequipment.

Systems Design Concepts

Understandably, the introduction and almostunbelievable acceptance and expansion of the useof computers in government and industry wasaccompanied by a pre-occupation with thecomputer equipment. Here was a new tool ofincredible power and potential. With each passingyear in fact, with each passing month greaterspeeds, increased capabilities, broader horizonsappeared,and much was accomplished.

Gradually, however, as the novelty wore thin, asperformance failed to meet expectations, and ascomplexities compounded, re-appraisals ofcomputerization became increasingly common.Clearly, computer equipment was not at the rootof the problem. More and more, attention wasfocused on objectives of computerization. Insimplest terms and for many different reasons,most failures, and there were many, lay in thebroad area of systems design.

3-9

Existing manual, bookkeeping machine, andtabulating machine applications were oftensimply "put on" the computer. Little if anyattention was given to the computer abilities thatcould radically change and economize operationsor to the informational needs that could beserved. Certainly in the early days virtually nothought was given to integration of existingsystems. Even when identified, however, thisfundamental problem of poor systems design hasgenerally not been quickly resolved. There arethree reasons for this which must be wellunderstood if appropriate planning for systemsanalysis and computerization is to beaccomplished and then successfully implementedin Minnesota.

Organizing the Systems Function

The first of these is the organization of thesystems function. The systems approach must bebased upon the definition of broad objectiveswhich are to be accomplished by the function tobe systematized. It then proceeds to the detailedspecification of all activities essential toaccomplishing the function. This detailed effortmust involve all administrative activity (not onlycomputerizable activity), including organizationstructure, plans, policies, procedures, andmethods. Unfortunately, and this has often beentrue in the State's efforts, many so-called systemsefforts have been in fact "computer methods"oriented. Unless the overall systems approach istaken and unless 'responsibility for all systemsanalysis is centered in one organization,cost/benefit relationships can never be optimized.

Nature of Systems Design

The second reason is the nature of systems design.It is extremely difficult work. To be effective, itmust consider all facets of the system. Anadministrative system of any size is highlycomplex. Any one process in a system mayappear to be simple. As it is related to otherprocesses and as it is organized in such a way as toproduce the data necessary for managementcontrol and planning purposes, complicationsbegin to arise. As each of the necessary processes

3-10

is viewed in its complete detail, as all processesare interrelated, and as all management needs areprovided for, the complexity further increases.Another way to say this is that the managementof any activity is complex; the systems necessaryto assist with that activity are also complex. It isthis very complexity and the often resultantfailure or inability to design a comprehensivesystem that has lead repeatedly to disappointingr9sults.

Management Responsibility

The third reason for inadequate performance isfailure of management personnel, top throughmiddle, to become concerned with systems. Sincecomprehensive systems design vitally affects theirresponsibilities, managers must be deeply anddirectly involved. The professional systemsanalyst and the computer expert can be roosthelpful in many ways, but they cannot possiblyproduce a satisfactory product if managementabdicates its responsibility for specifying systemsrequirements. The management directive to "putit on the computer" can neverbe a successfulsubstitute for the rigorous management effortnecessary (a) to establish the systemsrequirements that is, to specify the "it", and(b) to clarify what the computer must produce.

Personnel

As computer hardware costs go down, computerpersonnel costs are rising. In the late fifties,computer hardware costs typically approximatedtwo dollars out of every three spent forcomputerization. Now a typical installation withall costs accounted for will show one dollar spenton computer equipment out of every three spenton computerization. The trend is still continuing.How long it will continue depends on manyfactors.

The important point here is that computerpersonnel costs are the most significant cost incomputerization. Obtaining maximum benefitfrom that resource will pay off accordingly. Lossof an experienced systems analyst or programmertypically represents a cost far beyond the

retraining and orientation costs of most otherprofessional personnel. This is because of thedetailed knowledge of a system or systems thathas been absorbed over a period of intensivestudy. The problem has been complicated inrecent years by the great demand for systems andprogramming personnel with resultant rapidturnover and constant training of new personnel.Recently because of the economic condition ofthe country and perhaps because of some basicfactors at work in the computer industry,turnover of such personnel has dropped. It is tooearly to detect with any certainty what futuredemand will be.

In any event, there is a very large payoff incontrolling turnover. Continuing technicaltraining, making available other avenues forprofessional improvement, the opportunity forapplication of knowledge, sound controls overprogress on projects, informed evaluation ofindividual performance, and carefully plannedpromotion and financial rewards are all vital inthis effort.

Equipment

Efficiency of the computer equipment and itsoperation is affected by many factors. Importantamong them are:

Internal speed and memory size of thecomputer

Multi-programming capability

Utilization level

Scheduling sophistication

Peripheral equipment availability

Productive vs. non-productive use.

Because of the importance of this subject andbecause of the interest that has been expressed init by legislators and by top state management, abrief review is included here. This discussion iscentered on the central processing unit (CPU) ofthe computer.

Internal Speed and Memory Size

Computer throughput that is, unit outputincreases substantially as the internal speed andmemory size of the computer increases. Over thefifteen years since introduction of the firstsignificant commercial computer, throughput perdollar of computer cost has increased radicallywith estimates varying from 20 to 50 times. Theprecise measure of increase depends upon manyfactors; the important point is that computerhardware costs per unit of output have been on aconsistent downward trend a trend that is stillcontinuing.

Multi-Programming

A second factor which takes on increasingimportance as larger computers are used is theextent to which the computer is being operatedunder multi-programming. "Multi-programming"is a term which means the ability of a computerto accomplish more than one task at the sametime. For example, part of the CPU (CentralProcessing Unit) and part of the peripheralequipment may be accomplishing one task,another part of the CPU and other equipment isdoing something else under control of anotherprogram, while a third part of this same CPU ishandling a communications system. Theefficiency of multi-programming is heavilyinfluenced by the sophistication of the operatingsystem (a program) which "manages"the internalactions of the computer.

Utilization

Another related factor is utilization of thecomputer. Computer rental charges are (a) basedon a one shift (176 hour) monthly usage plus asmall additional charge for overtime use or (b)established at a figure which includes unlimiteduse. In either case, the more intensively thecomputer is used, the less is the cost per unit ofoutput. The measurement of utilization is itselfdifficult; interpretation of the results of themeasurement must be carefully done to avoiderroneous conclusions. There are three measuresthat may be used to indicate utilization.

3-11

"Clock-tirbe" which indicates the numberof actual hours that the computer centralprocessor is operating that is, turned onand executing, or waiting to execute, aprogram.

"Billable time" which indicates the numberof "clock time" hours that the computerhas worked for the customers. Undermulti-programming, "billable time" mayexceed the number of clock hours.

"Meter time" which indicates the numberof hours during which the centralprocessing unit was actually executingprograms.

Scheduling Sophistication

The mix of programs run at any time undermulti-programming is important because of thepotential conflict over peripheral units such asprinters, random access devices, card readers, etc.To the extent possible in meeting deadlines,scheduling should minimize the conflicts (and theresulting waiting periods) so as to maximizeoutput. In the dynamic environment that typifiesmost large computer operations and certainly isfound in CSD, the scheduling task is complex andis extremely important in maximizing output.

Peripheral Equipment Availability

The proper mix of peripheral equipment is yetanother important factor in attaining hardwareefficiency. It is particularly difficult to establishoptimally because of the changing mix ofprograms. Often this decision is not so critical asdecisions on other factors since the unit cost ofsuch equipment is relatively low.

Productive vs. Non-Productive Use

A further factor in measuring computer efficiencyis the amount of make-ready time; unscheduleddowntime for maintenance; and re-run time, thelatter caused by program error, operator error,erroneous data, or computer malfunction. Theseare considered non-productive time. (Thequestion of the efficiency of the use ofproductive time is important but is related mainly

3 -12

to the quality of systems design andprogramming.) How much productive use is

obtained from a set time period may be measuredby billable time minus non-productive time.

In any event, an authoritative measure of theefficiency of the computer equipment and its usecan be determined only with a full knowledge ofthe computer configuration, operating systems,scheduling requirements, and detailed records ofproductive and non-productive time.

Minnesota Computer Utilization

The State of Minnesota central computers havegenerally experienced high utilization in recentyears. Utilization at the Centennial installationhas been extremely high, occasionally so great asto affect customer service adversely because ofschedule slippages due to overloaded systems.Funds were simply not available to add necessarycomputer capacity. Utilization of the IBM 360-50at the Highway installation has been reducedsince the 360-40 was installed to handle thedrivers license and motor vehicle systems.Therefore, capacity exists for the first time inrecent years to absorb increasing workloadwithout stress on existing applications. Thissituation, however, is only temporary.

Table 2.1 shows both the "meter time" and the"billable time" for the two principal computers."Billable time" is the total hours billed to usersand is obtained from computer-generated datashowing the elapsed clock time charged to eachjob. The figures on the chart exceed the numberof hours in the month, indicating thatmulti-programming is occurring. Note that theIBM 360-40 at MHD and the UNIVAC 418-11 atCSD both are scheduled to run continuously,except for preventive maintenance, 720 hours amonth. They are both handling on-line, real-timesystems that must be available at all times. The360-50 at the Centennial Building is alsoscheduled to run 24 hours a day, seven days aweek.

In February 1970, a 360-25 replaced a 1401computer at the Centennial installation. Its use

Table 2.1. State Government Average Monthly Computer Utilization by Calendar Quarter

IBM 360/40 (256K) 1 IBM 360/50Meter Hrs.

(256K)Billable Hrs.Meter -Hrs. I Billable Hrs.

Third Quarter 1968 382.97 596:.94 423.75 828.17

Fourth Quarter 1968 460.66 729.62 389.76 823.10

First Quarter 1969 524.07 917.22 442.24 886.08

Second Quarter 1969 587.65 1,637;03 458.78 1,208.40

Third Quarter 1969 575.90 1,402.99 468.84 559.68

Fourth Quarter 1969 569.64 1,399.65 436.13 609.14

First Ouarter 1970 461.42* 1,117.62 583.41 1,040.21

Second Quarter 1970 586.04* 1,597.86 497.47 599.93

*Model 40 was upgraded in February 1970 to a Model 50 (384K).

has increased from 209 billable hours in Februaryto 358 in July. Many presently operations'systems on the IBM 360-25 are processed underan "emulation" mode. This mode is generallyconsidered less than efficient. Much progress ispresently being made in the redesign andreprogramming of these applications for moreefficient third generation programming systems.As this work continues, it will slow the rate ofcomputer utilization increase because of greaterefficiency.

Computer utilization in the Department ofManpower Services has not kept pace with that inthe Centennial and Highway installations.Although precise figures were not available, use ofthe 1401 system during the past two years hasbeen running between one and two shifts or 176to 350 hours per month. This rate will fall stilllower with the installation of a substantiallylarger computer in the near future and is expected.....

to remain at a relatively low level for at !east ayear.

Total utilization of the State computers willgradually increase as additional applications are

_ added. BecaUse of the scheduling flexibilitiesinherent in the rather substantial base of

computer power, the utilization growth can bewell planned and controlled. Additional capacitycan be added when necessary in a mannerconsistent with overall plans. The recentreorganization of systems and computerresponsibilities enhances the State's ability toattain more efficiency in computer use. Maximumefficiency, however, will be possible only whenthe central computers are in a common location.It is important to note also that full compatibilitybetween the systems is necessary to attain highestutilization. This means, for example, that the two360-50 computers should be operated under thesame operating system, that programmingconventions must be identical, that teleprocessingand graphics network and interface controlprograms must be standardized. These conditionsdo not now exist.

The question is often asked: What is goodutilization of computer equipment? Thisquestion, as indicated at considerable lengthearlier, is extremely difficult to answer. Variousfigures have been suggested as saturation points,ranging from as low as 60 percent to usually notover 85 percent. But the figures are generallymeaningless because of the lack of a generallyaccepted definition of utilization. The one thing

3-13

that can be said authoritatively is that a groupingof computers provides the possibility for higherutilization for reasons discussed earlier. If forexample, an acceptable, carefully definedutilization rate for the computer in an installationwith one computer is set at 70 percent, such arate for the same computer in a computerinstallation with several computers could besafely set much higher perhaps as high as 90percent. This is true because of the flexibility of agroup of centralized computers to handle peakloads and unusual problems, whereas the singlecomputer must plan for the indeterminate impactof unusual problems as well as the effect ofunavoidable peaks.

COMPUTERIZATION COSTS, WORKLOAD,AND FUNDING PROJECTIONS

Past Expenditures

Table 2.2 shows the readily ascertainable cost ofsystems and computerization in Minnesota for thefiscal years 1966 through 1970. These figuresreflect costs for personnel, equipment, andrelated factors, for the most part, in theComputer Services Division, Department ofHighways, and Department of ManpowerServices, but also include such costs for otheragencies of government, exclusive of highereducation. They do not, however, reflect the costof non-computer systems personnel, data inputpreparation done by a user agency, or the inputand output control costs incurred by those usingcomputer services. Nor do they include thesignificant costs of space and power since the

State does not presently charge the computeroperations for either.

It is emphasized that whereas the costs within thethree main computer operations have been wellaccounted for and clearly reported, the total coststo the State of processing many applications arenot known. To carry out requirements of soundfeasibility analysis and subsequent evaluation ofapplications, such costs must be accumulated.These requirements are being considered in thedevelopment of a new State accounting system,currently underway.

Trends Which Will Affect Future Requirements

Pressure for increasing State budgets forcomputerization is going to continue in theforseeable future' as a result of a number offactors. Important among these are:

Basic economics of computerization vs.manual processing methods. Total costs perunit of computerized output and files aregoing down whereas sharp cost increasescharacterize manual procedures andmanually maintained files. Computerapplications, both operational andinformational, formerly marginal, willbecome economically feasible.

Increase in the volume of transactions inexisting applications, resulting frompopulation growth, affluence, or publicpolicy.

Increasing experience with andunderstanding of information systems andother computer capabilities by user

management personnel.

Table 2.2. Total Costs Minnesota Systems and Computer Operations

Fiscal Year 1966 1967 1968 1969 1970

Total Costs $1,999,000 $2,221,000 $2,934,000 $3,856,000 $5,105,000

3-14

I

I

I

ri

I

I

I

I

I

I

r.

Probable new social service programs, bothFederal and State.

Increasing requirements for data by Federalagencies.

Increasing sophistication in systems designresulting in part from continuingtechnological developments (as intelecommunications and microfilming) andin part from greater strength andexperience in systems personnel.

Increase in the State's servicing of localgovernment through computerization;MINCIS and its further development is anexample.

More rapidity and lower cost ofimplementing new systems because ofprogressively more efficient programming,more experienced programmers, and theexpected use of programs developed inother jurisdictions.

A number of these factors, while contributing tothe total effect of upward budget pressure forcomputerization, will at the same time be placinga significant downward pressure on total Statecosts. This will occur both through direct andindirect cost/benefit.

Potential Budget Levels

All projections of future budgets for informationsystems are compounded not only by the factorsmentioned just previously but by many otherfactors such as the extent and rate of inflation,the State salary plan, impact of technologicalchange, and the legislative decision as to controlover systems analysis and computerization. In thisconnection, it must be again emphasized that thehistorical costs are not all-inclusive since it wasnot possible in the course of this study to obtaincomprehensive cost data. The historical costs areunderstated, probably substantially. To theextent that future projections are based on thosecosts, they will likewise be understated. It is alsoemphasized that these are only suggestions. Theymay be materially different from actualrequirements both because of timing changes orany of the many factors mentioned previously.

F our factors were evaluated in developingpotential budget levels for the next five years:

The number of major applications.

Expenditure trends in recent years.

Information systems expenditure as a

percent of total State expenditures.

Comouter hardware cost comparison withother states.

The results of these evaluations are outlinedbelow.

Major Applications

A "major application" defined in a broad sense isa function performed by the computer system fora unit, division, or department. It may consist ofone or many programs. For example, in anapplication called Forestry Time Summary (toobtain Federal reimbursement on workperformed by Conservation employees in stateforests), three programs are required tosummarize hours spent on specific tasks. Anotherexample of a major application is the processingof the State payroll. This application is composedof more than 60 linked programs.

In June, 1970, the inventory of existing majorapplications for all State agencies was 332. Therewere then 45 additional major applications instages of systems design, programming, orimplementation. During the course of this study,174 new major applications were identified. Someof these are major enhancements to presentsystems, and others will require development ofnew systems and programs. Some will be found tobe impractical candidates for computerizationbecause of unfavorable cost/benefit analyses. Forpurposes of this analysis, it is estimated that 140

(80 percent) of these applications will ultimatelybe implemented in one form or another.

It is further estimated that 125 additional majorapplications will be identified and implemented inthe next five years, 75 by the 16 agencies in thestudy and 50 by all other agencies of the State.

3-15

This represents a total of 265 new majorapplications. Generally, the new applicationsboth because of their nature and because of thepresent level of sophistication of the State'sinformation systems effort will be far morecomplex than those currently implemented. It isconservatively estimated that these newapplications will require on the average twice thedevelopmental, operating, and maintenance costof the existing applications. Applying this factorto the 265 major new applications gives a 530equivalent figure to add to the existing 332 for atotal 1975 count of 862. It is estimated, however,that 100 of the existing applications (30 percent)wiii be eliminated by obsolescence or by revisionsso major as to be included in the newapplications.

This, then, would produce by 1975 the datashown in Table 2.3.

Figure 2.2 shows the above workload projectiondata in graphic form. Line A represents theanticipated decrease of 100 applications over afive-year period based on the el imination ofsome of the 332 applications existing in 1970.Line B shows the effect of the 140 newapplications which have been identified in thisstudy. Line C is the additional workload expectedfrom the 125 additional new, but as yetunidentified, `applications in the 1971-1975period. The trends shown in Lines B and Cprovide for the increased complexity of futureapplications by applying a factor of two asdescribed above. Lines A, B, and C arecumulative; Line C, therefore, represents, on a'straight line' basis, the total projection ofapplications workload.

The translation of this straight line relationshipinto annual costs required for implementationwill not result in a linear cost relationship. Therate of growth and associated costs of aninformation processing organization aredetermined by its ability to expand at a

reasonable rate, the priorities given toimplementation of various applications,limitations imposed by funding, and many otherfactors mentioned earlier. The overall trend,however, is clear. The workload in Statecomputer operations will increase by at least 130percent over the 1970 level in the next five years.

Trends in Information Systems Expenditures

The State's five-year information systemexpenditure history and the annual rates ofincrease are shown in Table 2.4. The high rates ofincrease in 1968, 1969, and 1970 reflect in partthe effort begun in 1968 to accelerate thedevelopment of information systems. Asdiscussed in detail elsewhere in this report,substantial progress has been made in many areasin the three-year period; the funds have beenproductively applied. Growth at that rate on acontinuing basis, however, may impose burdenson management that cannot be handled as

effectively as has been the case up to this time.

Information Systems Cost Related to the StateBudget

The substantial increase in information systemscosts has occurred during a period of sharpincrease in total State expenditures. The figuresfor the most recent five years are shown in Table

Table 23. State Government Estimate Major Applications by 1975

Existing6/70

Eliminate7/70-6/75

Add7/70-6/75

EquivalenceLoading*

EstimatedTotal

Actual

Equivalent

332

332

(-100)

(-100) ...-265

265--0

265*

497

762.

*This reflects the conservative estimate that applications added during the next five years will require on the average at leasttwice the developmental, operating, and maintenance cost of the existing applications.

3-16

1000

900

800

700

600

500

400

300

200

100

0

TOTALWORK

PROJECTEDLOAD

590

676

762

504512

418 404 440476

332 368

292 272232

../4312

252(AVERA .';E 30/YR)

1969 1970 1971 1972 1973

FISCAL YEARS

1974 1975 1976

Figure 2.2. Projected Workload State Government Computer Operations

2.4 and are accompanied by annual rates ofchange. The table also includes the annualpercentages of total State expendituresrepresented by information systems expenditures.Aside from noting the trend of such figures,caution should be exercised in drawingconclusions from them. The expenditures aretotal and include figures not relevant to theexpenses for State government computerization,such as Federal aid, State grants-in-aid toelementary and secondary education, and Stateappropriations to higher education. The final linein the table relates the annual rates of growth oftotal expenditures and information systemsexpenditures.

Comparison With Other States

1977

In the course of this study, it was not feasible toobtain total costs of information systems in otherstates for comparison with Minnesota. However,it was possible to obtain an estimate of computercosts for comparative purposes. Figures 2.3 and2.4 compare (1964-1969) the States of Illinois,Indiana, Iowa, Michigan, Minnesota, Ohio, andWisconsin on two points:

o Estimated value of computer hardware perstate resident.

Estimated value of computer hardware perstate employee.

3-17

Table 2.4. Total Expenditures and Annual Growth Rates for State of Minnesota Governmentand State Computer Operations (Dollars in Millions)

Fiscal Year 1966 L 1967 1968 1969 1970

Total State Expenditures $1,088 $1,168 $1,507 $1,727* $1,860*

Annual Rate of Increase in State 9.2% 7.4y. 29.0% 14.6% 7.7%Expenditures

State Computer Operations $2.0 $2.2 $2.9 $3.9 $5.1Expenditures

Annual Rate of Increase in Computer _ 11.0% 31.8% 34.5% 30.8%Operations Expenditures

State Computer Operations 0.184% 0.190% 0.195% 0.223% 0.274%Expenditures as a Percent of TotalState Expenditures

I

Af stimated

Significant observations from this table as relatedto Minnesota are:

In terms of computer hardware cost perresident in 1969, Minnesota was clusteredwith Illinois, Indiana, and Michigan in agroup far below Wisconsin and Iowa andsubstantially higher than Ohio. In 1967,Minnesota's computer hardware cost perresident was the lowest of the selectedstates, moving up sharply in 1968 and1969.

Minnesota ranks low in computer hardwarecosts per state employee. In 1967 and1968, Minnesota ranked the lowest amongthe selected states, but in 1969 movedahead of Ohio and equalled Michigan.

Information this sketchy cannot be used fordefinitive analysis. For example, there is notabulating equipment cost included. Further,qualitative questions are omitted such as the levelof multi-programming, cost effectiveness of thesystems, etc. It is clear, however, that Minnesotacomputer hardware expenses have been relativelylow and only ,.ecently have begun to moveupwards toward the average levels of the otherstates shown. Table 2.5 below relates to

3-18

Minnesota as of June 30, 1970 and shows theestimated annual computer hardware costs bothon a purchased and leased basis. The estimatedtotal information systems costs relative to stateresidents and state employees are also shown.

Potential Funding Levels

As discussed above, four factors were evaluated inthe development of projected budget levelsapplicable to State information systemsrequirements for the next five-year period. Theseare the number of major applications, anticipatedexpenditure trends in recent years, informationsystems expenditures as a percent of total Stateexpenditures, and historical computer hardwarecosts in Minnesota as compared to other states.Figure 2.5 and Table 2.6 show three of thesevarious alternatives projected in graphic andtabular form.

Curve A in Figure 2.5 reflects the "majorapplication" analysis. It is based on theassumption that the rate of implementation ofapplications over the five-year period will beconstant and, accordingly, that the number ofapplications implemented each year will increase.

'64I I

'65 '66 '67CALENDAR YEARS

'68 '69

Figure 2.3. Seven State Comparison EstimatedValue of Computer Hardware PerState Resident (Based on PurchasePrice)

The total costs of operating existing systems willincrease each year as will maintenance costs, atapproximately the same rate as the newapplications. Curve B is simply a projection onthe basis of 32 percent a year, the most recentthree-year average of annual rate of increase inState information systems. Curve C represents aprojection based on the historical relationshipbetween the trend in total State expenditures andthe trend in State expenditures for informationsystems, using the projected State budget figuresas the base line for the future. It assumes that thehistorical trends of the increasing percent of thebudget going to computerization will continueand that State expenditures will grow in thefuture at the average rate of the past five years.

