census data for researchers thursday feb 23, 2012 3:30-4:45 pm
TRANSCRIPT
Census Data for Researchers
Thursday Feb 23, 20123:30-4:45 pm
Getting Acquainted with Research Data
Six 1.5 hour seminars this Spring Census (Content & Issues): Today Foreclosure Crisis Data : March 8 Data from IGO’s : March 16 Geoportal/GIS resources : March 22 AddHealth : April 12 Census (Resources) : April 26
The Census Bureau spends a LOT of money – ($7 billion (2010), 11 billion dollars (2011)) - each year in its mission as the primary statistical agency of the US. It gathers an enormous amount of data about individuals, households, establishments and firms from a broad array of surveys and censuses. All of this is made possible by ongoing investments in developing new content, maintaining sampling frames, evaluating quality of responses, and responding to methodological issues and concerns.
Today, we will identify the primary data collections of the Census Bureau, looking at content, geography, access levels, standard data products, value-added research resources, and issues in cross-time and cross-survey analyses.
Today
Recurring Questions
Basic distinctions about collections Survey vs. CensusPopulation & Households vs. EconomicTitle 13 vs. Title 15Microdata vs. Aggregate data
ContentWhat questions are asked and how?
GeographyWhat data is available for what areas?
Multi-legged stoolsDrawing on multiple resources, surveys, time periods and geographies….and the strengths
and drawbacks.
Why is the Census Bureau important?
Huge data collection budget Even more money allocated on basis of data
collection (~400 Billion annually) Most widely used social science data
– High quality sample frames– Large samples sizes, small geographies– Consistency
Broad Data Collections
Population & Housing Census - every 10 years Economic Census - every 5 years Census of Governments - every 5 years
American Community Survey – annually Many additional surveys -- both Demographic &
Economic Economic Indicators - each indicator is released on a
specific schedule
Supplementary Resources
Population Projections and Estimates
Small Area Income and Poverty Estimates Small Area Health Insurance Estimates
Geographic shapefiles & resources
Behind the scenes:Sampling Frames (Household Surveys)
Master Address File (MAF)– Official inventory of known living quarters– Linked to TIGER
Housing Units– Based on Census 2000 MAF and updates from
the USPS’ Delivery Sequence File Group Quarters
– updates from the administrative records and the FSCPE
Behind the scenes:Sampling Frames (Business Surveys)
Business Register– Census Bureau’s master business list– Industry classification - NAICS– Geographic classification – states, counties, etc.– Legal form & tax status
Establishments– Places of Business
Enterprises– Firms
Behind the scenes:Sampling Frames (Business Surveys)
Source
Payroll Tax Returns
(IRS Forms 941 & 943)25 million
Sole Proprietorships’ Business Income Tax Returns
(IRS Form 1040, Schedule C)22 million
Other Business Income Tax Returns (Forms 990, 1065, 1120) 10 million
Social Security Administration Industry Codes (Form SS-4) 1.8 million
Bureau of Labor Statistics Industry Codes 1.5 million
Behind the scenes:Sampling Frames (Geography/Other)
Topologically Integrated Geographic Encoding and Referencing system (TIGER)
Boundary and Annexation Survey (BAS)– Annual; legally defined geographies
Population Estimates– Based on Vital statistics data, IRS migration,
Medicare enrollment data
Broad Data Collections
Population & Housing Census – Every 10 years – Full enumeration– Mixed mode (mail-in, CATI, in-person)– Long form/short-form (2000 and earlier)– Multiple data releases
Census 2010: ContentCensus 2010: Content
10 Questions10 Questions NameName SexSex AgeAge Relationship (to Household Head)Relationship (to Household Head) Hispanic OriginHispanic Origin RaceRace Owner/Renter StatusOwner/Renter Status
PlusPlus Whether each member sometimes lives/stays elsewhereWhether each member sometimes lives/stays elsewhere
Total number living in residenceTotal number living in residenceProbe for unreported personsProbe for unreported personsTelephone contactTelephone contact
Census 2010: ProductsCensus 2010: Products
Reapportionment data – December 2010Reapportionment data – December 2010
Redistricting data – February-April 2011 Redistricting data – February-April 2011
SF 1 – June – August 2011SF 1 – June – August 2011
SF 2 - Dec 2011 – April 2012SF 2 - Dec 2011 – April 2012
Same Sex Couple Summary File – Nov 2011Same Sex Couple Summary File – Nov 2011
Congressional District Summary File – Jan 2013Congressional District Summary File – Jan 2013
AIAN Summary File – April 2013AIAN Summary File – April 2013
State Legislative District Summary File – June 2013State Legislative District Summary File – June 2013
PUMS - TBDPUMS - TBD
Census 2010: Product DetailCensus 2010: Product Detail
P.L. 94-171 (Redistricting Data)
State and sub-state counts down to the block level are shown for the total population and the population 18 years and over for 63 race groups; and not Hispanic or Latino origin by 63 race groups. Also shown are housing unit counts by occupancy
status (occupied units, vacant units).