Curve A appears to represent a reasonableapproach to funding for State informationsystems. It is based on a measure of workload.Admittedly, that measure is far from precise bothin terms of number of applications andcomplexity. But it is a measure and one whichresults in a fairly manageable growth rate. CurveB is based on a "catch-up" period in Statecomputerization and attempts no relationship toworkload or to overall State growth. Further, the

'65 '66 '67CALENDAR YEARS

'68 '69

Figure 2.4. Seven State Comparison EstimatedValue of Computer Hardware PerState Government Employee(Based on Purchase Price)

rate of increase may be difficult to manageeffectively. Curve C is extremely dependent uponthe level of the total State budget. It does notrelate directly to the workload in informationsystems which remains regardless of the level ofoverall State spending.

The projection indicated by Curve A is suggestedas a basis for future planning but not forlegislative appropriation. The deficiencies in thepresent cost data; the rapidly changing nature ofFederal, State, and local informationrequirements; developments in systems andcomputer technology and art; inflation;

3-19

Table 2.5. Minnesota Estimated Costs of Computer Hardware and Information SystemsPer State Resident and State Employee 1970

Estimated CostPer State Resident

INIIIINOEstimated Cost

Per State Employee

Computer Hardware: Purchased Basis $1.59 $240.00

Computer Hardware: Leased Basis 27% of Purchase $0.43 $ 64.80Total Information Systems Costs $1.35 $207.00

"CS

U.CaUJ

C.)

OUJ

zn.2OCZoz-w

w

500

3-20

25

20ACTUAL PROJECTED 20.430

CURVE C

CURVE

11.730B

9.700

CURVE A

12.000

5.105

6.740

8.890

6.300

7.820

1.9992.221

2.9343.856

15 -

10

5

1966 1967 1968 1969 1970 1971FISCAL YEARS

1972 1

Figure 2.5. Expenditures and Projected Funding Trends State Government Computer Operations

16.100

14.900

Table 2.6. Funding Projections for State Government Computer Operations (Dollars in Millions)

Fiscal Year 1971 1972 1973 1974 1975Projected State Budget $1,93V $2,161** $ 2,419** $ 2,708** $ 3,031**Annual Increase 3.9% 11.9% 11.9% 11.9% 11.9%Funding for the State Computer $6.300 $7.820 $9.700 $12.000 $14.900Operations Based on MajorApplications Workload

Annual Increase 23% 24% 24% 24% 23%

Funding for State Computer $6.740 $8.890 $11.730 $15.480 $20.430Operations Based on StateInformation Systems AnnualGrowth Trends

Projected Trend of 0.326% 0.37" 0.428% 0.479% 0.530%Ratio of ComputerExpenditures toTotal State Budget

Funding for State Computer $6.300 :8.150 $10.350 $12.290 $16.100Operations Based on Relationshipof State Information SystemsExpenditures to Total StateBudget

er

*Estimated 1971 Budget.**State expenditures are projected at an annual increase of 11.9%. Data available from the Statistical

Abstract of the United States for 1969 indicates the national annual increase in State Governmentexpenditures for the years 1965 through 1967 was 12.6%

availability of personnel competent to accomplishthe necessary systems analysis, programming, andoperational functions; and many other factorsmake a definitive recommendation for futureappropriations not feasible. It should also benoted that implementation of certain massiveprojects, such as a new MEDICARE system orFederally prescribed Manpower ServicesDepartment projects, may materially increasethese levels in the near future. Based on allexperience to date in public and privateenterprise, it is probable that the projection willbe conservative and will have failed to account forgrowth in computerization which has almostalways greatly exceeded expectations. Finally, it

is postulated that Curve A would notsubstantially alter Minnesota's position in thenext five years relative to the states referenced inthis section.

Projection of the level of funding beyond fiveyears is extremely hazardous. The only point ofcertainty is that computerization will becomemore inclusive year by year until alladministrative activity is computer-based. Whatthis means in terms of costs, however, is far toodependent on many of the uncertaintiesmentioned previously to permit projection on anyreasonably sound basis.

3-21

3. COST/BENEFIT CONSIDERATIONS

DOES COMPUTERIZATION REALLY PAY?

The introduction to this report includescomments about the economics of systemsanalysis and computerization. The rationalealways comes down to the need to providebenefits greater than the cost of attaining them. Itappears fashionable, however, in some circles todiscount the direct cost/benefit capability ofsystems analysis. "Computerization never savesmoney." "We never cut our payroll." "Everytimewe put something on the computer, it costsmore." These and similar remarks are frequentlyheard from computer personnel as well as users.Two simple answers to statements such as theseare: (1) applications should not be computerizedwhere cost effectiveness does not result, and (2) itis hard to believe that both private enterprise andgovernment are massively embracing disciplinesand techniques that cost more than their worth.More complete answers follow:

Generally, the costs of the existing systemare not known. Thus, it is often impossibleto determine the cost savings of the newsystem.

Often feasibility studies are not made todetermine expected net cost or savings.Even when studies are made, often there isno follow-up to assure attaining thebenefits.

Frequently, systems analysis producesadditional results of substantial value tomanagement, not previously available.

Increasing workloads are sometimesoverlooked in making cost comparisonsafter the fact.

In many cases, the effect of wage increasesand inflation on the previous system arenot properly considered in makingcomparisons.

It must be repeated here that neither the State'saccounting system nor its past approach tosystems have contributed to cost/benefit analysis.It is difficult, therefore, to demonstratecost/benefit on a general basis. Two examples ofdirect cost/benefit in Minnesota are presented, ananalogy is drawn from the experience of anotherstate, and some comments made about theindirect cost/benefits of selected present andplanned systems.

Income Tax Processing

The figures in Table 3.1, supplied by theMinnesota Department of Taxation, give somemeasure of impact of computerization on incometax processing.

Thus, during the six years shown, income taxreturns increased by 29 percent while the hoursworked by seasonal employees on those returnsdecreased by 7 percent. If the seasonal employeehours had increased proportionately to theincrease in returns, 46,700 additional hourswould have been required. Although precisefigures are not available in full time employment,it is conservatively estimated that no morepersonnel are now being used than theapproximately 90 assigned to these functions in1964. The 29 percent growth factor, however,could have been expected to add about 26 fulltime employees to the work force since 1964.Thus, the 1969 fiscal year direct personnel costavoidance in these activities is conservativelyestimated at the annual rate of about $230,000(46,700 seasonal employee hours at $1.90 and 26full time employees at $5,400 per year).

A more general set of income tax cost data ispresented in Table 3.2. The figures in the firsttwo lines, supplied by the Department ofTaxation, show the trend in total costs ofadministering all Minnesota income tax laws. The

Siszla 3-23

Table 3.1. Minnesota Income Tax Processing, Seasonal Hours vs. Returns Filed (Figures in Thousands)

Fiscal Year 1964 1965 1966 1967 1968 1969

Total Hours Worked by Seasonal Employees 130 125 127 135 129 121

Annual Rate of Change -3.8% +1.6% +6.3% -4.4% -6.2';',

Income Tax Returns Filed 1,370 1,426 1,501 1,586 1,659 1, 760

Annual Rate of Change +4.1% +5.3% +5.7% +4.6% +6.7%

Table 3.2. Estimated Cost of Administering Minnesota Income Tax Laws Fiscal Years 1964-1969

I 1964 1 1965 1 1966 I 1967 I 1968 I 1969

Costs per $100 of Gross Collections $1.34 $1.19 $0.93 $0.90 $1.01 $0.90

Cost per Return Filed 2.11 2.08 2.11 2.07 2.37 2.30

Cost per Return Filed Assuming No 1.98 1.99 1.77 2.03 1.83State Salary Plan Increase Since 1964

1

last line contains estimates of what the cost perreturn filed each year would have been if therehad been no salary plan increases from 1964through 1969. It is emphasized that the tablereflects the results of all administrative actiontaken over the years, not only computerization.

Three additional points must be noted. First, noconsideration is given in the above figures to thefact that processing requirements have becomemuch more complex and costly over the years asa result of changes in the tax laws and researchrequirements. Second, the data shown abovebegin with 1964. Actually, the income taxcomputerization represents the continuation ofan effort begun in the 1940's on tabulatingequipment. The impact measured since 1964 is,therefore, only a part of the story. Third, noeffort is made to evaluate the intangibles ofservice to citizens and information formanagement.

In summary, even on the basis of incompletedata, it is clear that direct cost/benefits have

3-24

resulted from the systems analysis andcomputerization related to income tax processing;further, indirect benefits have been substantial.

Engineering Tasks

In a carefully documented study dated October1966, the Minnesota Department of Highwaysshowed substantial direct cost/benefit. 2 Theannual pre-automation costs of performingcertain engineering tasks were $2,748,875. Aftercomputerization, the costs were $1,129,971. The$1,618,904 saving amounted to nearly 59 percentof the original cost. It can be presumed that theannual benefit has increased materially since thenbecause of (a) volume increases, (b) increasingsophistication of the computer applications in theanalysis, and (c) the rapidly rising costs of manualoperations and the relatively slower increase incomputer related costs.

Another State

Claims for benefits in computerization made inanother state although not directly applicable

I

I

I

I

I

1

I

to Minnesota may have some relevance toMinnesota. Through its program of automationup to 1968, Pennsylvania claims to have avoidedthe cost of 20,000 employees representing$120,000,000 annually. 3 Assuming the sameNavel of accomplishment in Minnesota, the benefitwould have approximated $42,000,000 annually.Even a quarter of that figure would be well overthree times the $2,934,000 spent by MinnesotaState agencies on information systems in 1968.

Indirect Cost/Benefit

There are many types of indirect cost/benefitsthat may result from well designed,computer-based information systems. Twogeneral types are mentioned. The MINCIS systemoffers indirect cost/benefits that are directlyrelated to operations. The benefits result from theincreased speed and accuracy of response toinquiries from law enforcement officials relatingto motor vehicle registration, driver license status,stolen property, and persons involved with thecriminal justice system. The benefits from thesystem, which is essentially nationwide, can beextremely significant.

As the MINCIS criminal history file is completed,it can serve another general type of indirectcost/benefit namely, for research, planning, andmanagement control. Two systems now in thedevelopment process provide further importantexamples of this type of indirect benefit. Theseare the new state accounting system and the stateland inventory system. These systems, it shouldbe noted, will be valuable to the Legislature aswell as to the Executive Branch of the State.

For the purpose of this report, no attempt ismade to put a dollar value on such benefits. It isprobable, however, that the indirect cost/benefitsof well-conceived State, information systems willexceed direct cost/benefits many times over.

BENEFITS THROUGH CENTRALIZATION

Another area for cost/benefit analysis is in theeffectiveness of the computer operation itself.

Minnesota has one of the most centralizedcomputer operations of any state in the country.This report recommends a continuation of thatmethod of operation and completion of theexisting centralization plan. A hypotheticalexample is presented here to illustrate theinherent cost advantages of the Minnesotaapproach.

Figure 4.1 and Table 4.1 in the Higher Educationpart of this study present relative costs andprocessing capacities of certain sizes or classes ofcomputers. The data reflect Grosch's Law whichstates that computer power generally doubles asthe square of the cost. Trebling the cost increasesraw computing power nine times, etc. Computingpower, however, is not directly translatable tooutput in an information systems environment.Output is governed much of the time by thespeed and availability of input and outputperipheral devices, by scheduling, and by theoperating systems. The cost advantage of thelarger computer is, therefore, difficult to specifyshort of an extremely detailed analysis.

Table 3.3 lists the processing speeds of small,medium, and large computer systems, theirmonthly leasing costs, and the relative costs pertransaction. The IBM 360 series is shown in theillustration because CSD uses that seriesextensively. Using Table 3.3 as a base, an exampleis constructed in Table 3.4 portraying whatcould have happened had the State permitted thedevelopment of a whole series of separatecomputer centers. The example assumes that sixindependently operated computer centers ofvarying capacities would exist in Minnesota.Workloads and the annual lease and other costsare shown for each of the six centers. These costsare compared with those of one large centralizedoperation having the theoretical capability ofperforming the same data processing workload.Costs other than equipment are shown as equal totwice the equipment costs. This reflects thegeneral rule at this point in time that such a costrelationship does in fact exist. None of theone-time start up costs for an installation is

shown although they could be substantial.

3-25

Table 3.3. Estimated Computing Power as a Function of Monthly System Leasing CostsIBM 360 Series

NominalTypicalMonthly

Capability(Number of

TransactionsRelativeCost Per

Speed AdvantageCompared with:

System Class Lease Cost Per Hour) Transaction* 360/30 360/40

360/30 Small $ 8,000 18,000 1.0 - -360/40 Medium 12 000 60,000 0.45 3.3/1 -360/50 Large 25,000 240,000 0.29 13/1 4/1

*Relative cost per transaction is not expressed in a unit of measure (such as dollars); it shows the greater efficiency of themedium and large system in relation to the small system which is used as a reference.

Table 3.4. Estimate of Equipment Rental and Other Costs for Six Separate Site Configurationsvs. A Centralized Configuration

Site

Trans.Load

Per HourSystems

Used

MonthlyLeaseCost

AnnualLeaseCost

Annual CostsOther ThanEquipment*

TotalAnnualCost

Total Cap.Trans.

Per Hour

#1 70,000 1 360/40 $ 12,000 S 144,000 $ 288,000 $ 432,000 60,0001 360/30 8,000 96,000 192,000 288,000 18,000

#2 150,000 1 360/50 25,000 300,000 600,000 900,000 240,000#3 25,000 1 360/40 12,000 144,000 288,000 432,000 60,000#4 10,000 1 360/30 8,000 96,000 192,000 288,000 18,000#5 175,000 1 360/50 25,000 300,000 600,000 900,000 240,000#6 170,000 1 360/50 25,000 300,000 600,000 900,000 240,000

Total 600,000 $115,000 $1,380,000 $2,760,000 $4,140,000 876,000

Centralize. 600,000 3 360/50's $ 75,000 $ 900,000 $1,800,000 $2,700,000 720,000Site Con-figuration

Annual Cost Benefit Difference $ 480,000 960,000 $1,440,000

*Includes other expenses such as supplies, power, etc.

The costs resulting from the analysis -$4,140,000 for the separate installations and$2,700,000 for the single installation result in a1.52 to 1 ratio. That is, decentralization wouldcost about 50 percent more than centralization. Ifthis relationship were directly transferable to thepresent situation in Minnesota, the $5,105,000expended on information systems would haveapproximated $7,760,000. Applying the factor to

3-26

the potential budget figures in the years1971-1975 (Table 2.6), a total difference of$26,374,000 would be realized.

These figures, however, must be tempered by twomain factors. The first is that Minnesota is notfully centralized. The Department of ManpowerServices has not yet been organizationallyintegrated and has a separate computer. Second,

1

the numerical relationships in Tables 3.3 and 3.4cannot be assumed in actual application.Nonetheless, realization of as much as a half, athird, or a fourth of the theoretical savings wouldrepresent a substantial total. It is important tonote that a number of states, including California,Florida, Pennsylvania, Washington, and others,have taken action to reduce the number of dataprocessing installations with expectation of verysubstantial savings.

The case for centralization cannot be made solelyon such savings. Centralization's inherentstrengths must be capitalized upon and itspotential administrative disadvantages guardedagainst if optimum results are to be obtained. Anumber of recommendations in Section 4 arerelated thereto.

3-27

I

r

IT

4. THE MINNESOTA SYSTEMS PLAN

Throughout this report, there have beenreferences to the rationale for the development ofsound systems to aid in ordering the processes ofbusiness, industry, and government. The natureand increasing importance of computerizedinformation and decision systems have also beendiscussed. It has been stressed that the crucialfactors in successful accomplishment oforganizational objectives are first, clear definitionof those objectives and second, sound systemsdesign.

The plan outlined in this section and designatedas MI DS (an acronym for Minnesota Informationatx.; 1.9cision Systems) aims to achieve two broadobjectives:

To make available to the Legislature and toofficials of the Executive Branchcomprehensive and accurate information ina form and at a time to be of maximumvalue in decision-making and research.

Provide to all State personnel the mosteffective framework of systems,procedures, and methods within which tocarry out the operations of the State.

The implications of implementing MIDS areenormous. In its broadest terms, MIDS meansdefining, designing, and implementingcomprehensive systems for each area ofgovernment to serve as the framework forattaining optimum effectiveness of stateoperations. A major requirement is the ability toassociate and interrelate the data bases of theinformation systems serving major functionalareas of government. Corollary to this is theinstallation of procedures to assure the integrityand confidentiality of data.

IMPLEMENTATION OF MIDS

Much that Minnesota has done in recent years isconsistent with the requirements of MIDS. Action

is necessary, however, in five broad areas toassure the administrative environment mostconducive to attaining the MI DS objectives:

i

A more comprehensive systems approachto the problems of the State.

An organization structure effective incoordinating the design of the many majorsystems, building efficient operationalsystems for aii areas of State government,attaining maximum productivity from theState's systems and computerizationefforts, and providing assistance to localgovernment.

The planned, controlled development ofsystems, procedures, and methodscomputer -based or otherwise to optimizethe value of data bases and informationobtainable therefrom for decision-makingand research, and to provide the mostefficient tools for State employees to carryout the activities of State government.

The development, installation, and policingof subsystems for assuring the integrity ofcomputer-based and all other data and toassure confidentiality of such data bylimiting its accessibility only to properlyauthorized persons.

A leadership position in the developmentand coordination of intergovernmentalinformation systems and ofintergovernmental cooperative efforts insystems analysis and computerization.

i

1

1

Specific recommendations for action in each ofthese areas follow.

A More Comprehensive Systems Approach to theProblems of State Government

The widespread failure to date of systems analysisand computerization to reach their potential, asdiscussed earlier in this report, resulted generallyfrom failure to take a comprehensive systems

3 ...a.ss 3-29

approach. The State of Minnesota was noexception. Correction of this condition requires,in Minnesota's case:

Assignment of overall responsibility fordesign, coordination, and implementationof MIDS to one agency of the government

see further discussion on organizationbelow. The Computer Services Division,under Executive Order No. 56, has beengiven broad responsibility for andauthority over the development ofinformation systems and other dataprocessing activities. This order should beinterpreted, or if necessary expanded, toinclude similar authority over all systemsanalysis and ultimately incorporated in thestate statutes by legislative action.

More effective execution by the topexecutive in each agency of Stategovernment of his responsibility forclarifying the objectives of each functionof his agency and for specifically approvingthe general systems required to reach thoseobjectives. The developmental efforts ofthe State PPBS (planning, programmingbudget system) are helping to clarifyobjectives in many areas, but much remainsto be done. In the main, however, topState executives have abdicated ordelegated to a relatively low level theirresponsibility for specifying requirementsfor systems to assist them in accomplishingtheir objectives. A condition precedent toall future systems studies should be thesignature of the top executive of the useragency on a statement of the objectives tobe accomplished. The same signatureshould be required as approval on thegeneral systems design prior to beginningthe implementation efforts.

Attendance by top management personnelof all agencies of State government atintensive courses in systems analysis,information and decision systems concepts,and computer familiarization. To be mosteffective, this program should be planned

and coordinated by the Division ofInformation Systems. Ideally, it shouldinclude a mix of professional societyworkshops, vendor orientation courses, and

3-30

state directed sessions, the latter aimed atfamiliarization with MIDS and with theState's present systems andcomputerization efforts and capabilities.Legislators and legislative staff membersshould be invited to participate.

An Effective Organization Structure

The relatively strong position Minnesota occupiesamong the states in the broad areas of systemsdevelopment and computerization has resultedfrom organizational decisions made over a decadeago and updated more recently. These decisionswere based initially on the need to control themechanical means of processing the State'sbusiness. Later, more attention was directedtoward the overall management implications ofcomputerization as evidenced in the 1967 and1969 special appropriations and in the 1968 and1970 Executive Orders.

Implementation of MI DS requires that thefollowing further organizational changes beaccomplished.

The State information systems functionshould be strengthened by improvedorganization and by :statutorycentralization in that organization ofauthority over development and operationof information systems andcomputerization.

Information systems are, and will continueto be, increasingly important to theoperation and management of the Stategovernment, a condition which must bereflected in State organization structure.The MIDS impact on information anddecision systems throughout the State, itsrequirements for inter-departmentalcooperation and coordination, its need forfully informed representation at top levelmeetings with the Governor, its continuallygrowing impact on governmentaloperations, its requirements for legislativeinvolvement, and its impact on the publicall add up to heavy involvement at the toplevel of government.

Organizational improvement can beaccomplished in one of two ways: byfurther strengthening the function withinthe Department of Administration or byestablishing a new Department ofInformation Systems.

Principal reasons for retaining the functionas a Division of Information Systems underthe Commissioner of Administration arethat (a) the information systems functionis the key to the executive management ofthe state and therefore key to theCommissioner who is in effect theExecutive Vice President of the State; (b)there is a close relationship betweeninformation systems and budgeting; (c)

there are already too many agencies ofgovernment reporting to the Governor; and(d) continuity of leadership would beobtained through an assistantcommissioner or director under CivilService.

Establishing a separate Department ofInformation Systems would bring theadvantage of having an official at the toplevel of State government developinginformation systems. He could concentratesolely on the extremely important task ofobtaining both direct and indirectcost/benefit from information systems andother computer systems. He would beappointed by and serve at the pleasure ofthe Governor.

For the organizational alternative selected,the most important requirement is that theappropriate responsibilities and authoritybe given to the function as outlined below.The acronym DIS in this report refers to a

Division of Information Systems, and itsadministrative officer is referred to asDirector of Information Systems.

All systems and computer functions of theState administration should beadministered or controlled by DIS. Thismeans highly centralized control and

coordination of systems development and

1

computerization and is consistent with thebasic concepts already being implementedin Minnesota. The organization structure ofDIS should build upon that of the presentComputer Services Division. It is

recommended that DIS be organized withthe following sections.

1. Systems and Programming. Thissection should continue its presentheavy responsibilities except thatsoftware research (not softwaredevelopment and implementation)should be assigned to the Planning andCoordination Section. Theresponsibilities of the Managers ofMajor Systems should be broadened toinclude specifically all systems analysisand design in their areas ofresponsibility.

2. Planning and Coordination. Thissection should continue its presentresponsibilities for planning,broadened as necessary to updateMIDS on a continuing basis; for highereducation coordination with the State;for staff development; and for datastandards. Its research function shouldbe expanded to cover all areas of DISresponsibilities including theevaluation of new technology inhardware and software.

A u nit should be established tocarry out certain broad systemsactivities such as forms control,records management, and workmeasurement, and to provide certainsystems specialization such as inmicrofilming.

In addition, a management sciencefunction (sometimes calledquantitative analysis or operationsresearch) should be added in thissection with responsibilities for (a)establishing and maintaining a

comprehensive index of an State datafiles, whether manual or in some waymechanized; (b) advising Statepersonnel on the best method ofobtaining, analyzing, and presentingdata from such files; (c) advising State

3-31

personnel on scientific samplingtechniques adequate to producestatistically sound analyses; (d)assisting with obtaining special reportsto meet management needs quickly;(e) advising and assisting on theapplication of operations researchtechniques to State problems.

3. Computer Operations. This sectionshould continue as presently organizedwith line responsibility for Statecomputer centers and a staffrelationship to those in highereducation.

4. Telecommunications Section. TheState communications system isincreasingly involved with systemsdevelopment and computerization.Accordingly, all communicationsfunctions of the State should beadministered under control of thissection,. The section head would chaira statewide CommunicationsCommittee recommended elsewhere.

5. Local Government Section. Thedemands for computerization at thelocal level are discussed in detail inPart V of this report. The State musttake a leadership role in assisting withcoordinated development at the locallevel. To give appropriate emphasisand status to this function and tobuild the framework for greaterassistance in the future, the local

government activities of DIS should beorganized as a separate section. As itsprimary responsibility, this sectionshould serve as the staff of theI ntergovernmental Commission onInformation Systems recommended inPart V of this report. Initially, thesection's role will be to advise localgovernment on systems andcomputerization and to begin effortstoward a common data base in localgovernment to serve local, State andFederal needs. Then the State shouldbegin to assist in development ofcommon systems and computerapproaches to the needs of localgovernment. If funding is provided as

3-32

recommended in this report, directassistance to local government willassume major significance.