Census 2010: Product DetailCensus 2010: Product Detail
P.L. 94-171 (Redistricting Data)1980 1990 2000 2010
Race 5 Race ANDSpanish Origin
5 Race x2 Hispanic
63 Race x 2 Hispanic
63 Race x 2 Hispanic
Age --- Total, Age 18+ Total, Age 18+ Total, Age 18+
Housing --- Total Housing Units --- Occupied vs Vacant
Census 2010: Product DetailCensus 2010: Product Detail
Summary File 1 (SF1)Summary File 1 (SF1) About 300 tables Counts and cross tabulations Counts for detailed race, Hispanic or Latino groups, and
American Indian/Alaska Native tribes (to the tract) Tables repeat for major race groups alone, two or more
races, Hispanic or Latino, White not Hispanic or Latino Geography: block, census tract
http://www.census.gov/population/www/cen2010/glance/files/SF1_Final_1.5_Internet.xls
Examples of SF1 TabulationsExamples of SF1 Tabulations
P1 Total population (1)P3 Race (71)P8 Hispanic or Latino (17)P12 Sex by age (5-year groupings) (49)P14 Sex by age for the population
under 20 (single years of age) (43)P15 Households (1)P17 Average household size (1)
Census 2010: Product DetailCensus 2010: Product Detail
Summary File 2Detailed tables on age, sex, households, families, relationship to householder, housing units, and group quarters.
Tables are repeated by 141 race groups, 98 American Indian and Alaska Native tribes/tribal groupings, and 39 Hispanic or Latino origin groups.
Per Census Bureau Technical Documentation:
“The concept of race, as used by the Census Bureau, reflects self-identification by people according to the race or races with which they most closely identify. These categories are socio-political constructs and should not be interpreted as being scientific or anthropological in nature. Furthermore, the race categories include both racial and national-origin groups.”
“Origin can be viewed as the heritage, nationality group, lineage, or country of birth of the person or the person’s parents or ancestors before their arrival in the United States. People who identify their origin as Spanish, Hispanic, or Latino may be of any race.”
SF2 - Detailed Asian SF2 - Detailed Asian CategoriesCategories
Asian Japanese Asian Indian Korean Bangladeshi Laotian Bhutanese Malaysian
Burmese Nepalese Cambodian Pakistani Chinese Sri Lankan Chinese, except Taiwanese Thai Taiwanese Vietnamese Filipino Other Asian Hmong Indonesian
SF2 - Detailed Hispanic/Latino CategoriesSF2 - Detailed Hispanic/Latino Categories
Hispanic or Latino(of any race)
Mexican Puerto Rican Cuban Other Hispanic or Latino
DominicanCentral American
Costa Rican Guatemalan Honduran Nicaraguan Panamanian Salvadoran
South American Argentinean
Bolivian Chilean Colombian Ecuadorian Paraguayan Peruvian Uruguayan VenezuelanSpaniardAll other Hispanic or Latino
SF 2 - 42 American Indian CategoriesSF 2 - 42 American Indian Categories
American Indian
Apache Houma South American Indian
Arapaho Iroquois Spanish American Indian
Blackfeet Kiowa Tohono O'Odham
Canadian and French American Indian Lumbee Ute
Central American Indian Menominee Yakama
Cherokee Mexican American Indian Yaqui
Cheyenne Navajo Yuman
Chickasaw Osage American Indian tribes, Other
Chippewa Ottawa
Choctaw Paiute Alaska Native
Colville Pima Alaskan Athabascan
Comanche Potawatomi Aleut
Cree Pueblo Inupiat
Creek Puget Sound Salish Tlingit-Haida
Crow Seminole Tsimshian
Delaware Shoshone Yup'ik
Hopi Sioux
SF2 Subject ContentSF2 Subject Content
36 Population tables at census tract (PCT) level
10 Population tables at county level
10 Housing tables at census tract (HCT) level
Other Deccenial-based tabulations
Same-Sex Tabulation a single table, but tabulated as reported, not “edited” to unmarried partner
Congressional and State Legislative summary files – retabulation of SF1 to new boundaries
Other summary files draw upon ACS, rather than decennial
Where’s all the interesting stuff?