6. Administrative Services Section. Thissection would perform for DIS thefunctions normally performed bysimilarly named units in the majordepartments of State government.Among those functions is that ofcontrollership. The need for a strongcontroller staff is acute and should beaccomplished immediately, prior tosetting up DIS. The internal budgeting,accounting, and cost accounting tasksin DIS are substantial. In particular,cost accounting is complex because ofthe intricacies of billing computer timeunder multi-programming. A furtherimportant controllership function isassuring the validity of the feasibilityanalyses required for proposed newsystems. Corollary to that is follow-upto determine the extent to which costbenefit projections are realized.

Expenditure Control. Legislation should bepassed to continue in DIS the authority inExecutive Order No. 56 to review andapprove all proposed expenditures by anyagency of the State government forsystems development or computerization(equipment, personnel, programs, space,systems design, consultants, servicebureaus, etc.) and extension of theauthority to cover all systems activity.

Governor's Committee. Legislative action isrecommended to formalize the presentGovernor's Committee on StateInformation Systems or a similar group toadvise the Governor, the Legislature, andDIS on MIDS plans and progress and torecommend on the expenditure ofdevelopmental funds. The Committeeshould be organized ) as to providecontinuity of approact. for example, sixyear terms for members, with one-thirdappointed each two years.

Agency Systems Coordination. InMinnesota, the systems coordinationfunction, when identifiable at all, isI

i

typically assigned to an employee at thefourth or fifth level of management. Thedevelopment and implementation ofmanagement systems is a complex matterrequiring the attention, coordination, andauthority of top management of the useragency. Accordingly, a systems coordinatorshould be appointed by each major agencyof State government, reporting directly tothe agency head or to his chief deputy orassistant. Heavy experience in systemsanalysis is the primary qualification forsuch a position with computer experiencehelpful.

Communications Advisory Committee. Anexplosion is occurring in the demand forelectronic communications of all types.Proliferation of systems, leading toduplication and overlapping, will resultunless coordination is effected. As a vehiclefor assuring the needed coordination, it isrecommended that a CommunicationsAdvisory Co mrnittee be established,chaired by the DIS TelecommunicationsSection head. Membership should includerepresentation from higher education andlocal government; major State users such asthe Department of Highways; and Statecomputer, radio, and telephonetechnicians. A major concern of thiscommittee will be data communications.

MIDS User's Committee. A committee ofDIS and user personnel is recommendedand is discussed immediately below.

Planned, Controlled Development of M IDS

I

Given the responsibilities and authorities and thetop level involvement outlined above, DIS shouldbe in a position to move the State toward theMI DS objectives. But it must be emphasized onceagain that proper systems design andimplementation and successful computeroperations are exceedingly complex and difficultand require a high degree of coordination andcooperation between DIS and the user agencies.

The major ingredients required are discussed here.It should be understood that the State has alreadytaken some action on most of these items, in

some cases with good effect and in other caseswith improvement still needed.

First and foremost is the need for DIS andits user agencies (almost the entire Stategovernment) to work together efficiently.It is axiomatic in this field that only trulycooperative efforts are successful. Threerecommendations have been madepreviously in this section to strengthen theusers' abilities to get what they need andwant. Specific additional action thatshould be taken includes:

1. Establishment of a MIDS User'sCommittee. A User's Committeeshould be established to serve as aprimary communication device in theimplementation of MIDS. Thecommittee should consist of theSystems Coordinator from each of themajor agencies of government, the DIS

Director, and the appropriate sectionheads of DIS. The DIS Director shouldserve as chairman the first year. TheCommittee should elect its ownchairman in the second and succeedingyears. Regular monthly meetings, witha structured agenda, should be held.

This advisory committee should makerecommendations to the DIS Directoron any matters relating to DIS and itsusers. It should be updated monthlyon the status of MIDS and on anyunusual developments in DIS.

2. Agency Systems and ComputerPlanning. Long range and short termplans for each agency are essential forboth the agency and DIS. The agencySystems Coordinator and the DISMajor Systems Manager for thatagency should jointly develop thenecessary plans. It is stronglyemphasized that the SystemsCoordinator, as agency representative,must assume the key role in planning.The DIS Major Systems Manager canmaterially assist the SystemsCoordinator; but the plan must reflectthe agency's real needs, not the DISview of those needs.

3-33

3. Agency Orientation in Training. Inaddition to the orientation to systemsand computers recommended earlierf or agency top management,formalized and continuing trainingprograms are necessary for agencypersonnel at the operating level. Aspectrum of subjects should becovered including systems; controls;systems conversion techniques; inputalternatives; computer techniques fordata storage, management, andretrieval; general computer capability;etc.

4. I nter-agency Reporting. Formalizedmethods of communication betweenDIS and its users should be establishedand rigorously administered. These

should include procedures forrequisitioning service, for advising onprogress, and for reporting onproblems. User management must bekept appraised of status so that timelyremedial action may be taken ifnecessary.

5. DIS Newsletter. A large number ofsignificant and exciting developmentshave occurred, are occurring, and willcontinue to occur in the State'ssystems and computer efforts. Manyof these would be of interest to useragency personnel. A simple newsletter,prepared monthly, could be helpful inthe continuing orientation of agencypersonnel to the accomplishments ofDIS users and to DIS capabilities.

The following guidelines should be adoptedfor MIDS:

1. A systems philosophy aimed atassisting the decision-making processas contrasted with simply mechanizingor otherwise improving a productionsystem.

2. A systems design to permit integratedsystems. This requires thedevelopment of standard identificationsystems for persons, organizations,physical assets, etc.; standard datadefinitions, and standard computertechniques for file management.

3-34

3. Cost/benefit analysis to establish

system feasibility.

4. Computerization developed in an

environment based on centralizedcentral processing units and whateverremote terminals are necessary to serveusers.

5. Operational objectives of on-line filesfor real-time retrieval of informationand on-line entry and update of dataoriginating where feasible in the useragency.

6. Attaining maximum output from bothequipment and personnel.

Planning and implementation based insofaras feasible on these guidelines are essentialto arriving at the MIDS long range goals.

The centralization of administration andcontrol of systems development andcomputerization should be completed.Justification for this recommendation isfound primarily in the cost/benefit analysisin Section 3 of this part of the study. Theimportant point to the user is that thenecessary systems assistance and computercapability must be available through DIS.

Alternatives to this approach include: (1)anumber of autonomous computer centers,(2) functional centers serving groups offunctionally related departments, and (3)regional centers serving groups ofdepartments located in the immediatevicinity of the center. Autonomousdepartment centers, however, tend to berestrictive to departmental needs, allowinglittle inter-agency data interchange andleading to duplication of file update,storage, and management. Functional andregional department oriented centers tendto create bias in favor of the prime user ofthe center, and permit interchange of dataonly among the users of each center.Furthermore, these alternatives present

other problems such as staff and fileduplication; inadequate overall dataexchange; more difficulty in setting data,control, or operating standards; equipmentinefficiency; and either a lack of, orextremely expensive, back-up facilities.

i

Specific action necessary to complete thecentralization includes:

1. Transfer of the computer functions ofthe Department of Manpower Servicesto DIS. Appropriate systems personnelshould remain assigned to MSD, aswith other major departments, toassure systems design responsive to theneeds of that department. The presentresistance by MSD to integration of itscomputer functions is typical amongsuch departments across the country.These departments, alone among allfunctions of State government, viewtheir operations as essentially Federalbecause of Federal financing and have atendency to remain apart from theregular processes of State government.The rationale, however, that applies tothe centralized computer approachpreviously adopted by Minnesota andrecommended for continuance hereinis equally applicable to MSD. It is

important to note that the U.S.Department of Labor's Bureau ofEmployment Security, the federalagency involved with MSD, issued onDecember 15, 1969, a policy oncomputerization which includes thefollowing:

"Maximum cooperation will be givento State governments in establishingcentralized ADP facilities which willprovide increased efficiency for Stategovernments and State employmentsecurity operations and which meetFederal and State manpower prioritiesand requirements."

2. Developing a schedule for phasing outtabulating machine installations infavor of computerization, includingprovision for remote terminals whereappropriate. Tabulating procedures arerelatively costly; careful systems andconversion planning must precedeconversion to the computer to assureacceptable agency service.

Systems activities must be economicallyjustified. Each systems design which is

recommended for implementationwhether to be on a computer or otherwise 1

must be analyzed, and the feasibility ofimplementation determined. If feasible,and if implemented, then a plannedfollow-up must be completed to assure thatthe "pay-off" was realized or to discoverwhy not. Unless the economics arecarefully identified in advance andpotential areas of increased costs anddecreased costs agreed to in some detail,potential savings may evaporate.

For example, presently three of the largercomputer-based systems of the state(driver license, income tax, and motorvehicle) should be analyzed to determinewhether cost/benefit projections have beenrealized; and, if not, to determine whetherthe projections still can be realized. Theuser agencies and the State Budget Divisionshould be involved in such analyses, alongwith the DIS Controller's staff.

Adequateprovidedallocated

physical facilities must beDIS. At present the spacefalls short of minimum

requirements. The two computerinstallations are on opposite sides of theCapitol Mall, preventing optimum use ofthe computer equipment. Air conditioningfacilities in the Centennial Buildinginstallation have frequently been unable tohandle the load and, occasionally, havecompletely failed. Both computer roomsare jammed so full of equipment as toimpair the efficiency of computeroperating personnel. Neither room is builtso that acceptable security standards canbe assured. There is in both areasinadequate fireproof space for storage ofmagnetic historical master and transactionfiles. Space assigned to the CentennialBuilding keypunch section is extremelycrowded and generally not conducive tomaximum productivity.

Space for computer professional personnelis at the crisis point. Systems analysis andprogramming require heavy concentrationand, accordingly, call for office space withsome degree of privacy. In the Centennialarea, the present space is totallyinadequate, with desk space assigned tofour persons in each of a number of smallcubicles. The general noise level, constant

3-35

3-36

interruptions, inadequate file and workingarea, and inadequate ventilation all militateagainst top performance.

Stop gap measures have been adopted overthe past three years but have beenoutdated even before completion. It is

essential that appropriate space beobtained as rapidly as possible for housingthe entire computer operation in one area.If this can only be done by acquisition orconstruction of new space, authorizationshould be given now to obtain interimspace to bring the two computer centerstogether. The space must be large enoughto accommodate equipment to serve theDepartment of Manpower Services.Professional personnel could be housed inanother area temporarily if necessaryduring the interim period.

Recent emphasis has been placed by CSDon the security of the computer facilities,and many improvements made. Currentevents give great urgency to the need tostrengthen security even further. (It hasbeen noted elsewhere in this report thatthe Department of Administration is

projecting approximately $3,000,000 as abuilding cost for 60,000 square feet ofspace for computer purposes.)

Personnel costs aro constantly growing as apercent of the overall costs of systemsdevelopment and computerization, withthe result that increasing emphasis must beplaced on proper management of the staff.Personnel utilization is now far moreimportant than computer utilization.Selection of personnel is at least as

important as selection of a computer.

The Computer Services Division, in itsrecent reorganization, has taken a numberof steps to strengthen its management ofpersonnel, including appointing a

Manager-Personnel Development withresponsibilities for training anddevelopment of CSD professionalpersonnel. Guidelines in personnelmanagement should include:

1. Establishment and enforcement ofsystems design and programmingstandards.

I

2. Administering training anddevelopment programs for personneland encouraging involvement inprofessional activities.

3. Setting up schedules for all systemsand programming projects andcontinually measuring progress againstthe schedules.

4. Conducting periodic performancereviews including outlining programsto strengthen individual weaknesses.

5. Periodic review of classification andpay plans to assure competitiveposture, with particular emphasiscurrently on the positions in computeroperations.

6. Continuing effort to identify andimplement successful systemsdeveloped elsewhere to make moreproductive use of personnel.

7. Use of consultants in systems analysisand programming (1) in preference toemploying a specialist for one-time orsporadic requirements, (2) to avoiddisrupting schedules on existingprojects when a crisis project appears,or (3) to serve upon occasion as ashort-run overload staff.

8. Comprehensive documentation ofsystems design and programming sothat knowledge does not leave with adeparting employee.

These guidelines in the main have beenfollowed by CSD. The quality of resultswill be in direct proportion to thesupervisory attention given to theirenforcement.

Computer capabilities of the State willrequire expansion as the applicationsincrease. Centralization provides theframework within which that expansioncan be done most efficiently.Recommendations related to equipmentrequirements are:

1. The "Model State Contract" forleasing or purchasing computerhardware should continue to be usedby the State. Modifications of terms ofthe contract, necessary in the past,

I

I

I

I

I

1

I

...

should not be necessary in the future.The National Association forInformation Systems (NASIS), anaffiliate of the Council of StateGovernments, is revising the contractwith assistance from the NationalAssociation of State PurchasingOfficials and the computer industry.The contract provides the best presentbasis for assuring performance byvendors.

2. The requirements for compatibilitymust be given appropriate weight inobtaining additional or replacementequipment. There must be thecapability of directcomputer-to-computercommunications where required.Direct programming compatibility (orin the alternative, highly efficienttranslators) must be available whennecessary to avoid enormousreprogramming cost.

3. Communications networks should becentered on one computer with strongcommunications handling capabilities(until its capacity is consumed) toavoid the expense of designing,programming, and maintainingcommunications systems on differenttypes of computers.

4. Appropriate back-up equipmentshould be available. This isparticularly important for on-linesystems where working files aretotally computerized. It is emphasizedthat total back-up is expensiveand that the extent of back-upnecessary must be determined bycareful cost/benefit analyses.

5. To attain the substantial savings

available through purchasing ratherthan leasing, computers and relatedequipment should be purchasedoutright or under lease purchase planswhen there is reasonable expectationthat equipment obsolescence will notoccur prior to the breakeven point.The State appears to be in a goodposition to take advantage of savingsthrough purchase because of the size

of its computer requirements andparticularly as it increases its assistancewith computerization of localgovernment. Equipment obsolete inone installation might adequately servea need in another installation. If thereis any question about the legality ofthe State's sub-leasing to localgovernment, enabling legislationshould be obtained.

6. Peripheral equipment from other thanthe vendor of the central processingunit should be obtained where costeffective. Substantial savings are

possible, but careful planning and

complete vendor understanding is

necessary to assure propermaintenance support.

7. Computer utilization records shouldbe kept on all equipment and shouldreflect each of the methods ofmeasuring utilization. Continuingeffort should be made to increase theeffectiveness of utilization. Note,however, that interest in attainingextremely high utilization rates mustnot be permitted to endangercustomer service or to affect adverselyeither the flexibility or supportafforded by the computer hardware tosystems analysts and programmers. Nospecific utilization rate isrecommended as a guideline. Rather,the State should plan its computerexpansion so as to maintain itsgenerally high levels of utilization ofrecent years.

8. Special emphasis should be placed onevaluation of methods and equipmentfor input preparation because of itslarge potential for cost reduction.

A number of further actions are necessaryin the implementation of MIDS. The majorpoints that have not been covered earlierare:

1. Data Dictionary and Standardizationof Data Elements. In the recentreorganization, CSD set up the newposition of Manager-Standards andQuality with the major objective ofestablishing a data dictionary and,

1

3-37

corollary thereto, standardizing thestructure of data elements andidentifiers. Because there is probablyno single function more important tothe long-range objectives of M I DS, thisprogram must be given top priorityand top level support. Furtherexplanation is appropriate here, usingas an example, an organization doingbusiness in Minnesota. A typicalbusiness is "doing business" with theState in many ways: for example,withholding State income taxes fromemployees and remitting the amountwithheld to the State; collecting salestaxes for the State and remitting them;obtaining economic data from one ormore State departments; paying aState corporate income tax; payingState unemployment compensationtaxes; reporting accidents to theWorkmen's CompensationCommission; registering motor vehicleswith the State; selling merchandise tothe State; and many more dependingupon the type of business. The abilityof the State to service all of theseactivities promptly and efficientlywith a minimum of duplication is

important not only to the State itselfbut also to the business. An integratedsystems approach to any business'sbusiness with the State should have apositive effect on the business climate.Among other things, such an approachrequires the use of a unique,identifying code for each businessentity.

2. Controls. No principle incomputerization is so basic as the needfor controls. GIGO (garbage ingarbage out) has scuttled morecomputer installations and damagedmore user relations than all otherproblems. It is essential that adequatecontrols be established; continuedfailure or unwillingness by any useragency to provide control informationor by DIS to establish and administeraccuracy controls is a matter forprompt discussion at theCommissioner level.

3-38

3. Standards for Systems Analysis andProgramming. A major determinant ofquality production in an informationsystems organization is thedevelopment and implementation ofstandards to be followed by systemsanalysis and programming personnel.These standards should cover allelements of work performed by suchemployees for example,documentation of existing systems,feasibility analyses, programmingconventions, and programdocumentation. Although the State iswell along with this program, it isemphasized that the enforcement ofthe standards must be a primaryfunction of management in theSystems and Programming Division.

4. Unified Operational Environment.Probably the outstanding technicalachievement of the State's computereffort was the unusual speed andsuccess in converting both centers tofull operating systems. This was anenormous task. Remaining, however,are a large number of secondgeneration programs. Such programswhich are handling important, sizablesystems should be third generationprograms. But straight conversion maynot be the answer in some cases.Rather, the system may be in need ofredesign for any one of a number ofreasons. In addition, some of theprograms are so infrequently used thatreprogramming would not be costeffective.

The unified operation also requiresadoption of a single set ofprogramming conventions, similarityof operating systems, andstandardization of teleprocessing andgraphics network and interface controlprograms. Unless this is done, thecost/benefit of centralization will fallfar below its potential.

5. Customer Billing. Two billing systemsare presently in use as a result of theunification of the separate operations.Development of a new billing system

should be given high priority as amajor aid to the DIS internalmanagement information system andas an aid in user relations. The newsystem must, of course, be consistentwith both the new State accountingsystem and the State program budgetrequirements.

Adequate financing is essential. With theexception of the 1967 and 1969 specialappropriations, computerization inMinnesota has been financed by directappropriation to the user agencies. Asservices were billed by CSD, the useragencies transferred funds to the CSDrevolving account. This procedure wasfollowed both for existing applications andthose under development.

The two special appropriations made sumsavailable to accelerate development of 'newsystems or otherwise advance the computereffort of the State. In 1969, however,unlike 1967, certain of the newapplications recommended fordevelopment were eliminated fromdepartmental budgets, with theunderstanding that funds from the specialappropriation could be used for thosepurposes.

It is recommended that future funding forall new systems development follow theprecedent set in the 1969 specialappropriation. That is, all funds forproposed new applications orresystematizing of major existingapplications should be included in a specialappropriation to DiS. Assuming thecontinuance of the Governor's Committeeon State Information Systems, thatcommittee should advise the DISCommissioner on the expenditure of thosefunds. It is also recommended that anamount in addition to that necessary tofund the known new applications beincluded in the special appropriation for"crisis" applications or applications wherethe time table should be accelerated. Apossible level of funding for MIDS isdiscussed in Section 2 of this report.

I n addition, two further fundingrecommendations, made elsewhere in the

report, are repeated here. The first is forproviding adequate space for DIS(approximately $3,000,000). The second isa special appropriation in the amount of$1,900,000 for the next biennium to assistwith local government systems andcomputerization efforts. The secondrecommendation, discussed in more detailin the local government part of the report,calls for expenditure authority in the DISCommissioner upon recommendation bythe Intergovernmental Commissionon Information Systems.

Finally, in anticipation of implementingthe earlier recommendation concerningpurchase rather than lease, the computerrevolving fund should be increased. Theoriginal working capital was set at$250,000. An additional $250,000although inadequate for outright purchaseof a sizable system, would give DIS theflexibility to purchase some equipment toeffect savings.

Performance evaluation is desirable for allfunctions, but it is mandatory for DISbecause of the growing impact it has onState government functions and decisions.The evaluation should be comprehensive,including all DIS activities and those userresponsibilities that affect DISperformance. This includes but is notlimited to a review and analysis ofbudgeted vs. actual cost by project; actualsystem vs. the originally planned system;controls over the systems input and outputand run-to-run controls to assure validityof the files; cost/benefit actually realizedvs. those projected in the feasibility report.An evaluation is already being performedto a limited extent by the State PublicExaminer's Office and should continue tobe the responsibility of that office. Thefunction requires a high level computerprofessional with both substantialmanagement experience and a, heavyaccounting background.

Integrity and Confidentiality of Data

The guidelines just discussed includedrecommendations for establishing andadministering proper controls. In this

3-39

recommendation, attention is focused on theneed for assuring integrity of data files bypreventing willful tampering with data and byproviding detailed procedures for the correctionof erroneous data or elimination of obsolete data.Further, and most important, is action necessaryto safeguard private rights by establishingprocedures to guarantee confidentiality of data. Itis emphasized that all data are included, not justcomputer-based data. The problems of integrityand confidentiality of data are as old asdemocracy. The advent of computerized files hasdramatized the problems; the computer hasbrought opportunities both for better control andfor abuse.

Accordingly, it is recommended that theLegislature enact legislation establishing State andlocal government policy on integrity andconfidentiality of all data and directing (a) theCommissioner of Information Systems, inconsultation with appropriate officials of theState government, to establish programs andprocedures to implement such policies and toinsure their enforcement at the State level and (b)the I ntergovernmental Commission onInformation Systems (see Part V of this report) todevelop the programs and procedures toimplement those policies in local government.The Commissioner of Information Systems andthe I ntergovernmental Commission onInformation Systems should give priority to thisresponsibility.

(Substantial research is being conducted by anumber of different groups on the developmentof legislation and operating guidelines on thesetopics. One of these efforts was in conjunctionwith Project SEARCH, a federal six state(including Minnesota) joint effort in the criminaljustice area. The results of the SEARCH researchon these matters was published in August 1970,and covers the topic in depth.)

Intergovernmental Coordination and Cooperation

In the area of information systems, DIS stands ina key position with respect to the Federalgovernment, higher education, local governments,

3-40

and vendors of systems and computer equipmentand services. In every way possible, DIS shouldencourage and participate in cooperative effortsamong and between these groups. Actions ofwhich the following are representative should betaken by DIS.

Encourage Federal agencies to fund thedevelopment and implementation of majorinformation systems for State and localgovernments as, for example, has beendone by the U.S. Department of Labor,Bureau of Employment Security, for thestate departments (Minnesota Departmentof Manpower Services).

Corollary to the above, encourage directFederal and state participation along withlocal government in the development andimplementation of local governmentinformation systems. Only in this way isthere assurance that the resultant systemswill have general applicability to localgovernment.

To the extent feasible, promote MinnesotaState or local government agencies' servingas prototypes in systems development assuggested above.

Encourage joint local governmentdevelopment of information systems,assisting where possible with Statepersonnel and funding.

Promote the utilization of existing,successful applications including individualprograms as well as broader groupings ofapplications such as those developed byT I ES; in the resulting implementationeffort, assist with . State personnel andfunding when possible.

Support with personnel and funding anyprograms of NASIS (The NationalAssociation for State Information Systems)to promote common systems,intergovernmental data interchanges, dataand operating standards, and any otherNASIS efforts furthering State and localinformation systems.

Coordinate and assist with thedevelopment of data bases of statewide orregional value, such as MAPS at the

TI

University of Minnesota and RAFTdeveloped by the Citizen's League and theUpper Midwest Research and DevelopmentCouncil.

It is noted that Part V of this report emphasizesstate-local cooperative and coordinative efforts.

1.

Part II of this report contains a summary of therecommendations in Parts III, IV, and V alongwith a discussion of the interrelationships amongthe three main parts.

REFERENCES

North Star Research and Development Institute, A Plan for an Information System forMinnesota State Planning, (A Report to the State Planning Agency, State of Minnesota)January 1967.

2. Department of Highways, State of Minnesota, Meeting the Electronic Data Processing Needsof the Minnesota Highway Department, October 1966.

3. Planning Research Corporation, PIMISS: Pennsylvania Interagency Management InformationSupport System, (Prepared for the Bureau of Management Information Systems,Commonwealth of Pennsylvania) September 1968.

3-41

I

I

i

I.0.0

7

PART IVHIGHER EDUCATION

PREFACE

The length and somewhat tedious detail of thisstudy of computers in Minnesota highereducation is at least as apparent to its authors asit will be to its readers. Section 1 of the reportargues the necessity of computers in highereducation today, in terms of the objectives of thefive systems of post-secondary education inMinnesota. The final section, Section 5, presents30 specific recommendations for providingMinnesota higher education with the computingcapacity it needs over the next five- and ten-yearperiods, complete with estimated budgetschedules and commentary on the relationship ofthe projected costs to national standards andother costs of higher education. By reading thesetwo sections, the reader who lacks time to digestthe entire report will be exposed to most of theimportant points and recommendations containedin it. In between, Section 2 covers the presentstatus of computer facilities and activities inMinnesota higher education; Section 3 contains aquantitative analysis of the needs for computingcapacity over the next ten years; and Section 4converts this needed capacity into generalcomputer hardware configurations andgeographical deployments, and estimates the totalcosts involved. Those readers who are unfamiliarwith computer systems using remote terminalsmay find it especially helpful to read Section 4.3,and perhaps Sections 4.1 and 4,2, to learn justwhat is involved in them.

Frequent reference is made in this part of thereport to ten Appendices, labeled H.1 through

H.10. These appendices contain information onthe conduct of the study, papers written onspecial topics of importance to computerapplications in higher education but which couldnot be covered adequately during the study, anddetails and tables of results from the analyses inSections 2, 3, and 4 of the main report. Theseappendices are not bound in copies of the generalreport which also addresses information systemsin state and local government. The interestedreader may obtain a copy of the higher educationreport bound together with its Appendices fromthe Higher Education Coordinating Commission.

Finally, we should like to emphasize that,although the recommendations and developmentschedules in Section 5 are in a sense a "MasterPlan" for computing in higher education inMinnesota, they cannot and must not beinterpreted as specific, rigid implementation plansor schedules, but rather as guidelines for thedevelopment of specific implementation plansand for the evaluation of those plans. It isproperly the function of the various systems andinstitutions of higher education to develop andpropose specific, detailed) implementation plans,and the function of the coordinatingorganizations and the State Legislature toevaluate and fund the plans. If this report isacceptable, it may provide a useful basis forplay, ping and evaluation, but the hard work ofdetailed planning and implementation must beleft to higher education itself.

4-i

ACKNOWLEDGMENTS

Whatwer acceptance this report receives will be aresult in part of numerous consultations withfamily members and administrators in manyinstitutions and all of the systems ofpost-secondary education in Minnesota, as well aswith several representatives of the computerindus':ry. Most of these individuals are listed inAppendix H.1. The authors are most grateful tothem for their generosity and hospitality inmeeting with us, and apologize for any nameswhich have been omitted inadvertently,misspelled, or incorrectly identified. ProfessorRusse I W. Burris and Dr. Frank Verbrugge of theUniversity of Minnesota, Dr. John E. Haugo ofthe State College Board, Richard Peterson of the

4-ii

State Department of Education, BanningHanscomb, Donald Wujcik and Dr. HowardBergstrom of the State Junior College Board, andR ichard Paulson of Macalester College wereespecially helpful to us. Periodic reviews andencouragement from the Computer AdvisoryCommittee of the Higher Education CoordinatingCommission were important in shaping thereport, as were information, advice, and technicalsupport received from the staff of theCoordinating Commission, and especially from itsExecutive Director, Richard C. Hawk.

Minneapolis, MinnesotaSeptember 1970

Peter G. RollPeter C. Patton

I4

L.

PART IV

HIGHER EDUCATION

TABLE OF CONTENTS

Description Page

1. INTRODUCTION 4-1

THE OBJECTIVES OF POST-SECONDARY EDUCATION 4-1

Instruction 4-2

Research 4-3

Public Service 4-3

Institutional Support of the Objectives of Higher Education 4-3

THE ROLE OF THE COMPUTER IN MEETING THE OBJECTIVES 4-4

OF HIGHER EDUCATION

Administrative Data Processing 4-5

Educational Computing 4-5Research Computing 4-6Public Service Computing 4-8

Special Applications of Computers in Higher Education 4-8

The Necessity for Computing Facilities and Services 4-9in Higher Education

OBJECTIVES OF THE STATEWIDE STUDY OF COMPUTING FACILITIES 4-10IN HIGHER EDUCATION

2. A SUMMARY OF PRESENT COMPUTING FACILITIES AND PROGRAMS

IN MINNESOTA HIGHER EDUCATION

THE UNIVERSITY OF MINNESOTATHE STATE COLLEGES

THE STATE JUNIOR COLLEGES

THE PRIVATE COLLEGES

THE AREA VOCATIONAL-TECHNICAL SCHOOLS

OVERALL SUMMARY OF COMPUTING CAPACITY

3. GOALS AND THE COMPUTER RESOURCES NEEDED TO MEET THEM

EDUCATIONAL COMPUTING

Instructional Computing

Instructional Computing for Graduate Students

Research Computing

Summary of Educational Computing Needs

ADMINISTRATIVE DATA PROCESSING

4-11

4-11

4-12

4-13

4-13

4-13

4-13

4-17

4-17

4-17

4-19

4-20

4-20

4-20

Description Page

SPECIAL APPLICATIONS

A Statewide Automated Library System

Computer-Assisted Instruction

Information Services

4.24

4.24

4-27

4-28

4. FACILITIES AND COSTS TO MEET THE GOALS 4-29COMPUTER SYSTEMS 4.29MASS STORAGE 4.39INPUT/OUTPUT FACILITIES: TERMINALS AND COMMUNICATIONS 4.40

Configuration of Remote-Terminal Computer Systems 4.40Terminals 4.40Multiplexors or Concentrators 4.40

Modems or Data Sets 4.42Communication Lines 4-42

Communications Controllers or Concentrators 4.43Staffing of Teleprocessing Networks 4.43

STAFF AND OPERATING SUPPORT 4-44DEPLOYMENT ALTERNATIVES 4-44

Procedures and Assumptions 4-44Specification of the Alternative Deployment Configurations 4-45

University of Minnesota 4-45State Colleges 4-46State Junior Colleges 4-47

Private Colleges 447Area Vocational-Technical Schools 4-47

A Fully Regionalized Deployment for Higher Education 4-48Summary of Results for the Deployment Alternatives 4-48

5. RECOMMENDATIONS AND DEVELOPMENT SCHEDULES 4-51

RECOMMENDATIONS FOR MINNESOTA HIGHER EDUCATION 4-51

Recommendations for Each of the Five Systems of Higher Education 4-53The University of Minnesota 4-53

The State College System 4-55The State Junior College System 4-55Area Vocational-Technical Schools 4-57Private Colleges 4-57

Recommendations Addressed to the Higher Education 4-58Coordinating Commission

A TIMEPHASED FIVE-YEAR PLAN 4-62A SECOND FIVE-YEAR PLAN 4-65THE PLANS IN PERSPECTIVE 4-72PRIORITIES 4-74REFERENCES 4-76APPENDICES 4-77

4-iv

I

I

LIST OF FIGURES

Figure Page

1.1 An Organizational Structure Supporting the Objectives of an Institutionof Higher Education (Suggested by Dr. R. W. Brady, Ohio State University)

4-4

2.1 Relative Spending for Computer Facilities, 1964-1969 Publicly Supported 4- 'i 6

Colleges and Universities

2.2 Relative Spending for Computer Facilities, 1964-1969 Privately 4-16Supported Colleges and Universities

4.1 Computer Capacity Versus Cost 4-37

4.2 Configuration of a Remote Terminal Computer System 4-41

5.1 Statewide Plan for Computers in Higher Education 4-63

5.2 Public Funding for Computers in Minnesota Higher Education 4-71

1

hi v

LIST OF TABLES

Table Page

2.1 Existing Computing Facilities Used for Instruction and Administration 4-15in Minnesota Post-Secondary Education, 1970-71 (includes thoseapproved for acquisition)

3.1 Summary of Terminal and Peak Load Computer Capacity Requirements 4-21for Instructional Computing in Minnesota Post-Secondary Institutions

3.2 Need Estimates for Administrative Data Processing Capacity in 4-25Minnesota Colleges

4.1 Estimated Computer Capacities Required to Serve Minnesota HigherEducation with Batch-Processing and Remote-Job-Entry Facilitiesin 1975-76

4.2 Estimated Computer Capacities Required to Serve Minnesota HigherEducation in 1975-76 with Time-Sharing and Remote-Job-Entry,Batch-Processing, Using the Same Computers for Both

4.3 Estimated Computer Capacities Required to Serve Minnesota HigherEducation in 1975-76 with Time-Sharing and Remote-Job-EntryBatch Processing, Using Separate Computers for the Two Modes

4.4 A Comparison of Three Ways of Supplying the Computer CapacityRequired at the University of Minnesota and at the Minnesota StateColleges. Computer Cost Estimates are Monthly Lease EquivalentsExtracted from Tables 4.1, 4.2, and 4.3

4.5 Summary of Monthly Equivalent Lease Costs Associated with Eachof the Deployment Configurations Analyzed (1970 Dollars)

4-31

4-33

4-35

4-39

4-49

5.1 Existing Computing Facilities in Public Higher Education by 4-66Function, 1970-71 (Including those Approved for Acquisition)

5.2 Proposed Funding Schedule for Computing Equipment 1971-75 4-67(Purchase Basis, 1970 Dollars)

5.3 Proposed Funding Schedule for Computing Equipment to be Added 4-68During 1971-75 (Lease 1asis; Annual Lease = 27% of Purchase Price,in 1970 Dollars)

5.4 Approximate Annual Incremental Operating Budget Needs (1970 Dollars) 4-69

5.5 Approximate Annual Total Operating Budget Neets (1970 Dollars) 4-70

5.6 Annual Cost Per Student of the Computing Capacity Proposed for 4-74Minnesota Higher Education in 1975, in 1970 Dollars (Derived fromTable 4.5)

4-vi

1. INTRODUCTION

During the past two decades, the development ofelectronic computing and data processingmachines has had a profound impact on allaspects of American society an impactunmatched by other technological developmentsin our history. Starting in the late 1940's as theesoteric research tool of a few scientists andmathematicians, computers have invadled almostevery part of our economy and society. Scienceand engineering were the first areas; then cameapplications to fiscal operations of business andother large organizations, and to the operations ofgovernment. Computers have become the tools ofthe artist and musician, the literary scholar, theprinter (automatic typesetting), the surveyor, thecarpenter (job estimating), and even the taxpayer,who today can avail himself of several serviceswhich make use of remote, time-shared computerterminals to assist in filling out IRS Form 1040.Rather than merely automating traditionalprocedures formerly carried out by large numbersof clerks or engineers, electronic computing anddata processing machinery has begun to changeprofoundly the systems and organizations whichuse them business, industry, government,hospitals, education. To quote from a report ofthe President's Science Advisory Committee in19671,

After growing wildly for years, the field ofcomputing now appears to be approaching itsinfancy.

A technology with this social impact must, andhas, invaded higher education. The computer hasevolved in higher education from an object andtool of research to both an object of instructionand a--tool of instruction. This evolution hasproceeded in a relatively uncoordinated way, inMinnesota as well as elsewhere in the nation. Theevolution has by now progressed to the point atwhich the computer is beginning to havesignificant effects on the methods, content, andorganization of higher education. It is the need

for coordination of these efforts, and the costs ofproviding the computer and data processingfacilities which they imply, which led to thepresent study.

This report will endeavor to provide answers tothe following questions:

Why are computers needed in highereducation?

How much computer capacity is needed toserve higher education in Minnesota?

How much will this computer capacitycost?

How can computer applications inMinnesota higher education be coordinatedand organized effectively to provide themaximum benefits to students and society,at the least cost to the taxpayer?

As the location of a significant fraction of theAmerican computer and computer-relatedelectronics industry, the State of Minnesota has a

special stake in the health of that industry, and inthe vigor of programs in higher education whichuse or relate to the products of the computerindustry. However, it will be the objectives ofhigher education in the various systems andinstitutions which form the basis for the analysisdirected at the above questions, and for therecommendations at the end of this report.

THE OBJECTIVES OF POST-SECONDARYEDUCATION

Higher education in the United Statestraditionally has had three objectives:

Instruction

Research

Public Service

4-1

Some kinds of institutions divide their effortsamong all three missions, and others specialize injust one or two of them. To understand the placeof computers in meeting these objectives today, itis important to have a clear understanding ofwhat the objectives really mean, and which of theobjectives apply to each of the five systems ofhigher education in Minnesota:

The University of Minnesota

The State Colleges

The State Junior Colleges

The Area Vocational-Technical Schools

The Private Colleges

Although referred to in this report as "systems",the last two listed are not coherent systems in thesame sense as the first three. The AreaVnrational-Technical Schools are creatures oflocal government Independent or IntermediateSchool Districts coordinated by the StateBoard of Vocational Education and the StateDepartment of Education. As such, they arerepresented indirectly in the other two majorparts of this study, covering state agencies andlocal government. Sixteen of the private collegesare members of the Minnesota Private CollegeCouncil, through which they are represented onthe Higher Education Coordinating Commission.They do not form a "system" in any but theloosest sense of that word.

Each of the three objectives of higher education isnow addressed in turn.

Instruction

Instruction of students is the primary reason forthe existence of post-secondary institutions inthis country, and for some of the institutions, it istheir only objective. But there are three kinds ofinstruction which are distributed among the fivesystems

General education of students tounderstand and appreciate themselves,other people, and the physical, social, andintellectual world in which they live. Thiskind of instruction is an importantfunction of the university, the state

4-2

colleges, the private colleges, and transferand some non-transfer programs in thejunior colleges. In the classical sense, thisobjective is the heart of a liberal artseducation. However, it does not play

a major role in the AreaVocational-Technical Schools.

Vocational training. This kind ofinstruction is the exclusive objective of theVocational Schools, the vocationalprograms in the junior colleges, and certainvocational programs in the colleges anduniversity. It is aimed at preparing studentsto enter a job and begin earning their paywithin a few days after they start, on thebasis of the specific training they havereceived. The jobs for which vocationaltraining is provided are well defined, and itis relatively easy to establish a successfultraining program and determine whether ornot a student has completed itsatisfactorily.

Professional education. The objective ofprofessional training is not to producegraduates who can step into a job andbegin producing immediately (althoughthis may happen), but to prepare a personwith a basic understanding of his field, themethods, techniques and tools used in it,and the discipline necessary to learn whatis needed to do a job and to direct andmanage others in getting it done. A personwith professional training at theundergraduate or graduate level may beunable to perform many tasks as well asthe graduate of a non-degree vocationalprogram. But it is from this group thatmost of the managers and executives come,and the people whose work leads to newkinds of jobs for others, and new ways ofdoing old jobs. Most students in a liberalarts curriculum today acquire in addition aconsiderable amount of training relevant tosome vocation or profession. Professionaland pre-professional training is theresponsibility of the state and privatecolleges and the university. Most of theadvanced graduate education, andprofessional training in areas such as law,engineering, medicine, and other health

sciences, is concentrated on the Universityof Minnesota's Twin Cities campus.

To maintain a viable system of post-secondaryeducation meeting the needs of the state fortrained manpower, an informed and articulatecitizenry, and a satisfying life, Minnesota mustsupport all three kinds of instruction. The resultsand payoffs of vocational and professionaleducation are easiest to recognize, but the generallevel of understanding and sensitivity of thepeople, enhanced more by general education, maybe as important to the welfare of our society inthe long run.

Research

Research may be defined as the expansion ofknowledge and the development of techniques,methods and devices for solving problems relatedto the natural world and to man and hisinstitutions. It is on research that the progress ofsociety depends. The principal resource of theState of Minnesota for research is the Universityof Minnesota. Within higher education, the statecolleges and some of the stronger private collegesalso have a mission which includes research,though to a lesser extent than the university. Thejunior colleges and area vocational-technicalschools do not serve as research institutions,except in the broad sense applicable to allinstitutions, whether they are in education oroutside, public or private: all institutions mustcarry out research on their own operations, todetermine whether they are meeting their goalsand how to increase their effectiveness andefficiency.

Much research is of value to the state and thenation in terms of its results and applications; thisis a public service of higher education. Butresearch is a necessity for institutions whosemission includes graduate and post-graduateprofessional education. The basic understanding,the ability to use the tools and methods, andmost especially the self-discipline required tolearn and to generate new knowledge when it isneeded these attributes are imparted to

graduate and post-graduate professional studentsthrough their participation in research. Whetherthese students remain in research when they leavethe university or pursue their field as practitionersin some other way, it is the research-generatedself-discipline and self-motivation for continuallearning which characterizes them asprofessionals, and in which the value of theireducation resides.

Public Service

In a broad sense, education and research are aservice to society. In addition, however,institutions of higher education in the UnitedStates have traditionally provided more directservices to the public as part of their mission. Thisbegan with accurate time-keeping service to thelocal community by the college astronomer(before the days of widespread telegraphiccommunications and railroad time zones), andagricultural extension services through theland-grant colleges. Today public services areprovided by colleges and universities not only toagriculture, but to businesses, schools and schooldistricts, the health care industry, governmentalagencies, the housewife, urban renewal projectsalmost all segments of society. These servicesmost often grow out of the other two functionsof higher education. Instruction has led toevening arid adult education classes, and largeExtension and Continuing Education Divisionswithin various institutions. And many of theresults and methods of research, particularly atthe University of Minnesota, have been developedby faculty members, working in cooperation withoutside groups, for application to some of theproblems of a segment of society or of the stateas a whole. If higher education were notperforming these services, other institutionswould have to be created to provide them.

Institutional Support of the Objectives of HigherEducation

Just as have most other institutions of oursociety, higher education has grown large,complicated, and expensive. The organizationalstructures required to support college and

4-3

university faculty in carrying out their threemissions have become more conspicuous, moreimportant, and also more costly. Two supportservices which directly affect and are necessaryfor the primary instructional function of highereducation are:

Student services financial aid, assistancein obtaining living accommodations andmeals, extra-curricular activities, etc.

Learning resources libraries, audio-visualmedia, television, laboratory equipment,and even the computer itself. All ofeducational technology and appliedlearning psychology would fall in thiscategory.

These two activities draw upon the results ofreievant research, and may carry out research anddevelopment work where it is needed to improvetheir support services. The results of this king ofdevelopment effort, as well as many of theresources used in higher education, are alsoavailable as a service to the public. The Universityof Minnesota Libraries, for instance, are a

resource available to the people of the State.

OBJECTIVES

SUPPORTACTIVITIES

4-4

INSTRUCTION

Supporting and managing the operations of eachinstitution and system of higher education is anadministration which, through its respectivegoverning board, is legally responsible to thepublic for the operation of the institution orsystem, and for the successful accomplishment ofits missions. Administration must carry out theresearch and develop the information and fiscalsystems necessary to effective and efficientmanagement of the educational enterprise.

Figure 1.1 may be useful in illustrating therelationships between the three objectives ofhigher education, which are always uppermost,and the supporting activities required to achievethem effectively and efficiently.

THE ROLE OF THE COMPUTER IN MEETINGTHE OBJECTIVES OF HIGHER EDUCATION

Having defined the goals of higher education, it isnecessary now to define the functional areaswhich need computing services or facilities, andthen to justify the needs in terms of the goals ofhigher education. For ease of discussion, severaldifferent functional areas have been defined andwill be discussed in turn.

RESEARCH PUBLIC SERVICE

i

LEARNINGRESOURCES

STUDENTSERVICES

ADMINISTRATION

FACULTY

ADMINISTRATIVESTAFF

Figure 1.1. An Organizational Structure Supporting the Objectives of an Institution of HigherEducation (Suggested by Dr. R. W. Brady, Ohio State University)

I

IAdministrative Data Processing

IStudent records, personnel files and accounts,financial records and transactions, property andspace inventories, and curriculum and schedulingrepresent most of the administrative and studentservice activities in a college or university. Theseare tasks which lend themselves to electronic dataprocessing, just as have the administrative dataprocessing requirements of business, governmentagencies, and other large organizations. There isno essential difference between administrativedata processing in institutions of higher educationand in businesses or other organizations ofcomparable size and complexity. For this reason,the benefits of electronic data processing inhigher education are similar to those which haveinduced many large businesses and industries toautomate their administrative activities; namely,increased effectiveness in the form of better andquicker service, and increased efficiency in termsof decreased costs or costs which do not rise asfast as they would otherwise. The real benefits tobe gained from administrative applications ofcomputers are the development of effectivecollege management information systems toimprove the quality of the administrative supportto students and faculty; and program planningand budgeting systems, so that it will be possibleto determine true costs of programs, and to assigninstitutional priorities on a more rational basisusing this information. Just as has been theexperience in business, applications of computerswill force a complete review and revision ofadministrative objectives, organizations andprocedures, providing the institutions themselves,their governing boards, the Legislature, and thepeople of the State with much more adequateinformation on which to base decisionsconcerning the wise allocation of the state'sresources.

_ Educational Computing

This is the computer application most closelyrelated to the primary mission of higher_....

education instruction. But within this category,

there are several kinds of computer applicationswhich are quite different, and require someelaboration.

Instructional Computing will be defined asthe use of generally-accessible computersby students in all courses requiringcomputing, and by graduate students fortheir thesis work; that is, computing whichis directly related to the student's learningactivities and which is normally carried outon large or small general-purposecomputing facilities. Except for courses inprogramming, the computer is used as atool for learning (or a "learning resource"),rather than an object of instruction.

The value of the computer as a

problem-solving tool is readily apparent forstudents in such quantitative areas as

engineering, science, accounting, andstatistical analysis in the social andbiological sciences. It may be less obvious,however, that computers have manyinstructional uses other than problemsolving in academic areas ranging fromengineering to the fine arts. Programs are inuse which permit students to interact witha computer simulation of a bridge orbuilding structure, an atom, a businessoperation, the political system of a cityduring an election, or a hospital patient.Using such simulation programs, studentscan acquire experience in making decisionsand investigating the consequences of thosedecisions. The computer is being usedsuccessfully to analyze the expositorywriting of students in journalism courses atthe University of Michigan. It is makingpossible student experience with using andinterpreting information from large socialdata bases, such as the census data. And ithas become an important medium forstu dents in art and music at several

col leges. These and many otherinstructional applications are described inthe Proceedings of a Conference onComputers in the Undergraduate Curricula,held in June 1970 at the University ofIowa2.

Computer Training includes courses andprograms in which the computer is an

4-5

object of instruction rather than a tool.For instance, students in data processingand programming courses at areavocational-technical schools need hands-onuse of either a computer or a

remote-job-entry terminal. This is theenvironment in which they will be requiredto function when they get a job. Manyvocational school graduates, in fact, begintheir careers as computer operators, forwhich training in operating a specificcomputer system is not so important, butgenerai familiarity with and experience inhandling some kind of computer orterminal equipment is most important. Forthose students who have regular accessonly to a terminal, it is important that,sometime during their training, they visit acomputer site for a few days of observationand hands-on experience in how acomputer system functions preferablyone that handles remote as well as localprocessing. (These same remarks apply tothe one junior college data processingprogram, at Lakewood State JuniorCollege.)

There are also programs for computermaintenance technicians in operation orbeginning at two of the area schoolsMankato and St. Cloud. Such programsrequire substantial access to computerhardware which can be taken apart andreassembled. Therefore, it simply cannotbe used to provide computer service. Thetwo schools with these programs have twoof the largest computers in the state a

UNIVAC Athena and an RCA 501,respectively. Fortunately, these have beenobtained as surplus at virtually no cost tothe Minnesota taxpayer, so that theseprograms can operate on a reasonable costbasis.