In 2000 (and earlier) censuses, the census used more than one form:
A “short” form, which asked basic demographic data, just like the 2010 census form (AKA – 100% data)
A “long” form, which collected both the items on the short form and a broader set of items about income, education, ancestry, language, disability, employment, etc.. (AKA – sample data)
Now, decennial census focuses solely on basic demographic data, and social and economic data are collected in the American Community Survey (ACS)
Broad Data Collections
American Community Survey Annual Replacement for the “long form” of the decennial
census. HH sample fully implemented in January 2005,
annual sample of around 3 million. Multi-mode: mail, CATI, CAPI Multiple Data releases
– 1 year, 3 year, 5 year, PUMS
ACS Content - Basic
ACS: Design of the Sample
Annual Sample Size of 3 million addresses Series of Monthly Samples of 250,000 addresses HU sample in each of the 3,141 Counties Areas with smaller populations sampled at higher
rates than those with larger populations HU Address sampling rate set by Block Final sampling rate varies between 1.6% and 10% No HU address can be sampled more than once in 5
years
ACS: Data Collection
HU addresses by three modes – Mailout of paper questionnaire in 1st month– Telephone (CATI) non-response follow-up in 2nd
month – Personal visit (CAPI) non-response follow-up in
3rd month to a sub-sample
GQ– Personal visit within 6 weeks of sample selection
ACS: Sample Design
GQ facilities sample for each state Two stratum
– Small (15 or fewer residents)– Large ( more than 15 residents)
Small – Data collected on all residents– Facility eligible once in 5 years
Large– Groups of ten residents sub-sampled– Number of groups determined by size of facility– Facility eligible every year
ACS Content Tests
2006Health InsuranceMarital HistoryVeteran's Service-connected Disability
2007Field of Degree (BA)
2010Computer Ownership-Internet AccessParental Place of Birth
2011-2013Testing of Internet Response mode
Distribution Formats
Like former decennial census data, released in both aggregate and microdata formats
Because of change to continuous sampling, however, aggregate data released at different geographic levels with differing collection frames
Sample Data Summary FilesSample Data Summary Files
Summary File ….. 3?Summary File ….. 3? 813 tables of data
Counts and cross tabulations of sample items (income, occupation, education, rent and value, vehicles available)
Lowest level of geography: block group
Multi-year estimates
Larger geographies have multiple options for estimates – 1 year, 3 year, 5 year
Comparing and interpreting overlapping multi-year estimates not intuitive: only differences come from the non-overlapping period.
EEOC: Fall 2010
37
New this time: EEOC
ACS 2006-2010 5-year file
Margins of error
2010 Census population base
2010 SOC Occupation categories
Additional variable: Citizenship
38
Basic Census Geography
Legal/Administrative Entities
Statistical Entities
A public use microdata area (PUMA) is a decennial census area for which the U.S. Census Bureau provides specially selected extracts of raw data from a small sample of long-form census records that are screened to protect confidentiality. These extracts are referred to as ‘‘public use microdata sample (PUMS)’’ files. For Census 2000, two two types of PUMAs were delineated within states.
PUMAs of one type comprise areas that contain at least 100,000 people. The PUMS files for these PUMAs contain a 5-percent sample of the long-form records. The other type of PUMAs, super-PUMAs, comprise areas of at least 400,000 people. The sample size is 1 percent for the PUMS files for super-PUMAs.
PUMAs cannot be in more than one state or statistically quivalent entity. The larger 1-percent PUMAs are aggregations of the smaller 5-percent PUMAs. PUMAs of both types, wherever the population size criteria permit, comprise areas that are entirely within or outside metropolitan areas or the central cities of metropolitan areas.
Non-Nested Geographies – PUMAs
Some Key Points to Remember
Census Geographies include nested and non-nested geographies
Some geographies defined politically, others for statistical and reporting purposes
Geographies range in size from a block to the nation as a whole, but different sorts of data available depending on type of geography
Demographic (Household) Surveys
Survey of Income and Program Participation Survey of Program Dynamics American Housing Survey Current Population Survey Consumer Expenditure Survey (CES)
And more…..
Survey of Income and Program Participation
The Survey of Income and Program Participation (SIPP) program, initiated in 1983, is a longitudinal, multi-panel survey primarily of adults in households in the United States.
Sampled households are interviewed at least nine times at four-month intervals and followed over the life of the panel. New samples (panels) are drawn periodically, ranging in size from around 13,000 HHs to around 40,000 HH’s. (annually 1984-1993; 1996, 2001, 2004, 2008)
The SIPP attempts to interview all members age 15 and older in the household during the first wave of interviewing. Subsequent interviews may be in-person or by phone, with the same interviewer speaking to the same respondents.
New members who join the household are interviewed after they join; departing members are interviewed at their new address.
Survey of Income and Program Participation SIPP information falls into two categories: the core
information, and other questions (found in "topical modules") that produce in-depth information on specific subjects and are asked at only one or two interviews.