The programs of the Computer,Information, and Control ScienceDepartment at the University ofMinnesota, although aimed at a muchdifferent kind and level of student, are incertain respects similar to those in the areavocational schools. The graduate andperhaps advanced undergraduate trainingand research functions of this department,

4-6

now in its second year of operation,require access to software and hardwaresystems with which the students canexperiment. It is not feasible to carry outthis kind of training on a computer systemwhich must also provide service. Therefore,most major computer science departmentsthroughout the nation have acquiredspecial medium-sized or somewhatout-dated large computers for their ownuse (examples: Kansas University, Purdue,Case-Western Reserve). The University ofMinnesota will need to do this in the nearfuture.

Special-Purpose Laboratory Computing is aform of educational computing which isbecoming important and even necessary

in certain specialized professionalprograms, particularly in engineering. Forinstance, the use of process controlcomputers has become widespread in thechemical industry. As a result, manychemical engineering departments are usingsmall computers as a piece of instructionallaboratory equipment, to control unitoperations and provide real-time analysesof chemical engineering systems forstudents in their laboratory work. In orderfor engineering education to remain abreastof engineering practice, this kind ofeducational computing, in which thecomputer is again a tool rather than anobject of instruction, will become moreand more necessary. While this type ofeducational application is now mostapparent in a few engineering fields, it is

very likely to become a significant factor inother sciences, and especially in the healthsciences, over the next decade.

Research Computing

Computers were originally developed as researchtools, and over the past two decades they havebecome indispensible to research in many of thefields pursued in major universities. Researchcomputers can be grouped in three classes.

General-Purpose Service Facilities availableto the entire academic community forresearch and instruction. The major share

I

I

1

I

I

I

tII1

iII

1

of the research computing at colleges anduniversities is usually handled bycomputing facilities of this kind.

Special-Purpose Computing Facilities areconfigured for and dedicated to specificapplication areas. Although they may beused for both instruction and research, andgenerally by students and faculty in certainacademic areas, they usually are not suitedto or readily available for use outside ofthese areas. Two examples of facilities ofthis kind at the University of Minnesota arethe CDC 3300 in the Division of HealthComputer Sciences, dedicated to andfunded largely by the Federal Governmentfor applications in medical science; and theHybrid and Digigraphics ComputerLaboratory, specializing in hybridanalog-digital computation and high-speedinteractive computer graphics. Both ofthese facilities are used and are availablefor instruction and public service, althoughthe major fraction of their use is forsponsored research. Correspondingly, theyare funded largely through federal grantsand income from services rendered tosponsored research projects (also mostlyfederal grants) and outside agencies. (SeeAppendix H.4 for a breakdown of costsand funding sources. The Division ofHealth Computer Sciences facility, forinstance, was acquired with only 28%university or state funds, while aboutone-third of its computing load is directlyrelated to instruction of students in thehealth sciences.)

Special-Purpose Computing Devices builtinto research apparatus as part of theapparatus. Such computers are performingfunctions which cannot be accomplishedeffectively or efficiently, if at all, by ageneral-purpose computer. Functionally,they are serving not as computers, but asan integral part of the apparatus in whichthey are used. A useful analog is a

telephone exchange: a modern telephoneexchange could be called a computer theequipment can and does do most of thethings a computer does. Functionally,however, it is not a computer, but rather amessage-switching device. Because of this

functional distinction, and because theyare funded almost exclusively by researchgrants and contracts or service fees, thesespecial-purpose devices will not beconsidered further as computers.

These are the three categories of researchcomputing facilities. As pointed out earlier, thefunctions of research and instruction are closelylinked for graduate and post-graduate professionaleducation. In this sense, research grants andcontracts have contributed an important andsubstantial subsidy to this kind of education. It isimportant, therefore, to distinguish as

instructional computing the unsponsoredcomputing by graduate and post-graduateprofessional students related to their course workand thesis projects. This is computing for whichthe State must pay if it wishes to provide suchadvanced training. Sponsored research and otherprojects directly under the control of facultymembers and related to professional activities intheir academic field will be defined as researchcomputing, even if some graduate students maybe involved. This is not the best way to make thedistinction, but it is a practical way.

The computer demands of research are that thegeneral-purpose facilities and services be availablewhen they are needed, and at a level sufficient tomeet the demands; and that the institutions beorganized to allow and assist faculty members toobtain and use appropriate special-purposecomputing devices and facilities when outsidefunds are available for this purpose. Suchcomputer needs should not and have not madedemands upon institutional or state funds in anyway different from the many other kinds ofresearch activity which are supported from withinhigher education. For these reasons, this reportwill not undertake a detailed analysis of researchcomputing in higher education: the servicecapacity must be there as excess aboveinstructional needs. A major investment of statefunds is not required, because most of thiscomputing is supported from outside grants andcontracts.

47

Public Service Computing

Much of the activity normally labeled "research"in hither education actually is a public service,involving development and research activities forpublic agencies. There are, for instance, manygovernmental agencies using the variouscomputing facilities of the University ofMinnesota, both in the Twin Cities and Duluth.State college regional centers may be expected toundertake many similar activities. Two examplesexist 'or which a facility is almost exclusivelydedica:ed to a public service application.

The BM 360/30 computer located at theAgricu tural Experiment Station in St. Paul isdevoted to and funded by services to the DairyHerd Improvement Association, and for soilanalyses, farm management, and otheragriculiure-related activities. These applicationsare self-supporting. At the University Hospitals, aBurroughs 2500 computer is dedicated to hospitalad rninistrative data processing, patientmonito-ing, and other internal hospital functions.Again this facility is self-supporting as part of thehospita operations.

Since public service computing within highereducation is almost completely supportedthrough the services provided and does notrequire a direct outlay of state funding throughthe institutions of higher education, its needs willnot be analyzed in this report. The computercapacity and willingness to provide service of adevelopmental nature must be there, but therequired funding will come from the agenciesserved.

During the course of this study, it becameapparent that there were three identifiable areasin which significant planning and developmentwas underway for special applications ofcomputers to the goals of higher education.Development effort devoted to each of the three

4-8

areas has the potential of leading to moreeffective service for instruction, research, andpublic service, and for making the operations ofhigher education more efficient in providing theseservices that is, reducing costs or preventingcosts from increasing as rapidly as otherwise. Thethree special areas are discussed in more detail inSection 3, and two of them are dealt with inspecial reports appearing as Appendices H.2 andH.3. They are:

A Statewide Automated Library System.

The library is a major and expensivelearning resource at every college anduniversity, directly serving the needs ofinstruction, research, and public service.With the recent explosion in volume ofprinted information, libraries (andespecially academic libraries) are beingsorely taxed just to keep up with thevolume of ordering, processing,cataloging, and shelving, to say nothingof providing adequate reference andretrieval services. To be optimallyefficient, however, an automated librarymust serve the needs of all libraries inthe state, not just the libraries in highereducation.

Computer-Assisted Instruction andComputer-Managed Instruction.

These are direct applications of thecomputer to the instructional processusing the computer as an integral part ofthe process to administer and manageinstruction. To do this successfully willrequire a careful assessment of thespecific goals, methods, and content ofinstruction in each course and academicfield, followed by an equally carefulinstructional design to utilize computersand other media in an optimum way.This will require the greatestdevelopment effort and will encounterthe greatest resistance from thoseconditioned to traditional methods ofinstruction. But if successful, these

applications also will bring the greatestpayoff in terms of effectiveness andefficiency.

Information Services.

Information on populationcharacteristics, natural resources, andsimilar distributed properties of thenatural and social worlds is becomingincreasingly important for purposes ofpublic and governmental planning at alllevels. Likewise, information of thiskind, and the ability to process it, is animportant learning resource for researchand instruction, especially in the socialsciences. These latter needs have led toan increasing involvement of facultymembers with the development of largedata bases of information which can beread and processed by computers. Thepotential value of information services ofthis kind, both to higher education anddirectly to the public, cannot be ignoredby a Statewide Study of ComputingFacilities in Higher Education, and is, infact, linked closely to the other twoparts of this overall report, coveringState Government Agencies and LocalGovernment.

Each of these three special application areas willbe discussed in more detail in Section 3 of thisreport. Recommendations concerning themappear in Section 5.

The Necessity for Computing Facilities andServices in Higher Education

Having discussed the objectives of highereducation and the functions that computers canserve in meeting these objectives, it is time tosummarize three of the important reasons whycomputers are a necessity in higher educationtoday.

First of all, there is a need for people trained andeducated in computer programming and thevarious other aspects of computer science anddata processing. With the size of the computerindustry today, this need is obvious, but the sizeof the need is not. To quote from the 1967 reportof the President's Science AdvisaryCommittee, Computers in HigherEducation.1

There is inadequate information about thenumber and level of skills of personnel nowemployed in the field of computers, and thereare no meaningful forecasts.

We recommend that the Federal Governmentcollect meaningful data concerning computersand the jobs, personnel, and educationalfacilities associated with them, and endeavor tomake useful annual forecasts.

This recommendation has not been followed.Two articles in a recent issue of the trademagazine Datamation quoted rather disparateestimates of the national need for computerpersonnel:

50,000 per year 3, and 125,000 per year 4 .

Even considering the present recession in thecomputer industry, 2% of the lower estimate, or1,000 per year at all levels, is a reasonableconservative estimate of Minnesota's needs.Existing and planned programs will not saturatethis need for some time to come.

Secondly, we need people educated to understandcomputers, what they can do and how to usethem, and how to manage and direct their use inparticular areas of application. More specifically,it is doubtful whether society can effectively use,or even tolerate, more social scientists,educational and business administrators, andworkers in any number of other fields who arenot trained in quantitative methods. Computersin higher education make it possible to trainstudents in all fields in the use of quantitativemethods to attack and solve problems.

And finally, there is a need for a generalunderstanding of what computers can and cannotdo, and how they affect the society in which welive. It must be understood that an incorrectbilling, a lost order, or an inaccurate income taxrefund cannot be blamed on computers, butusually are the results of human error. Only whena broad level of public understanding is achievedcan the computer begin to reach its full potential.Higher education can help achieve this level ofunderstanding directly by including some

4-9

exposure to computing in the general educationof most students, and indirectly by adequatelyeducating those who will become school teachersand parents and affect the minds of futuregenerations.

OBJECTIVES OF THE STATEWIDE STUDY OFCOMPUTING FACILITIES IN HIGHEREDUCATION

Having argued the objectives of and needs forcomputing in higher education, it is the intent ofthe remainder of this report to:

Summarize the existing (1970-71)computing facilities and applications inMinnesota higher education and their costs(Section 2);

Analyze the needs of Minnesota highereducation for computing facilities andservices over the next five and ten years(Section 3);

Determine the costs of various ways ofmeeting these needs, identifying the leastexpensive ways (Section 4);

Recommend specific actions by theinstitutions and systems of highereducation and the legislature to meet theidentified needs and lay a sound basis forfuture development (Section 5).

The entire thrust of this study has been todetermine the computing facility and serviceneeds of the post-secondary students ofMinnesota, in terms of the objectives of highereducation, and to find and suggest ways ofsatisfying these needs which have the greatestlikelihood of being implemented. To this end,educators and administrators from numerousinstitutions in all five systems of higher education

and throughout the state were consultedextensively before the recommendations wereformulated. The extent of these consultations canbe judged from Appendix H.1. There is muchevidence, in the form of similar studies in otherstates, to suggest that plans and recommendationsdeveloped without consulting extensively thosepeople and institutions which must eventuallyimplement them are unlikely to be accepted andimplemented, regardless of their technical merits.In addition to this consultation within the state,the study benefited from the broader perspectiveof two consultants from other parts of thecountry. Dr. Ronald W. Brady, the ExecutiveAssistant to the President of the Ohio StateUniversity (and now AdministrativeVice-Chancellor of Syracuse University) offeredadvice on most of the matters related to thestudy. Dr. Robert M. Hayes, Director of theInstitute for Library Research at the University ofCalifornia and Vice-President of Becker andHayes, Inc., was consulted on technical andorganizational facets of plans for a statewidelibrary automation system.

This study presents some estimates for neededcomputing facilities and services in isolation fromthe other important requirements of highereducation. The final determination of the weightof the arguments and recommendations here, andthe priority of computing among all of the needsof higher education, resides properly with thefaculties, institutions, and systems of highereducation, the Minnesota Higher EducationCoordinating Commission, and ultimately thepeople of Minnesota through their electedrepresentatives in the State Legislature. If it is

generally acceptable, this report can provideguidelines for planning and evaluation by thesegroups, but it cannot be used as a plan. Specific,detailed planning and implementation is theresponsibility of higher education, and this reportrepresents only the beginning of that job.

I

I

I

I

I

I

I

I

I

1

1

.1.

2. A SUMMARY OF PRESENT COMPUTING FACILITIES ANDPROGRAMS IN MINNESOTA HIGHER EDUCATION

The 1967 report of the President's ScienceAdvisory Committee the Pierce Reportlindicated the need for improvement in computingfacilities and programs in higher education, andframed ten recommendations designed to bringthe facilities of all higher educational institutionsin the United States up to the standards of thoseschools which were best equipped at that time.These recommendations largely went unheeded,since the recommended federal funds of up to$400 million per year were not forthcoming. !nseveral states Pierce Report goals were met to agreater degree than has been the case nationwide.In such cases, however, funds have been suppliedby the respective legislatures of those states andhave not come mainly from federal agencies.Computing in higher education is currently in atransition phase in this and many other ways:from federal funding toward state funding, fromprimarily research use toward instruction, andfrom use by the few toward use by the many.

The situation in Minnesota is not atypical. Mostof the institutions of higher education in the stateare not nearly so well equipped as the Universityor Carleton College. Yet even the University fallsfar short of the Pierce Report objectives and lagsbehind similar institutions in, say, Ohio orWisconsin. On the other hand, the need forcomputer education and training may be greaterin Minnesota than many other states, since it is acenter for the design and manufacture ofcomputers, and has a larger number oftechnology-based industries than many states.

As pointed out in Section 1, the requirements forcomputer education and training, and thecorresponding programs and facilities at thepost-secondary level, vary widely depending onthe institution, its goals, and the career intentionsof its students. The ensuing discussion willaddress the five areas or systems of higher

education in Minnesota and will attempt tosummarize their facilities and programs and toevaluate them with respect to their mission incomputer education or training. Computers foradministrative data processing will also besummarized.

THE UNIVERSITY OF MINNESOTA

The University is the largest institution of highereducation in the state and is the best equippedwith computer facilities, These computer facilitiesare summarized in Table H.5.1 of Appendix H.5,and their uses are summarized in greater detail inAppendix H.4. Three major general-purposefacilities serve the entire university for bothinstructional and research purposes. These are theCDC 6600 at the University Computer Center andthe CDC 3200's at the West Bank ComputerCenter and the Duluth Campus. The overallinstructional use of these facilities amoums toabout 45% of their capability. However, theirinstructional use is growing much more rapidlythan their research use (See Appendix H.4). Tomeet their needs for limited amounts oftime-shared computing, which the University hasno facilities to provide, several departments andresearch groups are spending their own supplyfunds to purchase service from commercialvendors.

Special-purpose computing facilities serveinstructional, research, and public service needsbut are more specialized in function or servicearea than the general-purpose facilities. TheUniversity's special-purpose facilities include theCDC 3300 of the Division of Health ComputerSciences, the IBM 360/30 on the St. Paul campus,used primarily for agriculture-related publicservice activities, and the Hybrid ComputerLaboratory, with a CDC 1700 serving both a dualanalog-digital hybrid computer system and an

4-11

interactive graphic display system. AdministrativeData Processing at the University is served by anIBM 360/50 configuration which, if augmentedwith additional disc storage, would have sufficientcapacity to handle the batch and on-lineprocessing requirements for an institution of thissize.

1 n general, the computer facilities of theUniversity are adequate for the programsunderway or planned in the near future, with afew exceptions. The University has one of twoundergraduate computer science degree programsin the state and the only graduate program.Especially for the graduate program, a computeris essential for student laboratory work andexperimentation. The Center for Research inHuman Learning has done considerable researchand development in computer-assisted instructionbut needs a computer system able to support thecontinued development and early productionphases of this work. The IBM1500 instructionalcomputer system, used for the early phases oftheir work and funded largely by the NationalScience Foundation, was discontinued by IBMand the NSF in early June 1970. Finally, the CDC6600 has considerable unrealized potential forwider service which could be realized byproviding central-site hardware and software tosupport remote terminals, both within theUniversity and at other institutions. Also, it couldbe further enhanced as a job-shop computer byimproving printer facilities. Most large-scaleresearch computers used in an instructionalenvironment are badly printer-bound, and the6600 is no exception.

THE STATE COLLEGES

The computing capability in the State Collegeshas evolved over the past ten years from severalelectronic accounting machine installations.Current facilities include IBM 1401 installationsfor administrative data processing, IBM 1620centers for instruction, and some IBM 1130computers for both, as shown in Table H.5.2(Appendix H.5). Most of the existing applications

4 -12

are administrative in nature, leaning heavilytoward student record-keeping, with somebusiness or fiscal applications and a fewinstitutional research programs. Approximately1500 students within the state college system, orslightly less than 5% of total enrollment, areregistered each quarter in classes which involvesome sort of computer use. Only one of thecolleges offers a major in Computer Science; butthe others offer mathematics, physics, business,economics, or statistics courses which involvecomputer solution of problems.

As will be apparent from the analysis in Sections3 and 4 of this report, the existing (September1970) computing facilities within the state collegesystem are inadequate by the norms of the PierceReporti . However, this situation will be correctedby the major facility to be installed at Mankatolate in 1970. If a similar machine is installed in St.Cloud in two or three years, and if the other statecolleges have terminals to one of these machinesin addition to their current capability, thenfacilities will be adequate for well over fiveyears5. Demand for computer services at the statecolleges has not been severe primarily because theavailable computer resources were so meager as todiscourage serious use. The "double-hub"network planned for 1975 or sooner will be amajor step toward correcting this situation andwill encourage the development of computer andcomputer-related courses, as well as support theapplication of information systems technology tothe management of the State Colleges.

It may be difficult to justify more than oneadditional Computer Science degree program inthe state college system, and logically that wouldbe located at the college with the second hub. Itwould be feasible, however, for the other collegesto develop major or minor degree programs ininformation sciences, management informationtechnology, o r elsewhere in the interdisciplinaryareas between computer science, theadministrative disciplines, and the basic sciencesand mathematics. (Moorhead State College, forinstance, has a Computer Science optionassociated with its Mathematics program.)

THE STATE JUNIOR COLLEGES

The State Junior Colleges have a centralizedadministrative data processing system based on alarge IBM 1401 computer. Although this systemgives good service, it does have limitations forplanned increased enrollment and service levels.Until recently, the junior colleges had nocomputer service for instruction. Starting in theSpring of 1970, there have been teletypeterminals at each college connected to theHoneywell EDINET time-sharing system, as

shown in Table H.5.3. The 18 junior collegesshare three computer ports. These terminals willsupport instructional programs and courses forwhich the computer is used as a tool. The onlydata processing curriculum in the system is atLakewood Junior College, the site of theadministrative data processing center.

THE PRIVATE COLLEGES

The range of computing facilities andinstructional programs in the private colleges runsfrom more than adequate to non-existent. A fewschools are able to expose a hundred to threehundred students to the computer in seven to tenor more courses in Computer Science orcomputer-related disciplines, as indicated in TableH.5.4.

Although fewer than half of the private collegesand junior colleges employ computers forinstruction, most of them use computers, servicebureaus, or electronic accounting machines foradministrative data processing. Development costsappear high for a college to put administrativeapplications on a small computer. A group ofseven Twin Cities colleges has organized a

common program development and coordinationeffort to reduce these costs6.

Most of the gaps in computing for private collegesin Minnesota could be closed by coordination ofefforts to purchase or develop programs forcollege administration and by support forterminal facilities at each college and for central

or regional computers to service the terminals.Some of the colleges are doing an adequate jobwith a small IBM 1130 computer and do notreally need such terminal support. The University6600 is available to private college faculty fortheir research needs on the same basis as it is tothe University faculty and is now in the processof being made available via remote terminals atcolleges. The Mankato Regional Computer Centerwill provide similar service.

THE AREA VOCATIONAL-TECHNICALSCHOOLS

Several of the Area Vocational-Technical Schoolshave training programs for computerprogrammers (see Table H.5.5). However, somedegree of computer service is also needed at thoseschools which have one or two-year accountingprograms but do not offer programs in dataprocessing. As pointed out in Section 1, dataprocessing curricula in the AVTS have a

somewhat different mission than do similarprograms in colleges and the University.

There are no obvious gaps in the facilities orprograms at the schools with curricula in dataprocessing and with computers (Hibbing,Alexandria, and Mankato). They have excellentfacilities and are graduating students wellequipped to enter the business world as juniorprogrammers. Those schools with data processingprograms using IBM 1401 computers owned bytheir school district, or computer time donated orotherwise made available by local businesses, aswell as schools with one and two-year accountingprograms but without computer services adequatefor at least COBOL instruction, would be wellserved with terminals connected to regionalcomputers, service bureaus, or those AVTS withdata processing curricula and computers.

OVERALL SUMMARY OF COMPUTINGCAPACITY

There are two institutions offering degrees inComputer Science in Minnesota: Mankato State

4-13

College offers a B.S. degree, and the Universityoffers B.S., M.S., and Ph.D. degrees in Computer,Information, and Control Sciences. An additionalB.S. degree program at St. Cloud State Collegewithin the next five years would be tenable interms of supporting facilities if the secondregional center were to be located there. The dataprocessing programs in the AreaVocational-Technical Schools provide instructionin computer technology as it is applied to dataprocessing for business and industry. Theseprograms conceivably might be augmented by oneor two more such programs in the new vocationalschools soon to open in the Twin Cities area.

On the administrative data processing side ofhigher education, there are many facilities andmuch activity but also much redundancy.National efforts are underway to allowcomparability of data and use of commonprogram packages. Information systemstechnology is just now finding its way into collegeand university management and holds promise forrationalizing the current undisciplined growth ofapplications, files, and programs into a moreconsistent approach. The introduction of ProgramPlanning and Budgeting in universities andcolleges is also encouraging the informationsystems approach.

Table 2.1 gives a summary of the major computerfacilities currently existing in Minnesota highereducation and their costs. A somewhat qualitativeapproach to placing the present computingcapability of Minnesota higher education in

4-14

perspective is to compare spending with othersimilar states or states with similar educationalexpenditures. Over the past five years, from 1964through 1969, yearly expenditures for computerfacilities in Minnesota higher education grew by afactor of four, with the greatest increase comingin 1967. A marked leveling off occurred in 1968and 1969. Figure 2.1 shows the relative spendinglevels for computer facilities in the public sectorof higher education in selected states during thepast five years. Neighboring states were chosenfor their similarity to Minnesota in economy,population, and certain higher educationalfactors. Except for Indiana, the comparisons arenot surprising. In recent years public institutionsin Illinois, Ohio, and Wisconsin have beensomewhat better funded for computer facilitiesthan have their counterparts in Minnesota.However, the differences are not large. In theprivate sector, Minnesota colleges are relativelyworse off, not only as compared to privatecolleges in other states, as shown in Figure 2.2,but also as compared to publically-supportedschools in Minnesota.The computer industry marketing data fromwhich Figures 2.1 and 2.2 were drawn includefederally-funded computer facilities as well asstate-funded facilities. It was impossible toseparate these factors in the data for all the statesshown, but those which were checked indicatedthat the total value of computers amounted toalmost twice the cost of state-supportedinstructional facilities alone. Since the data wereessentially comparable for all the states, it is theirrelative ranking which is significant.

1

I

I

]

Table 2.1. Existing Computing Facilities Used for Instruction and Administration in MinnesotaPost-Secondary Education, 1970-71 (Includes those approved for acquisition)

FacilityI

Equivalent NetPurchase Price

UNIVERSITY OF MINNESOTA

Instruction and Research Systems

6600(a)(d) $2,560,000West Bank 3200 (a)(d)

450,000Duluth 3200(b) 440,000Health Sciences 3300(aXd) 825,000Hybrid Computer(aXd) 580,000Time-Shared Computing from Commercial Services 60,000Administrative S stem 1 510 000

STATE COLLEGES

Small Computers (at all colleges) 967,000Mankato Regional Computer 1,200,000

STATE JUNIOR COLLEGES

Administrative Computer 163,000HECC FUNDED PROJECTS(c) 253,700

Junior College Time-Sharing System

State College Projects:

Bemidji PDP-8 .Mankato Computer Science Curriculum DevelopmentSouthwest Academic ComputingMoorhead IBM 1130 SupportSt. Cloud Instructional Time-SharingTest Implementation of CAMPUS Management System

AREA VOCATIONAL-TECHNICAL SCHOOLS

Alexandria HW200 221,000

Mankato 360/25 233,000Hibbing HW200 232,000Moorhead 40,000Willmar 40,000

PRIVATE COLLEGES (ESTIMATED) 850,000

TOTAL $10,614,700

(a)These systems are used both for instruction and research. In Table 5.4, the operating budgets are for instructional usageonly. For 1969-70, the instruction versus research usage was as follows: (a) 6600, 31% and 69%; (b) 3300, 33%and 67%; (c) Hybrid 20% and 80%

(b)For planning purposes, these systems are assumed to be 100% instructional.(c)For both facilities and support.(d)Significant fractions of the costs of these facilities came from federal or private sources (see Appendix H.4).

4-15

RE

LAT

IVE

SP

EN

DIN

G F

OR

CO

MP

UT

ER

FA

CIL

ITIE

S, 1

964-

3.96

9P

UB

LIC

LY S

UP

PO

RT

ED

CO

LLE

GE

S A

ND

UN

IVE

RS

ITIE

S

IND

IAN

A

OH

IO

IWIS

CO

NS

IN

ILLI

NO

IS

MIN

NE

SO

TA

IOW

A

MIC

HIG

AN

RE

LAT

IVE

SP

EN

DIN

G F

OR

CO

MP

UT

ER

FA

CIL

ITIE

S, 1

964-

1969

PR

IVA

TE

LY S

UP

PO

RT

ED

CO

LLE

GE

S A

ND

UN

IVE

RS

ITIE

S

ILLI

NO

IS

OH

IO

WIS

CO

NS

IN

MIN

NE

SO

TA

IND

IAN

A

MIC

HIG

AN

IOW

A

i

I

I

I

I

..I

)'

3. GOALS AND THE COMPUTER RESOURCESNEEDED TO MEET THEM

How much computing capacity does the State ofMinnesota need to provide an adequate level ofcomputing services in higher education? Thissection presents quantitative estimates of the rawcapacity needed in the two major applicationareas: educational computing (instructionalcomputing and computer training) andadministrative data processing. The analysis onwhich these estimates are based is presented inAppendix H.6; the goals of the estimates are:

Achievement of the Pierce Reportlobjectives for educational computing forthe State of Minnesota, i.e., to provideeducational computing services for all ofhigher education in Minnesota equivalentto that available at the leading universitiesof the nation in 1967-68.

Establishment of machine-readable databases and an administrative data processingcapacity which is sufficient to support themanagement information needs ofinstitutions and systems of highereducation and lay the basis for program

,...... planning and budgeting.

These quantitative estimates of capacity will bepresented for the three major hardwareecomponents of computer systems:

IIIII

I

..4.-

Input/output or terminal capacity.

Central processor capacity.

Mass storage capacity.

Three special application areas are also discussed:a statewide library automation system;computer-assisted and computer- managedinstruction; and general information services.

EDUCATIONAL COMPUTING

Instructional Computing

Instructional computing is the most importantarea of educational computing for students inpost-secondary institutions. For undergraduatesthe analysis in Appendix H.6 follows the mei:hodsof References (1) and (7) based on Minnesotarather than national or hypothetical data. Thecomputing needs of graduate students were alsoincorporated into the analysis given in AppendixH.6. This analysis first classifies the variousacademic areas of study as to whether theyinvolve substantial, limited, or casual use ofcomputers. These categories are defined in TableH.6.1, and the major areas of study are classifiedin Table H.6.2 in Appendix H.6. These degrees ofcomputer utilization by academic area, based onMinnesota student populations, cannot beconsidered precise. However, there exists no morereasonable basis on which to estimateinstructional computing needs at the presenttime.

Because of the differences in the kinds ofprograms offered at the area schools and juniorcolleges, their computing needs were analyzedseparately from those of the degree-grantingcolleges. In applying the results from Tables H.6.2or H.6.3b in the Appendix, it must be kept inmind that they represent averages over a class ofstudents. Although a student enrolled in a coursein descriptive biology may use a computer not: atall, another student in a beginning genetics coursemay use it twice as much as the typical"substantial" user.

4-17

The analysis in Appendix H.6 averages thenumber of problems per academic year in TableH.6.1 over the three usage categories weighted byTable H.6.2, showing that 11.3 problems perstudent are assigned on the computer peracademic year. Assuming a 14-hour school day,200 school days per year and 80% available timeon the computer yields 2,240 computing hoursavailable per academic year7. This analysis allowsone to estimate the computing load for both atime-shared computing system (TS) usinginexpensive low-speed terminals and a

remote-job-entry system (RJE) with a muchsmaller number of medium-speed terminals.

Among computer scientists, computer centermanagers, and computer users there areconsiderable differences of opinion concerningthe relative merits and costs of time-shared andbatch-processed computing. Batch processingtends toward optimization of machine andoperating system efficiency, while time-sharedcomputing emphasizes the efficiency of the userat some apparent expense in hardware andsoftware. It is apparent to us that mostinstructional applications of computers can becarried out on a batch processing system, eitherthrough remote-job-entry terminals or at thecomputer site. It is equally apparent that some ofthese applications could be carried out moreeffectively (in terms of student learning) on aninteractive time-shared computer system, and thatthe educational value of other applications is

severely damaged if time-shared computing is notavailable. Programs in which students interactdirectly with simulations of engineering, social,business, political, medical, and other systems areprime examples of this latter kind of application:the student decision-making process needs to beon-line with the computer program. R is shown inAppendix H.8 that, if time-shared computingcapacity is used with a certain amount ofdiscretion, it is not necessarily much moreexpensive than batch processing. Withoutdwelling further on these arguments, this reportwill proceed on the basis that both kinds ofcomputing are necessary to meet the goals ofhigher education, and in about a 3:1 ratio ofbatch:time-shared computing.

4-18

A ime-shared computer system will be definedhere in terms of the user; it must include thefollowing characteristics:

Keyboard input/output, 10-30characters/second.

O User may interact with his program duringits execution.

It Response time of five seconds to a requestnot requiring significant computation.

The capacity of a time-shared computer system ismeasured in terms of the number of activeterminals.

Table H.6.4 gives estimates of terminal hours perproblem7. Averaging the loads estimated inTables H.6.1, 2, and 4 over the three categoriesand dividing by the total available time yieldls 150students per time-shared terminal in a college oruniversity. The same analysis applied to juniorcolleges (Table H.6.3b) yields 333 students perterminal. Time-shared terminals are not suitablefor vocational training in data processing, so theiruse has not been projected for areavocational-technical schools.

The dedicated computer capacity necessary toserve a given number of low-speed time-sharedterminals can best be estimated as the number ofactive terminals serviced. Over the past severalyears, experience with several time-sharingsystems of various sizes and costs has shown howmany active terminals they can serviceadequately. Experience also suggests an average ofone request per active terminal every thirtyseconds7. Assuming the equivalent of fifteenFORTRAN statements compiled per request anda factor of two to accommodate peak loads leadsto a compiling rate of 6.67 statements per secondper 1000 four-year college students, and threestatements per second for 1000 two-year collegestudents.

The mass storage capacity of a time-sharingsystem may be estimated by relying onexperience at the University of Minnesota andReference (7), assuming that each active userneeds an average of 3,000 characters of storage

I

I

for his own programs and data. An average of thenumber of courses using the computer leads to afigure of about 40 percent of the total students ina college using the computer at any one time(Appendix H.6). The analysis indicates a need for1.2 million characters per 1000 students at 4-yearcolleges and 600 thousand characters per 1000students at 2-year colleges.

A remote job entry system employing remotebatch access to a regional computer facility canalso satisfy the needs established above. Ananalysis of the required capacity for this mode ofoperation is given in Appendix H.6 based onTables H.6.5 and 6 and Reference (7). Assumingmedium-speed RJE terminals which read cardsand print at about 300 cards and 300 lines perminute, the total number of terminals required toserve 1000 students was estimated asapproximately one-half RJE terminal per 1000students at 4-year colleges.

Carrying out a similar analysis for the morelimited needs of students at junior colleges yieldsabout one-sixth RJE terminal per 1000 students.The obvious implication of this figure is that evena single RJE terminal has substantially morecapacity than the largest Minnesota junior collegeneeds for instructional computing.

The computer capacity needed to service theseterminals is estimated on a peak load basis as 1.2compiled FORTRAN statements per 1000four-year college students per second, with 0.4the corresponding figure for two-year collegestudents.

Another way to furnish remote-job-entrycomputing is via low-speed terminals rather thanmedium-speed terminals. Although it is ratherinconvenient for students to work with punchedpaper tape as an input medium from a simpleteletype terminal, it can be done. Mark-sense cardreaders may also be attached to such terminals,increasing both their convenience to students andcost. About 20 low-speed terminals would be

equivalent to one medium-speed terminal. Thesame amount of computing capacity will beneeded for remote-job-entry regardless of thespeed of the terminals used. The economic andpedagogical tradeoffs between low andmedium-speed remote-job-entry and time-sharingare discussed in Appendix H.B.

A mixed time-shared/remote-job-entry systemmay be a better model of the operation mode of aregional computer center than a system which isonly one or the other. As a basis for balancing thetwo types of usage, the analysis of Section 1.1.3in Appendix H.6 yields the result that 13.3time-sharing terminals are equivalent to a remotebatch terminal. Assuming an additional 30%overhead factor for the time-sharing mode ofoperation, the analysis also indicates a

requirement for seven times the raw computercompiling power to serve 1000 students in atime-sharing as compared to a batch processingmode. This is currently the mode of operation ofthe computer at the University of MinnesotaDuluth.

Instructional Computing for Graduate Students

The analyses in References (1), (7), and AppendixH.6 are directed to the computing needs ofinstitutions which offer primarily undergraduateprograms. As such, they suffice for all of theinstitutions in Minnesota except the Twin Citiescampus of the University of Minnesota. The statecolleges and the Duluth campus of the Universityall have graduate programs comprising a fewpercent of total enrollment.

Much of the instructional computing associatedwith these graduate programs will becourse-related rather than large thesis researchcalculations, and it may be lumped together withundergraduate instructional computing.

The information on computer utilization at theUniversity in 1969-70 in Appendix H.4 showsthat graduate students consume a substantialamount of computing capacity not specificallyrelated to their course work.

4-19

Using the data from Appendix H.4 on universitycomputer utilization for 1969-70, plus enrollmentdata summarized in Appendix H.10, it waspossible to develop a ratio between graduate andundergraduate student computer needs. This ratioof 3.5 between per-student graduate andundergraduate computing needs will be used,together with the enrollment projectionssummarized in Appendix H.10, to estimate thetotal needs of the University of Minnesota forinstruction-related computing.

The data processing, computer programming, andcomputer science programs and courses in all ofthe systems of higher education provide specificvocational raining for their students. The generalrequirements of these courses and programs havebeen includ3d in the analysis of Appendix H.6,along with the specific requirements of computertraining in tie vocational schools. This analysis isbased on '1974-75 enrollment projections inaccounting and data processing from the StateDepartment of Education. These presume thatthe 11 existing data processing programs will beincreased to 13 by 1974-75, with enrollmentsranging from 25 to 150. However, there is somequestion as to whether this many programs areneeded. This issue is addressed in therecommendations in Section 5; but for theanalysis in Appendix H.8, the present projectionsof the State Department of Education have beenassumed.

Research Computing

As discussed in Section 1, the two systems ofhigher education which have a defined mission inresearch and graduate training are the Universityof Minnesota, and to a lesser extent, the statecolleges. The computer capacity required toprovide research services at the University will beexcess capacity over and above the instructionalcomputing needs. If the state colleges possessenough computing power to accommodate theirinstructional needs, they should have littledifficulty in providing enough excess time to takecare of most of their research needs. Certainparticularly large computing or data processing

4-20

problems may be more economically processed atthe University's computer center, or eventually atone of the regional computer centers.

Summary of Educational Computing Needs

Table 3.1 summarizes the number of terminals,the computer capacity, and the special massstorage capacity required to meet the estimatedneeds for instructional computing of the specificinstitutions and systems of higher education inM i n nesota for the academic years 1969-70,1975-76, and 1980-81. For 1975-76 and 1980-81,the needs have been scaled according to totalenrollment projections from the MinnesotaHigher Education Coordinating Commission andthe various systems of higher education (seeAppendix H.10). The data obtained from thesedifferent sources is not always consistent, but anydiscrepancies and errors are far smaller than otheruncertainties in this analysis. If the enrollmentprojections change significantly for any of theinstitutions or systems, then the figures in Table3.1 must be adjusted accordingly. It should benoted that the capacity estimates are properly per1000 full-time-equivalent (FTE) students, ratherthan being proportional to total enrollment(headcount). The extrapolation scaling is based,therefore, on the reasonable assumption that theratio of FTE to total enrollment is constant from1969 to 1981.

ADMINISTRATIVE DATA PROCESSING

An administrative data processing system for aninstitution or system of higher education willconsist of data bases containing information forthe following areas:

Student Records

Staff and Payroll

Financial

Courses and Curricula

Space

Property Inventory

Table 3.1. Summary of Terminal and Peak Loadfor Instructional Computing in Minnes

Institution or System

1969-70

Enrollment(FTE)

Time-Shared System Remote Job-Entry System Th

Terminals

Peak ComputerCapacity

(FORTRANStatements

Per Second)No. of

Terminals

.

Peak ComputerCapacity

(FORTRANStatements

Per Second)

ProjectedEnrollment

(FTE)

UNIVERSITY OF MINNESOTA 44,610 460 460 34.6 83.0 56,364

Twin Cities

Undergraduate and Professional 33,519 224 224 16.8 40.3 38,000

Graduate 8,160(a) 190 190 14.3 34.3 11,520(a)

Duluth 5,279 35 35 2.6 6.3 8,050

Morris 1,565 10 10 0.8 1.9 2,900

Crookston 459 1 3 0.1 0.2 1,014

Waseca - - - - - 600

STATE COLLEGES 35,528 237 237 17.7 42.7 46,090

Bemidji 4,652 31 31 2.3 5.6 5,790

Mankato 10,960 73 73 5.5 13.2 13,640

Moorhead 5,249 35 35 2.6 6.3 7,000

St. Cloud 8,863 59 59 4.4 10.6 11,160

Southwest 2,153 14 14 1.1 2.6 4,100

Winona 3,651 24 24 1.8 4.4 4,400

PRWATE COLLEGES (4.YEAS) 27,627A

184 184 13.8 33.2 33,800 ITypical 1,000 6.7 7 0.5 1.2

STATE JUNIOR COLLEGES 15,914 48 48 2.6 6.4 26,173

Typical 1,000 3 3 0.2 0.4

TOTALS

A

123,679 929 929 68.7A

165.3 161,227 14

AREA VOCATIONALTECIV4ICAL 13,435(6) - - 23 84 Terminal Hours/WY 20,449111114001.8 (MEADCOUSM

(a)Total earolkoent figures have been used for graduate students, rather than full-time equivalents, for reasons explained in the text.Mei Appendix H.6 for annaplanation of the Area Vocational-Technical School requirements. Terminals may be business minicomputer or

*like terminals, or waif stand-alone computers. AVTS headcount projections are for 1974-75 and 1979-80.

,nputer Capacity RequirementsPost-Secondary Institutions

1975-76

Time-Shared System Remote Job-Entry SystemShared System Remote Job-Entry System

Peak Computer Peak Computer Peak Computer Peak ComputerCapacity Capacity Capacity Capacity

(FORTRAN (FORTRAN Projected (FORTRAN (FORTRANStatements No. of Statements Enrollment No. of Statements No. of StatementsPer Second) Terminals Per Second) (FTE) Terminals Per Second) Terminals Per Second)

1

602 44.8 107.8 66,610 734 734 54.9 131.6

254 19 45.6 42,400 283 283 21.2 50.8

270 20 48.4 14,865(a) 347 347 26.0 62.4

26.8 4.0 9.7 10,400 69 69 5.2 12.5

97 1.5 3.5 4,150 28 28 2.1 5.0

1.5 0.2 0.4 1,560 5 5 0.25 0.6

0.9 0.1 0.2 800 2 2 0.12 0.3

: 308 23.0 55.4 58,040 387 387 29.6 69.6, 39 2.9 6.9 7,180 48 48 3.6 8.6

91 6.8 16.4 17,000 113 113 8.5 20.4

47 3.5 8.4 8,710 58 58 4.9 10.4

74 5.6 13.4 13,870 93 93 6.9 16.6

27 2.0 4.9 5,750 38 38 2.9 6.9

30 2.2 5.3 5,530 37 37 2.8 6.6

225 16.9 40.5 37,200 248 248 18.6 44.6

79 4.2 10.5 32,200 97 97 5.2 12.9

INe1,214 88.9 214.2 194,050

11111,1,466 1,466 108.3

J258.7

24 146 Terminal Hours/Day 39,850(b) - - 24 164 Terminal Hours/Day

- .

2 4-21/4-22

Various applications programs access one or moreof these files to produce automatically thereports, forms, records, checks, and otherdocuments and data required for theadministrative operations of the institution orsystem. Within higher education in Minnesota, theUniversity of Minnesota has what is perhaps themost complete data processing system inoperation and in the process of implementation.For this reason, and because adequateinformation was not as readily available fromother sources, the University's system was takenas the basis for an estimate of the computersystem capacity required to provide completedata processing services to all institutions in thestate.

The units in which this capacity has beenestimated are:

Mass storage, in characters or bytes

Input, in cards or similar record entries per7114 year

Output, in lines per year

..... Computer transactions per year

Required output rate, in lines per minute

Required computing capacity in computertransactions per minute

The last two items are the limiting factors oncomputer system performance and, following thepractice of Reference (7, these factors wereestimated including peak load demands. Thedefinition of a "computer transaction" was takenfrom the Auerbach Reports8; it consists ofretrieval of a file from a random-access mass

storage device (disk, updating the file, andreturning it to mass storage and to a report file.

Details of the administrative data processingcapacity analysis, complete with annotations onthe various assumptions made in scaling theresults to the different institutions, may be foundin Appendix H.6. Summaries of the total capacityestimates for 1969-70, 1975-76, and 1980-81appear in Table 3.2. The overall capacity resultsare assumed to scale within a system ofinstituticns and with time according to the totalenrollment. From an examination of the details inAppendix H.6, it will be seen that several othermore relevant factors are used to scale variousdata file sizes or applications capacities. However,all of these factors increase and decrease withhead-count. To facilitate scaling of the individualprivate colleges and state colleges, the figures aregiven in the tables in terms of a private collegewith 1000 students and a state college with10,000 students. Area vocational-technicalschools have not been included in the estimates ofadministrative data processing needs because theirneeds are unlike those of the other institutions ofhigher education, and they are parts of localindependent or intermediate school districts. (SeeSection 2 in the Local Government part of theoverall study.

All of the factors entering into Appendix H.6 andTable 3.2 were reviewed by data processingmanagement staff at the University, the StateCollege Board, one state college, and one privatecollege. Despite numerous questions concerningspecific details, the estimates were confirmed asreasonable by these reviewers.

4-23

SPECIAL APPLICATIONS

As mentioned previously, the three specialapplications areas related to higher education andtreated in this study are automated librarysystems; computer-assisted andcomputer-managed instruction; and informationservices.

A Statewide Automated Library-System

The goals of a comprehensive statewide librarysystem are to furnish several kinds of services toall libraries in the state Through a centralizedprocessing facility. The major kinds of services tobe provided can be classified as:

1.

2.

3.

4.

Catalog production production of UnionCatalogs of holdings and of serials(periodicals).

Technical processing book ordering andpreparation, cataloging, and cardproduction.

lnrormation services retrieval ofbibliographic and archival informationfrom available machine-readable data bases,such as the Chemical Abstracts, Library ofCongress tapes, and archival census bureautapes.

On-line bibliographic services for librariesthroughout the state, providing rapidaccess and updating of holdings and otherbibliographic information to allparticipating libraries.

All of these services are to be based eventually ona common bibliographic data base which vviilform the core of the system. Because of themagnitude of the job of assessing the technicalsoundness of the plans developed by theUniversity Library Systems Staff, and identifyingviable solutions to some of the serious interlibraryorganizational problems that were apparent, anoutside consultant who is a specialist in librarysystems was retained Dr. Robert M. Hayes of

4-24

Becker and Hayes, Inc., and Director of theU.C.L.A. Institute of Library Research. On thebasis of an intensive three-day visit to the TwinCities and the documentation produced by theUniversity Library Systems Staff and variousother librarians, libraries, and library agencies inthe state, Dr. Hayes submitted a brief report andrecommendations, which appear as Appendix H.2of this study.

During and after his visit and subsequent todistribution of his report to a number of thelibrary people involved, we have receivednumerous comments on Dr. Hayes' evaluation ofMinnesota's statewide library automation plans.As far as the goals set forth in the Hayes Reportare concerned they are agreed upon as good andnecessary by the people concerned with librariesin the state. The overall development effortrequired to achieve the goals has been estimatedby Hayes in Tables 5 through 9 of Appendix H.2,in terms of dollars per year. It ranges from$700,000 for the first year to $3.3 million in thefifth year; after this investment for developmenthas been made, the operation would continue ona service-for-fee basis. The investment in a

statewide library system probably will not reducethe costs of library operation overall, but it cangreatly improve the services libraries can provide.And, if there is wholehearted participation andcooperation by all of the major public andacademia libraries, then the costs of libraryoperations can be prevented from escalatingrapidly at the same time that service deteriorates.For example, if this central service were availableit would not be necessary for each library in thestate to acquire its own technical processingcapability. On the basis of Dr. Hayes' report(Appendix H.2) and the reactions to it and to hisvisit, recommendations related to a statewideautomated library system are made in Section 5of this study. In addition, some comments appearthere on the problems of organizational structuresfor operations across system boundaries, as theyapply to library automation and as they arereflected in the Hayes Report.

i

Table 3.2. Need Estimates for Administrativ

1969-70

Headcount

MassStorage

(Millionsof

Characters)

RequiredOutput

(Lines/Minute)

Required PeakComputerCapacity

(Transactions/Minute)

Number ofQuery

Terminals for,Student

ReglitrationProjectedHeadcount

UNIVERSITY OF MINNESOTA 50,415 660 986 2,200 101 63,050

Twin Cities 42,884 85 50,150

Duluth 5,580 36 54 123 11 8,500

Morris 1,510 3 2,800

Crookston, Waseca 441 2 1,600

STATE COLLEGES 37,681 328 496 1,067 76 48,844

Bemidji 4,716 41 62 133 10 5,857

Mankato 12,090 105 159 343 24 15,045

Moorhead 5,235 46 69 148 11 6,988

St. Cloud 9,557 83 126 270 19 12,028

Southwest 2,206 19 29 63 4 4,209

Winona 3,877 34 51 110 8 4,717

STATE JUNIOR COLLEGE SYSTEM I 17,544 123 215 422 36 29,080

PRIVATE COLLEGES (4-Year) 27,137 400 325 1,055 - 33,189

Typical 1,000 11 12 39 2 1,000

I

ta Processing Capacity in Minnesota Colleges

finstorageillionsof

racters)

327

45

126

51

131

61

105

37

41

204

490

11

1975-76, . .