SIPP core content covers demographic characteristics, work experience, earnings, program participation, transfer income, and asset income.
Current Population Survey
The Current Population Survey (CPS) is a monthly survey of about 50,000 to 65,000 households conducted by the Bureau of the Census for the Bureau of Labor Statistics. The survey has been conducted for more than 50 years.
The CPS is the primary source of information on the labor force characteristics of the U.S. population. The sample is scientifically selected to represent the civilian noninstitutional population.
Households are in the survey eight times: four consecutive months, eight months off, and then a final four months.
Estimates obtained from the CPS include employment, unemployment, earnings, hours of work, and other indicators. They are available by a variety of demographic characteristics including age, sex, race, marital status, and educational attainment. They are also available by occupation, industry, and class of worker.
Supplemental questions to produce estimates on a variety of topics including school enrollment, income, previous work experience, health, employee benefits, and work schedules are also often added to the regular CPS questionnaire.
Current Population Survey
Annual Social and Economic Supplement (ASEC) – (formerly called the Annual Demographic Survey or March Supplement)
Voting and Registration (November) School Enrollment (October) Food Security; every year since 1995 Computer Ownership Fertility and Marital History Fertility and Birth Expectations Contingent Workers and Alternative Employment Displaced Workers Job Tenure and Occupational Mobility Race and Ethnicity Tobacco Use Work Experience Work Schedules
American Housing Survey
Provides information on the size and composition of the housing inventory in the United State, neighborhood characteristics, characteristics of occupants. household characteristics, income, housing and neighborhood quality, housing costs, equipment and fuels, size of housing unit, and recent movers.
The AHS returns to the same housing units year after year to gather data; therefore, this survey is ideal for analyzing the flow of households through housing.
Sample of ~ 65,000 Collected for HUD Separate national (fixed sample for ~50,000, followed since 1985) and
metropolitan samples (~3,200 – 4,800 per area, every 6 years, 12-14 areas/year)
More detailed data, less geographic detail, than census
Consumer Expenditure Survey
The Consumer Expenditure Survey (CES) provides information on the buying habits of American consumers and also furnishes data to support periodic revisions of the Consumer Price Index. A new sample is drawn annually, and includes about 60,000 households.
The survey consists of two separate components: (1) a quarterly Interview Survey in which each consumer unit in the sample is interviewed every three months over a fifteen-month period, and (2) a Diary Survey completed by the sample consumer units for two consecutive one-week periods.
The quarterly interview gathers retrospective data on purchases, and focuses on regular and large expenses.
The Diary Survey contains consumer information on small, frequently-purchased items such as food, beverages, food consumed away from home, gasoline, housekeeping supplies, nonprescription drugs and medical supplies, and personal care products and services. Participants are asked to maintain expense records, or diaries, of all purchases made each day for two consecutive one-week periods.
Selected Other Data
National Crime Victimization Survey– 48,000 addresses in 809 PSU’s in US– Operating since 1972– 7 interviews over 3 ½ year period
National Corrections Reporting Program – Prison Admission and discharges. Variables include incarceration history, current offenses, and
total time served. Background information on individuals includes year of birth, sex, age, race, Hispanic origin, and educational attainment.
A variety of surveys for NCHS, e.g.– National Health Interview Survey– National Hospital Discharge Survey– National Survey of Ambulatory Surgery
National Survey of College Graduates– Baseline Survey based on Census
1993 from 1990 census, 2003 from 2000 census Follow-up surveys every 2 years (4 total per decade)
Census – Over time
One of the great strengths of Census collections is that their temporal span is quite wide.Decennial Census- Aggregate data from 1790 onward- Microdata from 1850 onwardCPS- Aggregate data from 1940’s- Microdata from 1962 onward
Census – Over time
However…- Content/Questions change- Geography changes
Lots of value-added resources to help address these issues….
Historical Census Geography
Census Tracts– First created in 1910 – 8 cities tracted and 1910 and 1920– By 1940, 60 cities tracted– Substantial increase in tracting with advent of
Metropolitan Areas in 1950– Entire nation tracted/BNA’d by 1990– Tracts can split/merge
Historical Census Geography
Places– State-specific requirements for incorporation – In 1950, CDPs introduced by Census (called
unincorporated places).– Increase in size due to annexations– Increase in numbers due to incorporation– Merging of places possible– Between 1980 and 1990, 40% of places
experienced some change in boundary
In April…..
I’ve primarily talked about what’s available, not how to get it.
In April, I’ll talk about resources for online analysis and exploration, download resources, documentation, and local (Berkeley) resources…. And if you can’t wait, come by and visit me.
Geographic grain and Margin of Error