RequiredOutput

(Lines/Minute)

Required PeakComputerCapacity

(Transactiohs/Minute)

Number ofQuery

Terminals forStudent

RegistrationProjectedHeadcount

MassStorage

(Millionsof

Characters)

RequiredOutput(Lines/Minute)

Required PeakComputerCapacity

(Transactions/Minute)

Number ofQuery

Terminals forStudent

Registration

1,240 2,790 126 74,350 885 1,460 3,280 149

57,050 114

68 188 100 11,000 49 80 244 22

17 4,000 8

3 2,300 5

645 1,383 97 61,412 533 808 1,741 123

77 166 12 7,286 63 96 206 15

199 426 30 18,713 163 247 532 87

92 198 14 8,691 75 115 246 17

159 341 24 14,961 130 197 424 30

56 119 8 5,895 51 77 167 12

62 133 9 5,866 51 76 166 12

356 698 60 35,450 248 435 851 71

398 1,290 36,427 538 437 1,420

12 39 2 1,000 11 12 39 2o4-25 / 4-26

Computer-Assisted Instruction

As discussed in Section 1 of this report, thecomputer can be used not only as a tool bystudents (instructional computing), but also as aninherent part of the educational process as adevice to present material, problems, and decisionsituations and respond to a student's answers,solutions, and requests on the basis of hisindividual performance (computer-assistedinstruction, or CAI) and as a device to aid ateacher in managing the progress of a studentthrough an instructional experience tailored to hisindividual needs (computer-managed instruction,or CMI). Whether the computer can be efficientand effective in these tasks is not yet known;there are optimists and skeptics both. It is

imperative that higher education find out whetherit is feasible to use computers in this way, for tworeasons.

Firstly, the computer has the potential toenhance greatly the effectiveness of education incertain fields to increase the effectiveness andefficiency with which students learn certain kindsof skills and concepts, as in foreign language andsome areas of medical science; and toaccommodate the presentation and pace to thedifferent needs of individuals. Secondly, thecomputer may be an efficient, less expensive wayto provide education in those areas where it iseffective, as compared with traditional methodsin higher education. A major cost in education isthe cost of highly-trained manpower, which islargely determined by prevailing salaries andwages based on productivity of goods and servicesi n other parts of the economy. In highereducation there is no "product" with an

easily-defined market value in the usual sense.And there has been, until recently, no way ofincreasing "productivity" in higher education tokeep the total cost from soaring to levels thatrepresent a significant burden to the taxpayer, theparent, and the student. Computer-assisted andcomputer-managed instruction offer the potentialfor significant economies to reduce the rate atwhich the costs of higher education increase.Although success cannot be guaranteed, thepotential payoff is great enough to justify, or

even demand, a modest investment for researchand development.

On the basis of this kind of assessment of theeconomies and effectiveness of higher education,as well as some limited experience in the medicalsciences, the Ohio State University has recentlydecided to make a substantial investment in CAIdevelopment and has acquired a computer systemcostng about $25,000/month (Class D in TableH.7.1) for this and other instructional resourceapplications. With a much smaller investment overthe past two years, the Center for Research inHuman Learning at the University of Minnesotahas developed a staff with more experience and,in our opinion, a much sounder fundamentalunderstanding of the basic and applied problemsof instructional design and CAI than exists atOhio State, or at any but a very few universitiesin the world.

The Executive Officer of the Minnesota Centerfor Research in Human Learning, ProfessorRussell W. Burris, has prepared for this report apaper which appears as Appendix H.3. His papersummarizes in greater detail the potentials of CAIfor effectiveness and efficiency; some of thequestions to be answered and developmentproblems to be solved before it should beimplemented on more than an experimental basis;the costs of this development; and some of thepotential per-student costs if the developmentsucceeds. The eventual costs which ProfessorBurris cites $0.34 to $1.60 per student hourare less than or comparable to the present costs ofconventional instruction. However, they possessthe notable advantage that costs of computerapplications tend to decrease (or at least remainconstant) as technology advances, while wagerates, upon which the costs of traditionalinstruction are based, change in the oppositedirection.

In addition to this activity in CAI, two Minnesotainstitutions Southwest Minnesota State Collegeand Macalester College are planningdevelopment efforts in the computer managementof training for school teachers. This offers similarpossibilities for increased effectiveness and

4-27

attention to the needs of the individual student,as well as efficiency in terms of cost economies. Itis expected that the development efforts at thesetwo colleges would benefit substantially fromcontact with and support from the Center forResearch in Human Learning.

The specific recommendations of this report onCAI and CMI development, based on theseconsiderations and on Professor Burris' paper inAppendix H.3, may be found in Section 5. Thegoals toward which these recommendations aredirected are four:

1. To find out whether CAI or CMI cansignificantly increase the effectiveness ofhigher education in specific fields and morenearly tailor the instructional materials andstrategies to the diverse individual needs ofstudents;

2. To establish the real costs of CAI and CMIand determine to what extent theserepresent savings which will slow the rise intotal costs of higher education;

3. To involve faculty members from allsystems and institutions of highereducation in the state in the developmenteffort, and to disseminate information onthe work so that results may be adoptedand adapted to as many institutions aspossible and as quickly as they are provento be efficient and effective;

4. To develop pilot programs in a number ofdisciplines, and to establish CAI learning

4-28

centers from which formal instruction isavailable using CAI and CMI technologies.

Information Services

Several information retrieval projects havedeveloped in various departments of theUniversity of Minnesota as a part of their researchprograms. Some of these, such as the DiabetesInformation Center in the School of Medicine, arevery specialized. However others, such as RAFT(Rapid Analysis Fiscal Tool) and MAPS(Minnesota Area Planning Service), are muchmore general in nature and could serve statewideinformation needs which are much broader thanthe research programs of certain departments orschools. At the present time these informationservices are in a developmental rather thanoperational or "production" stage, and thus are inthe proper research province of the University;however, as they near operational status andbegin to attract a clientele from industrial firmsand government agencies, they should be fundedand managed like other statewide informationservices and resources.

Perhaps the best way to encourage thedevelopment of RAFT and MAPS and carry theminto operational status would be to fund themthrough the Computer Services Division of theDepartment of Administration by means ofcontracts to the University departments involvedin their development. As these systems becomeoperational, then the State can take over theirmanagement as a statewide information resource.

4. FACILITIES AND COSTS TO MEET THE GOALS

This section wi!I summarize the kinds of facilitiesavailable in mid-1970 to meet the goals estimatedin Section 3 for the instructional computingneeds of Minnesota higher education, and forproviding the administrative data processingcapacity needed for efficient and effective collegemanagement and program budgeting. Thefacilities and order-of-magnitude cost estimatesfor the computers themselves will be identified interms of remote-job-entry (RJE), time-sharing(TSS), and mixed (RJE and TSS) systems. Foradministrative data processing and vocationalschool computing, both batch processing and theRJE mode of operation are considered. Much ofthe supporting detailed analysis for this section iscontained in the tables in Appendix H.7.

It is not the intent of tnis report to configurecomputer systems in detail to meet the specificneeds of institutions but rather to give generalindications of the kinds of facilities needed forvarious applications and their costs. The mentionof specific hardware does not mean that thehardware configuration will necessarily performthe task indicated at the cost given. This kind ofinformation is included only to provide a generalidea of the size of the system involved for thosewho are familiar with computer systems. The costestimates could easily be low by a factor of twoor high by 25%. In other words, this report can

provide useful feasibility and backgroundinformation of a general nature, but it should notbe ,ised as the basis of a system design.

COMPUTER SYSTEMS

Table H.7.1 in Appendix H.7 establishes a scale ofcomputer systems according to their rawcomputing and data processing capacities andtheir cost ranges. This scale ranges from Class Athe most powerful and expensive computersavailable today through Class I at the oppositeextreme. Figure 4.1 displays the contents ofTable H.7.1 in graphical form. The resultssummarized in Table 3.1 for instructionalcomputing, and in Table 3.2 for administrativedata processing, have been converted into theunits used in Table H.7.1 and Figure 4.1. InTables 4.1, 4.2, and 4.3 are displayed theestimated computer capacities required to achievein 1975-76 the goals of Section 3 for Minnesotahigher education. The approximate costs shownin these tables are for the computer only; they donot include terminals, operating support, or anyof the other indispensible items discussed below.Furthermore, the computer costs are based oncommercial rates and do not reflect theeducational discounts and federal grants whichoften substantially reduce these costs to the state.

4-29

Each of the three tables shows a different way ofachieving the goals:Table 4.1: Systems which support only batch

processing via local and remoteterminals. However, it is not feasible touse medium -speed remote-job-entryterminals for junior colleges. Thelargest of them in 1975-76 (about 3000FTE students) could use only half thecapacity of a terminal, and two-thirdsof them would use a terminal to lessthan 20% of capacity. Therefore, thesecolleges (and the Crookston andWaseca campuses of the University)will require service from a Class Ztime-shared computer, as in Table 4.3,at a hardware cost of about614,000/month.

Table 4.2: Systems which support time-shared andbatch computing on the samecomputer. Reasonable assumptions aremade concerning the fraction ofinstructional computing on eachcampus which would be in each ofthese modes.

Table 4.3: Separate computer systems to supportbatch processing and time-sharedcomputing. This arrangement is basedon the observation that mostinstructional uses of time-sharedcomputing involve small programs orprograms which require interactionswith the user during execution ratherthan large amounts of computation.Programs requiring large amounts ofscientific computing are run moreeffectively, more quickly, and at lesscost in a batch processing mori.Hence, the power and variety ofprogramming languages and systemsavailable on a large computer are notneeded for instructional time-sharing.The least-expensive Class Z time-sharedsystem can be used to serve theseneeds.

Table H.7.1 and Tables 4.1 through 4.3 require anumber of comments. First of all, the capacityand system cost figures in Table H.7.1 are all veryapproximate. They are based upon benchmarktests published in the Auerbach Reports8, on

4-30

other published information, and on the personalexperience of the staff of Analysts InternationalCorporation. Computer systems of specificmanufacturers are mentioned only to providesome relation between the table and the re&world. Depending on the application, a particularcomputer at a specific price may performconsiderably better or worse than the tableindicates, or than another computer with which itis classed.

Figure 4.1 shows that, with the exception of thelargest computer systems (Class A and B), the rateof compiling FORTRAN statements increases indirect proportion to costs. In the middle of therange, the data processing capacity increases morerapidly than this, but that is not true for small orlarge computers. Neither the compiling nor thedata processing rates rise as fast as predicted byGrosch's Law, for which a doubling of cost isaccompanied by the quadrupling of the rawcomputing (arithmetic) capacity. This is becausethe computing associated with bothadministrative data processing and instruction ismuch more dependent on input/output andmanipulation and interpretation of non-numericalinformation than is the straight scientificcomputing to which Grosch's Law applies.

The required computer capacities listed in Tables4.1 to 4.3, in units of FORTRAN statementscompiled per minute, represent the minimumcompiling and computing speed for satisfactoryinstructional service during the busier times of theyear. In Section 3, a peak load factor of two wasincluded in these numbers, which means that if allinstructional computing were distributeduniformly over a 14-hour day, 200 day academicyear, then the computer capacity required wouldbe about one-half that listed in Tables 4.1 to 4.3.Such a uniform distribution is clearly impossiblein practice and, were it attempted, would destroythe educational effectiveness of the computer forall but a few students and faculty. As long as thiskind of rigid, uniform scheduling is not applied toother educational resources, such as classroom,office, and laboratory space; expensive laboratoryand audio-visual equipment; library facilities; and

Table 4.1. Estimated Conwith Batch-Prc

Administrative Processing

(Lines/minute)(Tr.

t--

UNIVERSITY OF MINNESOTA 1,240 2

Twin Cities

Duluth 68

Morris

Crookston, Waseca

STATE COLLEGES 645 .1.

Bemidji 77

Mankato 199

Moorhead 92

St. Cloud 159

Southwest 56

Winona 62

STATE JUNIOR COLLEGE SYSTEM 356

Typical Junior College (Enrollment : 1,000)

PRIVATE COLLEGES 398 1

Typical Private College (Enrollment: 1,000) 12

aFor administrative processing only.

bFor instructional computing only, including Twin Cities and Morris campuses.

cState college needs could also be handled by:1) a Class D computer for instruction + Class E for administrative processing,2) two Class D's at regional centers, providing about 70% of one as excess cal

dTo provide the two-year campuses with instructional time-shared computing underhardware.

eAssumes 50% of total Duluth workload processed locally and the remainder in the

uter Capacities Required to Serve Minnesota Higher Education'ssing and Remote-Job-Entry Facilities in 1975-76

?quired Instructional Capacity Required

Computer Class Required(Table H.6.1)

ApproximateMonsthly LeaseCost

;actionsiinute)

Number ofTerminals

(FORTRANStatementsAninute)

'90 E+a 30,000

39 5,640 C or 47% of Bb 2888 (47/0 of B)188e 4.0 582 F 12,000

1.5 210 _b

0.3 36 _d

383 23.0 3,320 D+c 45,000

1.66 2.9 414 G 8,000

126 6.8 985 F+ 15,000

198 3.5 504 G 8,000

341 5.6 805 F+ 14,000

1.19 2.0 294 H+ 6,000

[33 2.2 318 H+ 6,000

)98 4.2a d F..a,d' 10,000a'd

0.16

90 17 2,440 E+ 30,000

39 0.5 72 I+ 3,500

------ .

a total of $65, 000/month, or:ity, at $80,000/month.is plan would add about $14,000 per month for

yin Cities.

4-31/ 4-32

Table 4.2. Estimated Computer Capaciti,.Time-Sharing and Remote-Job

Administrative Processing Required

(Linesper

minute)

(Transactionsper

minute) % TS

UNIVERSITY OF MINNESOTA 1,240 2,790

Twin Cities 2.

Duluth 68d68 188d188 2:

Morris 5+

Crookston, Waseca 101

STATE COLLEGES 645 1,383 2

Bemidji - 77 166 2'

Mankato 199 426 2

Moorhead 92 198 2

St. Cloud 159 341 2'

Southwest 56 119 2'

Winona 62l

133 2'

STATE JUNIOR COLLEGE SYSTEM 356 698 10(

Typical Junior College (Enrollment 1,000)

PRIVATE COLLEGES 398 1,290 2E

Typical Private College (Enrollment :1,000) 12 39 2E

a For administrative processing only.bFor instructional computing only, Twin Cities and Morris campuses.

cMorris, Crookston, and Waseca are accommodated on the Twin Cities campus computer.dAssumes 507. of Duluth workload processed locally.

equired to Serve Minnesota Higher Education in 1975-76 withtry, Batch-Processing, Using the Same Computers for Both

arctional Capacity Required

Time-Shared Computing

Instructional Capacity Required

Batch Processing via Remote Job-EntryComputer A pprox .No. TS (Statements Class MonthlyTerminals per No. RJE (Statements/ Required Lease(Active) minute) %Batch Terminals minute) (Table 4.1) Cost $

E +a 30,000131 10,380 75 30 4,230 Bb 85,000

14 511 75 3 436 F 12,00010 383 50 1 105 c

5 426 0 0 0 c

77 3, 020 75 18 2,489 C 65,00010 382 75 2 310 F 12,00023 900 75 5 739 E 25,00012 462 75 3 378 F 12,00019 736 75 4 604 F+ 16,000

7 2/0 75 1.5 220 G 8,0008 290 75 2 238 G 8,000

79 2,350 E+ 30,000

55 2,230 75 12 1,830 D+ 52,0002 66 75 1 54 H- 4,000

4-33/4-34

4

.1

1

I

1

1

I

1

1

1

Table 4.3. Estimated Computer Capacities Required to ServeSharing and Remote-Job-Entry Batch-Processing,

Administrative Processing Required

Instructional Capacity Required

Time-Shared Computing

(LinesPer

Minute)

(TransactionsPer

Minute) N TS

No. TSTerminals(Active)

(StatementsPer

Minute)

UNIVERSITY OF MINNESOTA 1,240 2,790

Twin Cities 25 131 10,388

Duluth 68d 188d 25 14 511

Morris 50 10 383

Crookston, Waseca 100 5 426

STATE COLLEGES 645 1,383 25 77 3,020

Bemidji 77 166 25 10 382

Mankato 199 426 25 23 900

Moorhead 92 198 25 12 462

St. Cloud 159 341 25 19 736

Southwest 56 119 25 7 270

Winona 62 133 25 8 290

STATE JUNIOR COLLEGE SYSTEM 356 698 100 79 2,350

Typical Junior College (Enrollment 1,000)

PRIVATE COLLEGES 398 1,290 25 55 2,230

Typical Private College (Enrollment = 1,000) 12 39 25 2 66

aFor administrative processing only.

bFor instructional computing only, Twin Cities and Morris campuses.

cAssume a Class Z time-shared computer, the hardware cost of which is $6,000/active terminal, or $150/month per active terminal.This does not include operating support,.

dAssumes 50% of Duluth workload processed locally.

computerization game cannot be played by asimple decision to mechanize. What this report isall about is not installing a payroll system on acomputer. Rather it is about systernatizing andcomputerizing the operations of governments andinsofar as feasible, their decision makingprocesses. Information systems require statementsof objectives, programs, and policies as thestarting points.

There follows a series of recommendations andguidelines to serve as a plan for accelerating theefficient development of local governmentinformation systems and computerization inMinnesota.

They are in these groupings:

General recommendation for State policy

Organizational requirements

Approaches to cooperative centers

Area Computer Centers

Orientation and training

Funding

Implementation suggestions.

GENERAL RECOMMENDATION

The general needs outlined herein are widelyrecognized. Catalytic action is required, however,to set the stage for meeting the needs. Suchaction was almost initiated by the 1969Legislature; it is the strongest recommendation inthis report that the Legislature enact at its 1971Session effective legislation on local governmentinformation systems and computerization.

In summary, it is recommended that theLegislature establish an IntergovernmentalCommission on Information Systems withcoordinative authority over all cooperativeefforts in local government information systemsand computerization. The authorities andresponsibilities of the Commission are limitedto those necessary to attaining a coordinated,statewide approach to the J building of localgovernment information systems and toestablish the framework for the orderly and

5-28

efficient development of cooperative computerfacilities. Implementation of thisrecommendation would call for amending theMinnesota Joint Powers statute (a) to require asa final step, insofar as information systems andcomputerization are concerned, approval by theIntergovernmental Commission, and (b) forsuch joint efforts, involving any school districtthe approval of the State Board of Educationprior to Commission action. In addition,developmental funding by the State on amatching basis with local government is

recommended as is Commission control overFederal, State, or foundation funding for alllocal government information systems andcomputerization developmental efforts. Itshouid be noted that the recommendaiiOns arelimited to cooperative and coordinatedactivities. "Go it alone" efforts by localgovernment are not recommended to be in thepurview of the Commission although suchefforts may be affected by Federal or Staterequirements. More specific details of therecommendations, guidelines for theiraccomplishment, and implementationsuggestions follow.

ORGANIZATIONAL REQUIREMENTS

A formalized organization structure must beestablished in order to provide the framework andthe machinery for orderly development of localgovernment information systems andcomputerization. Four specific recommendationsfollow. The alternatives are discussed in each casealong with the reasons for the recommendation.

Intergovernmental Commission on InformationSystems

Responsibilities and Authority-

Some body concerned with the overalldevelopment of local government informationsystems and computerization is necessary. In asituation so complex and involving so manydifferent governing bodies, several levels ofgovernment, and diverse subject matter areas,effective coordination is hard to attainspontaneously. TIES is an example whereexcellent coordination was attained, an

accomplishment without parallel in Minnesotaand perhaps nationally. But TIES had a numberof favorable factors working for it (geographiccentrality, common subject matter, and federalfunding being principal ones) which may notalways be present. It would be unrealistic to trustto a series of TIES-like development throughoutlocal government in time to meet the problem.Further, TIES alone cannot bring about thestatewide standardization in data elements, etc.,necessary for compatible and transferablesystems.

Another alternative is to urge the State Board ofEducation to move ahead on its own withTIES-like developments under MESA (MinnesotaEducational Service Areas) or a similar regionalbreakdown, leaving other local governmentseither to their own devices or placed under someumbrella agency that is not concerned withschool districts. But there is a real need tocoordinate school district information systemsdevelopment with all other local governmentsystems. Standardization in data elements amongall jurisdictions is the basic requirement forinterchangeable data. There is also a need toassure most effective utilization of personnel andequipment statewide.

The recommendation of this report is that anIntergovernmental Commission on InformationSystems be established with coordinativeauthority over cooperative efforts in informationsystems and computerization in local government.Only in this way will it be possible to look at theentire problem in a manner that can maximize thecost effectiveness of such efforts. It isrecommended further that State Board ofEducation approval be required for any jointeffort of this nature on the part of schooldistricts, prior to submission of the plan to the,Intergovernmental Commission. It is believed thatthrough these two steps the statutoryresponsibilities of the State Board may be met,and the necessary coordination will be obtainedthrough the Commission.

The authorities and responsibilities recommendedfor the Commission must include only those

necesSary permit statewide Coordinated effortstoward compatible information systems and toassure carefully` planned development ofcomputer `.fiailities throughout the state. Inaddition, the recommended representation on theCommission is designed so is to assure aneffective voice for all groups. FOr example,educational" representation is such that alldecisions will have ample input from schooldistribt interests. With the Commission, alloptions on methods of cooperation andcoordination are left open; and in addition, theneed for standardization may be met. Therecommended responsibilities and authorities ofthe Commission, limited to those necessary forefficient coordination, are:

1. To establish goals and objectives forintergovernmental information systems inMinnesota as well as general policiesgoverning coordination, cooperation, jointefforts, and priorities.

2. To authorize establishment of AreaComputer Centers; to promulgate policiesand procedures assuring the coordinationof the efforts of Area Computer CenterBoards, so that an integrated statewidesystem may ultimately be developed; toestablish (a) criteria for determiningeconomic feasibility for the Centers andthe several information systems, and (b)procedures for measuring actual versusexpected cost/benefits.

3. To coordinate the development andimplementation of standards ofcompatibility of procedures, programminglanguages, codes, and media to facilitatethe exchange of information within andamong systems.

1

To review and approve or disapprove allapplications requesting Federal, State, orfoundation funding for planning anddevelopment of programs with potentialimpact on local or intergovernmentalinformation systems.

5. To review and comment upon policiesaffecting intergovernmental informationexchange established by the Governor'sCommittee on State Information Systemsand/or the Director of Information

5-29

6.

Systems and provide advice to the State onlong-range policies, programs, and plansrelative to intergovernmental informationsystems.

To develop policies and design proceduresfor the security of data in local governmentinformation systems and safeguarding theconfidentiality of such data consistent withstate statutes, local ordinances, andpertinent administrative rules andregulations (also see recommendation inPart Ill on security and privacy).

I

1

It is expected that until the establishment of thisCommission, the State-Local Data SystemsGroup, discussed in Section I, will be workingtoward the same general objectives.

Revision of Joint Powers Statute

The Minnesota Joint Powers statute has servedwell as the vehicle for permitting cooperativeefforts between and among local units ofgovernment. It would serve as the base for futurecooperative efforts in information systems as itdid for TIES. It is recommended, however, thatthe statute be revised to require approval of theIntergovernmental Commission on InformationSystems prior to establishing any futurecooperative effort in information systems andcomputerization and that any such joint effortsinvolving school districts be subject to approvalby the State Board of Education prior toCommission action. The reason for thisrecommendation is that the coordinated andefficient development of a statewide informationsystem has to be accomplished within somegeneral framework as discussed in more detaillater economic feasibility, geographicconsiderations, etc. Only the IntergovernmentalCommission will have this viewpoint. If this is notdone, it will be virtually impossible (a) to avoidthe overwhelming cost of redundancy, and (b) todevelop standardized and compatible informationsystems.

Intergovernmental Commission Membership

In order to attain a true intergovernmental base,members of the Commission should represent all

5-30

governmental echelons involved withintergovernmental information requirements.Majority representation, however, should be givento local governments as; a group. Accordingly, it isrecommended that the Commission membershipshould consist of two persons representingcounties, two representing municipalities, threerepresenting school districts, three representingthe State government, and one each from theGovernor's Committee on State InformationSystems and Minnesota regional government. TheState members should be the Director ofInformation Systems, a representative of theState Board of Education, and a representative ofhigher education. A Federal observer should beobtained from a department such as HEW orHUD that is heavily involved withintergovernmental information. Thus, localgovernment would have seven of the twelvemembers; school districts would have four,including the Board of Education representative;counties and municipalities would have a total offour.

It is further recommended that the appointmentsbe made by the Governor, to serve at his pleasure,and that local government representation be madefrom nominations submitted by the Associationof Minnesota Counties, by the League ofMinnesota Municipalities, and jointly by theMinnesota School Board Association and theMinnesota Association of School Administrators.Finally, it is recommended that the State Directorof Information Systems be chairman.

Local Government Division

The Intergovernmental Commission must have aneffective staff to carry out its heavyresponsibilities. One approach would be todevelop a separate and independent staff toimplement its program. A second approach, andthat recommended herein, is to use the staff ofthe Local Government Section of the Division ofInformation Systems. The advantages of thisrecommendation are that (a) the necessaryinterfaces with state agencies will be more readilyattainable, (b) staff costs will be minimized, (c)

staff specialists from the Division of InformationSystems will be more readily available and moreeffectively used, and (d) the relationship betweenthe recommended chairman (Director ofInformation Systems) and the staff should bemost effective. The disadvantage is the possibilitythat the State's interests may be given undueweight, a possibility that the Commission itselfshould be able to control.

The Local Government Section, then, would haveas its primary responsibility serving as staff forthe I ntergovernmental Commission onInformation Systems. As discussed later, it wouldalso be closely involved with the Area ComputerCenter Boards, particularly in systemsdevelopment and standardization efforts, andwould work closely with the EDP Coordinators inthose State agencies involved with localgovernmental functions, such as the Departmentsof Education and Welfare. (Note that a LocalGovernment Unit began operations in the StateComputer Services Division in July 1970.)

Area Computer Center Board

Among the duties recommended for theIntergovernmental Commission were authorizingthe establishment of Area Computer Centers andpromulgating policies and procedures assuringcoordination of the efforts of the Area ComputerCenter Boards. The centers, then, would beauthorized by the Commission and would beoperated by the Area Computer Center Boardunder broad policies established by theCommission to assure integrated, consistent, andeconomical systems analysis, programming, andcomputerization efforts. Without this, therewould be no assurance of attaining any of theobjectives discussed in this part of the report.

The Area Computer Center Board should beelected by the local governmental bodies involvedwith the Center. Since the demands of the Boardmay be substantial, the membership should belimited to a workable number such as seven. TheBoards may, of course, appoint standing

committees of an action or advisory nature. Inthe event that the computer base for the AreaComputer Center is either temporarily orpermanently a state, higher education, or AVIScomputer center, the I3oard would be concernedwith systems development and with the prioritiesand scheduling of the work of local governmentbut not with the actual operation of thecomputer facility.

Systems Coordinators

Certain departments of the State governmenthave supervisory or regulatory authority overcertain local governmental functions. Where thisis true and where State funds or Federal fundschanneled through the State are involved, it isrecommended that planned expenditure fordeveloping information systems and/orcomputerization by such local governments forthose particular functions must have approval ofthe related State agency prior to submission tothe I ntergovernmental Commission. Principalexamples are the Department of Education andthe Department of Welfare. The SystemsCoordinator in the affected State agency shouldhave responsibility for reviewing such requests.He is to assure that the development plan isconsistent with State legislation and with thepolicies of the Intergovernmental Commissionand his department prior to recommending itsapproval to his commissioner.

APPROACHES TO COOPERATIVE CENTERS

Alternate Methods

As indicated earlier, the only cost effectivecomputer cooperation for most local governmentsin Minnesota means common use of systems andcomputers. The question is how best to do this.There are many alternatives, but they may beviewed as essentially of three types:geographically oriented, functionally based, or acombination or overlapping of the two.

A geographically oriented Area Computer Centerwould be based, for example, on a county or acombination of counties, or a State region. A

5-31

functionally based Center would be limited toserving counties or more probably, schooldistricts. A combination or overlapping situationwould arise when all school districts in a Stateregion formed one Area Computer Center and asecond center is formed by all counties and theirmunicipalities.

Location: Regional Base Preferable

Designating geographic boundaries and specificlocations for each Area Computer Center is farbeyond the scope of this project. In fact, it willbe some time before quantitative criteria can beestablished for such decisions. A review of thegeneral criteria mentioned later in this sectiongives some indication why this is true. Inaddition, and most important, the computerrequirements of information systems for countiesand municipalities are simply not known;comprehensive systems do not exist. There isbetter understanding of the information systemsin school districts, but even that level ofknowledge is imprecise. Furthermore, theinstructional needs are not now quantifiable. It issimply not possible now to divide the State into"x" number of Area Computer Centers located inspecified municipalities. But there must be a startin the development of centers.

Notwithstanding the foregoing, it isrecommended that State regions designated in theGovernor's Executive Order Nb. 60 be used as abasis for the planning of Area Computer Centers.As guidelines at this time, a Center should cover afull region, two or three regions, or a part of aregion but not parts of two regions. The extent ofcoverage would be determined on the basis of thesystems requirements and the criteria discussedlater. One region might require more than oneCenter; two or more of the sparsely settledregions might be well served by one. The reasonsfor the recommendation are the probabledevelopment of regions in Minnesota as a serviceagency for both State and local government andthe need for a reasonable base on which to beginthe planning effort. As discussed in Section 2, theState Board of Education is considering the

5-32

establishment of MESA's (Minnesota EducationalService Areas), based on State regions. Deviationfrom this recommendation should be permittedonly after exhausting all possible alternativearrangements within the regional configuration.

General Criteria for Centers

Although specifying the quantitative criteria forthe selection of an area for cooperativedevelopment was not a part of this study, thefollowing are suggested as the major factors to beconsidered. Development of quantitative criteriais a major undertaking for the IntergovernmentalCommission; it need not, however, delayimplementation of at least two Area ComputerCenters as recommended later.

Service Base. There must be a service basesufficient to justify the center. If the center isto serve all local government needs in its area,then a population criterion would serve as theservice base. If, for example, school districts inthe area are to be served another way, then thepopulation criterion would have to be adjustedupward to reflect the loss of the school districtfunction. There are no known criteria of thisnature although one study has concluded that acenter satisfactory for school district use onlywould require a base of 100,000 students.9

Utilization Capability. Related closely toService Base above, is the general level ofreadiness to use the facility. Is the potentialclientele in the service base able to make use ofthe systems and computer facilities? To somedegree this is a function of orientation andeducation in computerization. It can beexpected to be a much more important factorin school districts than in other localgovernment jurisdictions because of theinstructional uses. Such uses are almost totallydependent on the individual teacher'sunderstanding whereas administrative uses areless dependent on one person. The need forcurricula in information systems andcomputerization is emphasized later.

Financial Support. There must be evidence ofample financial support for facilities,

1

1

processing, and overhead. It is anticipated thatthe developmental costs of systems analysis andprogramming will be funded jointly by Stateand local governments. The costs of updatingthe system and its programs after the initialimplementation should be covered in theoverhead expenses. (It is assumed thatdevelopment will be accomplished centrallywith costs distributed over all the centers. Tothe extent that this assumption is not accurate,higher costs will necessarily result.)

Contract for Service. Evidence of appropriateresources and continuity of service should bewritten agreements between the area center andthe local governments involved, covering costsof providing each service desired. The contractmust include agreement not only as to theservices to be performed by the center but alsothe tasks to be performed by the user and theuser's agreement to assign a qualifiedcoordinator to represent the user with thecenter.

Geographic Location. The area covered will, ofcourse, be a function of population density andlocal government participation. The time andcost of surface transportation for batchprocessing and the cost of common carrier linesfor on-line processing are important factors inthe feasibility of any specific geographicalcoverage. Although geographical contiguity isnot mandatory, it is highly desirable and shouldbe required except under extremely unusualconditions. Geographical compactness shouldnormally be sought, and county lines shouldnot be crossed unless to reflect a school districtor a municipality in more than one county.

Communications Costs. In addition to theirrelevance to geographic location,communications costs are influenced by the

_.... information systems requirements. If mostactivity involves on-line operations, thentelecommunications costs will be relatively lowper transaction or per unit of throughput. If

...1.

communication lines are little used, then theircost per transaction goes up.

It should be clearly understood that the precedingcriteria are exceedingly dynamic; each of them ischanging as a result of external forces or becauseof interaction with the other criteria. For thisreason, precise measurement is not desirable even

if possible. It appears likely that original demandestimates will inevitably prove to be low.

Area Computer Centers

Against the background of the precedingdiscussion, specifics of Area Computer Centersare now considered. There are three alternativesthat appear to be feasible in Minnesota.

A 'Center created anew under authority ofthe Intergovernmental Commission. Anexample of this type of Center is TIES, inwhich case twenty school districts set up anew computer center under the JointPowers Statute, and employed the systems,programming, and operating staff to carryout all phases of development andoperations.

A Center already existing in highereducation. Certain of the larger highereducation institutions have or will havecomputer configurations with the capacityto handle information systems andcomputerization requirements of localgovernments including school districts.Provision of such service would beconsistent with the institution's publicservice objectives. The development ofinstructional uses by such an institutioncould be most helpful to school districts.Depending upon developments in type andquantity of utilization over the next fewyears and the composition of the highereducation network as discussed in Part IVof this report, the institution may find thatit would be faced with a choice of servingonly higher education, limiting its serviceto higher education in favor of serving localgovernment, or greatly increasing itscomputer capacity.

A Center already existing in an AVTS.NoAVTS now has a computer configurationcapable of handling the kinds ofinformation systems and computerizationcalled for in this report. In fact, most ofthe present computer configurations in theAVTS's are less than adequate for theeducational needs of those schools becauseof their size and generation. The larger

5-33

configuration necessary for an AreaComputer Center could fill the educationalneed. Further, the programming andoperational requirements of a Centerwould offer substantial opportunity forexcellent student training. The AVTS's,under school district control, are a logicalchoice to provide service to school districtsand their schools. In addition, the AVTS'sin many communities enjoy a closerelationship with county and municipalgovernments.

Clearly, the first alternative is economicallyfeasible only when the previously discussedcomputer center criteria are satisfied, in particulara large service base in a small geographic area.Even then, temporary use of an existing largecomputer configuration would normally bedesirable until installed applications and volumesjustify the separate installations. In most cases,however, the latter examples appear as thefeasible alternatives. Either would, of course,require agreement of the appropriate state boardand of the institutional authority to serve as aCenter.

There has been general agreement at the boardlevel as to the need to obtain efficiency ininformation systems and computerizationthroughout the state. Putting such agreement intopractice, however, is extremely difficult in viewof the competing demands on institutionalleadership and resources. What would beunacceptable on a permanent basis might beacceptable temporarily as for example, in thecase of a major higher education institutionshould the higher education load become toogreat.

The functions of the Area Computer CenterBoards would vary depending upon whether theBoard actually operated a Center of its own orcontracted, for example, to obtain computerservice from an AVTS. In the first case, the Boardwould be in full charge of setting up andoperating the Center. In the latter event, theBoard would establish plans and priorities butwould simply contract with the AVTS or itsschool district for services. The Center

5-34

management would be responsible to the schoo;district but would have to work closely with theArea Computer Center Board to assure customersatisfaction.

In summary, both higher educational institutionsand AVTS's could serve as Area ComputerCenters. Certain higher education institutions arebetter equipped presently to handle such anassignment and should serve local government atleast in the developmental period of informationsystems and instructional computing. The AVTS'sappear to be a better choice for permanentassignment as an Area Computer Center. Aprincipal reason for this is that the AVTS's arealmost certain to serve their school districts aswell as others, a development that is inherent inthe State Board of Education policy. Anotherimportant reason is the often close relationship ofthe AVTS's to the officials in their constituentcommunities. The recommendations for AreaComputer Centers made later in this section areconsistent with these conclusions.

ORIENTATION AND TRAINING

Information systems, involved as they are withthe broad, complex, and largely innovativeconcepts of systems analysis andcomputerization, simply cannot be developedwithout a significant training effort. Curriculashould be established for elected officials, for topand middle management, for supervisorypersonnel directly involved in services likely to becomputerized, and for local governmentpersonnel who will form or be the coordinativelink with, the information systems cadre.

These courses should include systems analysis,information and decision systems principles, andcomputer courses, all geared to the level necessarydepending upon the function of the person.Courses should be offered in different areas in thestate. Much of this could be on a rotating basis;but some courses, particularly for those personsdeeply involved, should be offered throughappropriate educational institutions, particularly,of course, those presently offering such courses.

Educational programs of value to localgovernment information systems efforts shouldbe catalogued and periodically disseminated tolocal government through the existing channels ofthe Association of Minnesota Counties,. theLeague of Minnesota Municipalities, and theMinnesota Association of School Administrators.

Of equal or perhaps greater importance is theneed for such emphasis in teacher trainingsequences. Recommendation No. 7 in Part IV ofthis report recommends "adding requirements toexisting degree programs and certificationstandards in order to encourage pre-servicetraining in computer science for teachers andeducational administrators." Thatrecommendation concludes with this sentence:"The major problem in implementing moderninformation systems and computer instructiontechnology in local school districts is a lack ofpeople trained in these fields." That comment isborne out by the findings of the recent TIESevaluation report referred to in Section 2.

FUNDING

The key to initiation of the program outlinedpreviously is, of course, financing. It appearsunrealistic to expect developmental funding bylocal government alone. Some contribution,however, should be expected from localgovernment, perhaps on a matching basis. TIES,for example, was financed initially in part byFederal funding.

The school district distribution is suggested as a$1 augmentation of school aid per year (or $2 thefirst year) for all of the students in TIES,approximately 240,000, and all of the students inthe second TIES-like Area Computer Center,assuming the new center covers approximately100,000 students. The TIES suggestion is to givethat organization support for accelerating itsdevelopmental effort and assumes that all TIESsystems and programs would be available withoutcharge to other Minnesota jurisdictions. Themoney would be made available by the Directorof Information Systems to TIES and the newschool district Area Computer Center uponrecommendation by the IntergovernmentalCommission after reviewing plans fordevelopment to be financed by the funds. Thefunds could be made available over a one or twoyear period. It is, of course, imperative that StateDepartment of Education informationrequirements be met in any systems designundertaken by TIES or the other Center.

The amount of money recommended could easilybe doubled and spent productively, particularlyso long as the overall controls are being exercisedby the Intergovernmental Commission and theDirector of Information Systems. The odds arehigh that every dollar invested in this way willreturn an extremely rewarding pay-off to thetaxpayers in Minnesota in the future the nearfuture.

5-35

I

It would be appropriate, of course, to seekFederal and foundation funding assistance forboth of these efforts. Pooling of resources andefforts among the larger units of local governmentand the State in specific subject matter areasshould also be attempted. For example, theMetropolitan Council has listed in its 1971program a major effort on a metropolitaninformation system. As a further example, aschool district might be willing to assist withfinancing a particular project to move it ahead onthe priority listing.

To assure that requests for Federal or foundationfunding for information systems orcomputerization purposes are consistent withoverall plans and carry the potential for the mosteffective results, all such requests originating withlocal governments should be subject to the reviewand approval of the IntergovernmentalCommission. This requirement should apply tosuch funds whether intended to be granteddirectly to a local government or to be channeledthrough an appropriate State agency.

The possibility of Federal funding should beaggressively pursued by the Commission, either asa Minnesota project or in combination with otherstates. Further, joint development action withother states should be explored. The MidwesternStates Educational Information Project (MSEIP)is an example of the combination of joint actionby the Federal government and several states, as isthe Project SEARCH. The systems analysis andcomputer programs resulting from MSEIP arecurrently under careful review by the StateDepartment of Education for possible use byschool districts as well as the department itself.

IMPLEMENTATION SUGGESTIONS

Specific action to implement the generalrecommendations above is outlined below. Theseactions assume legislative adoption and enactmentof the general recommendations made previously.An interim approach is also suggested so thatdevelopment can proceed prior to such action.

5-36

2.

1

3.

The Intergovernmental Commission onInformation Systems should authorize anArea Computer Center dedicated to thedevelopment and implementation ofinformation systems for county andmunicipal governments. Although centersestablished later could well include schooldistricts as well as counties andmunicipalities, the development effort forthe latter two will take all the resources theArea Computer Center Board can marshall.( A second center, similar to TIES, is

recommended below.) Selection of the areato be covered, preferably a region or a partof a heavily populated region, should bebased on the criteria outlined earlier andon the interest and support available in thearea. The Metropolitan Council, forexample, might be an appropriate AreaComputer Center Board to serve part of itsseven county area. It is recognized thataccomplishing this task may be difficult.However, the local government needs, theavailability of some State and/or Federalfunding, and significant State technicalassistance should provide incentive to agroup of progressive jurisdictions.

The Intergovernmental Commission shouldspell out its delegation of authority andassignment of responsibility to the AreaComputer Center Board and then conductan election for the Board.

The Area Computer Center Board shouldappoint a Standards Committee and aSystems Development Committee. TheStandards Committee should undertake thesizable task of drafting a standards manualfor local government in Minnesota,including among other things, a datadictionary. Every opportunity should bemade available for other local governmentinput. In particular, TIES, the HennepinCounty data processing section, and the St.Paul-USAC project staff should be invitedto contribute. Draft manuals should becirculated to all counties, all municipalities(over 2500 population), TIES, theMetropolitan Council, and the StateDepartment of Education for review andcomment. The initial standards manualshould ultimately be presented to the

I

I

I

I

I

I

In

I

4.

Intergovernmental Commission forapproval and would then serve as the localgovernment standards manual. The LocalGovernment Section of the State Divisionof Information Systems should supplytechnical assistance in the standardsmanual development and should assureappropriate Federal and Statecoordination.

The Systems Development CoMmitteeshould be heavily aided initially by theLocal Government Section of the Divisionof Information Systems. The objective is toobtain or develop sound informationsystems. The systems selected for initialdevelopment should be those with greatestpayoff Further, they should be capable ofmodular development so that payoffobtainable from part of the system may berealized promptly without having to waitfor the completion of the entire system. Aswith standards, the systems design must besubmitted to all jurisdictions for review.Local governments with particular interest,expertise, or accomplishment in relatedsystems should be invited to participate indevelopment of transferable informationsystems. It is expected that certain alreadysuccessful systems or subsystems may beobtained free or by purchase from publicor private agencies to speed payoff andeliminate the waste associated withredundancy.

In the initial stages, the Area ComputerCenter Board should make use of acomputer capability other than its own. Itis expected that only a sizable computerconfiguration can handle the informationsystems that will be developed. The Centersimply could not utilize such equipmenteffectively at the outset. Prime candidatesfor such use include computers of the StateDivision of Information Systems andhigher education institutions The selectionof a site should be made in closeconsultation with the IntergovernmentalCommission to assure consistency with theoverall State plans and potential networkin all areas of computerization. A totallynew Center should be established only

I after exhausting the possibilities presentedIby existing facilities such as the AVTS's.

5,

6.

1

1

9.

The implementation of all except the mostproven of existing systems (such as payrollsystems) should be limited to one User.That is, one new application would betested with one user, another with anotheruser, etc. Until extensive testing is

completed, no additional users should beallowed to use a new application. Finally,comprehensive user's manuals must be inbeing prior to implementation.

The Area Computer Center Board shoulddevelop written policies, regulations, andprocedures governing the operation of theCenter including a model contract forservices; input and output controlrequirements; security, privacy, andconfidentiality of data; charging forservices rendered; etc. These should bereviewed and approved by theIntergovernmental Commission inasmuchas they should serve as a pattern for othercenters.

The Area Computer Center Board, inconsultation with the IntergovernmentalCommission, should establish an operatingplan for the first two years and a

corresponding budget.

Selection of a director for the centershould be made by the Area ComputerCenter Board. Staffing of the center shouldbe by the director subject to approval ofthe Board.

1

1

The Intergovernmental Commission shouldauthorize a second Area Computer Centercomposed of school districts. Like the firstcenter, it should preferably be based on aregion (MESA). It is recommended that theCenter be one of the AVTS computerfacilities, providing a good test of theadvantages and problems associated withhaving centers at those schools. It shouldutilize the TIES programs either as they areor reprogrammed to another computer, orit should obtain a comparable capability.An example of the latter is a combinationof the Newport-Mesa (California) systemfor administrative systems and the

5-37

EDINET (Honeywell) time-sharing systemfor instructional uses. Another possibilityto be researched is the end results of theMSE I P (Midwestern State EducationalInformation ProjeCti. In any event, thecenter should not begin from scratch butshould use systems now operationalalthough it may appropriately develop newsystems not readily availablelinder TIES orits alternatives.

INTERIM ACTION

I

Note that prior to legislative action, theState-Local Data Systems Group and the LocalGovernment Unit of the Computer ServicesDivision should lay as much groundwork aspossible for the initiation of the foregoingprogram. This should include:

5-38

Initiating a search for strong existingcomputer systems and accompanyingprograms for county and municipalgovernment which can be used as they,are;can be modified to meet Minnesota localgovernment needs; or, using the systemsdocumentation, can be helpful in

completing the design necessary few.

Minnesota local government.

Assuring the general applicability andacceptance of systems identified in (1)above.

On a pilot project basis, initiating localgovernment applications that would havepriority and be widely applicablethroughout the State.

Beginning the development of a standardsmanual for local governments includingamong other items a data dictionary toassure uniformity in terminology; dataelement specification to assure standardlength; programming language (forexample, USASI, COBOL, andFORTRAN); data communication codestructures; system analysis andprogramming documentation; feasibilitystudy and report formats; cost accountingfor information systems.

Validation and completion of the dataobtained from local governments in thecourse of this study and arranging forperiodic update of the data.

Establishing media and procedures forintergovernmental interchange of data.

Developing policies coveringconfidentiality of data and procedures forcarrying them out.

Setting up procedures for reviewingproposals for federal, state, or foundationaid in developing information systemswith local or intergovernmental impact.

Recommending policies and procedures forcharging for services rendered by areacenters.

j..e

REFERENCES

1. Hear le, F.R. and Mason, Raymond J., A Data Processing System for State and LocalGovernments, Prentice Hall, Inc., 1963.

2. Mitchel, William H., An Approach to the Use of Digital Computers in MunicipalGovernment, Doctoral Dissertation, January 1968.

3, Intergovernmental Task Force on Information Systems, The Dynamics of Information Flow,April 1968.

4. Mohrenweiser, Gary A. and Mueller, Van D., Characteristics of a Network Model forRegional Information Systems for Minnesota Elementary and Secondary School District, (AReport to the Minnesota National Laboratory, State Department of Education) 7969.

5. Keenan, W. W., A Study for the Coordination of Education Information and Data Processingfrom Kindergarten through College, Minnesota National Laboratory, September 1969.

6. Anderson, Gayle H., Preliminary Report for Development of Minnesota Educational ServiceAreas (MESA), Minnesota State Department of Education, September 1969.

7. EDP Technology, Incorporated, Evaluators' Report TIES (Total Information forEducational Systems), June 1970.

8. Weiner, Myron E., Cooperation as a Model for Municipal Use of Computers;, MunicipalInformation Technology Program, Institute of Public Service, University of Connecticut,1968.

9. U. S. Department of Health, Education, and Welfare, A Feasibility Study of a CentralComputer Facility for an Educational System, Office of Education, Bureau of Research,1968.

5-39