1. Employment and Earnings Outcomes for Young Adult Bachelor’s Degree Holders: Findings From the American Community Survey 25th Annual STATS-DC 2012 Data Conference July 12, 2012 Grace Kena William Sonnenberg National Center for Education Statistics

2. Overview Background of Analyses American Community Survey (ACS) and PUMS (Public Use Microdata Sample) Datasets Uses of ACS PUMS Datasets and Comparisons with Current Population Survey (CPS) Dataset Selected Employment and Earnings Outcomes Disclaimer: The content of this presentation does not represent the views of the National Center for Education Statistics or of the Institute of Education Sciences.

3. Background Indicators on postsecondary outcomes developed as part of Gaps in Access & Persistence Study (GAPS) produced under Congressional mandate and slated for August 2012 release* –educational attainment –labor force participation –median annual earnings, and –employment in STEM occupations *Higher Education Opportunity Act, H.R. 4137, 110 th Cong. §1109, 2008

4. American Community Survey (ACS) Continuous national survey of the Census Bureau which collects similar data on people and housing as Census 2000 long form. Sample of housing unit addresses surveyed monthly=> about 3 million addresses each year. Includes institutionalized and group quarters populations. Public User Microdata Sample (PUMS) data set: –According to the ACS website, “Microdata are for those users who want to create do-it-yourself tabulations, to be able to further draw on the richness of detail recorded in the ACS.” –PUMS data sets contain about 40% of the full ACS samples. –PUMS data sets are modified to deter individual record identification. –They are available in 1-, 3-, and 5-year varieties. –We used the 1-year PUMS, 2010 data set for GAPS analyses. SOURCE: http://www.census.gov/acs/www/

5. ACS – WHY USE IT? To obtain more reliable estimates for relatively small quantities. –Example: Tabulations for small states (e.g., Alaska) or small racial/ethnic groups (Alaskan Natives) –Example: Median earnings for workforce, by state (by race, age, educational attainment, etc.) 2010, full-time, full-year workers ages 25 to 34: –ACS 1-year sample=177,364 records; –ACS 3-year sample=545,222 records; –Current Population Survey (CPS) Annual Social and Economic Supplement (ASEC) sample=16,325 records

6. ACS – WHY USE IT? To obtain more reliable estimates for relatively small quantities. –Example: State-level analyses for smaller states Median annual earnings in 2010 for 25 to 34-year-olds in Alaska who worked full-time and throughout a full year –CPS sample size is 193, ACS 1-year sample size is 312, and ACS 3-year sample size is 879 –Yields median earnings of $38,431 (CPS), $43,741 (ACS 1-year), and $39,665 (ACS 3-year) => ALL valid BUT: less than High School Attainment yields CPS sample of 4 (no valid estimate), ACS 1-year sample of 16 (no valid estimate), ACS 3- year sample of 43 (valid estimate=>$31,101)

7. ACS – EXAMPLES OF OTHER USES Example: Analyses by race at the state level –For Colorado: only 6 Blacks in CPS sample; ACS 1-year sample size is 66 –For Mississippi: only 50 Blacks in CPS sample; ACS 1-year sample size is 518 –With regard to Native Hawaiians: 1-year ACS facilitates analyses in 3 States: Texas with 47 ACS records vs. CPS with 4, Utah with 59 records using ACS vs. 6 using CPS, and Washington state with 89 ACS records vs. 7 with CPS –These are ACS 1-year examples. ACS 3-year yields even more records and more valid cell sizes.

8. ACS – Distinguishing features of 1-, 3-, and 5-year estimates 1-year estimates3-year estimates5-year estimates 12 months of collected data36 months of collected data60 months of collected data Data for areas with populations of 65,000+ Data for areas with populations of 20,000+ Data for all areas Smallest sample sizeLarger sample size than 1-yearLargest sample size Less reliable than 3-year or 5-year More reliable than 1-year; less reliable than 5-year Most reliable Most current data Less current than 1-year estimates; more current than 5-year Least current Best used when Currency is more important than precision Analyzing large populations More precise than 1-year, more current than 5-year Analyzing smaller populations Examining smaller geographies because 1-year estimates are not available Precision is more important than currency Analyzing very small populations Examining tracts and other smaller geographies because 1-year estimates are not available SOURCE: http://www.census.gov/acs/www/guidance_for_data_users/estimates/

9. ACS – Differences between ACS and CPS ASEC SOURCE: http://www.census.gov/hhes/www/poverty/about/datasources/factsheet.html American Community SurveyCurrent Population Survey Principal Purpose Part of the 2010 Decennial Program, providing annual (or multi- year average) estimates of selected social, economic, and housing characteristics of the population for many geographic areas and subpopulations. Produce specific socioeconomic and demographic estimates for the United States, and estimates for states for selected characteristics and subpopulations. Provide timely estimates of income and health insurance, as well as official poverty estimates. Geography Nation, states, and cities and counties of 65,000 or more. Areas as small as census tracts using multi-year averages. Nation, regions, and states for selected characteristics. Sample Size About 3 million addresses per year. Data are collected from about one-twelfth of the sample each month. Annual sample size is about 100,000 addresses. Data Collection Method Mail, telephone, and personal-visit interviews for the 50 states, the District of Columbia, and Puerto Rico. The ACS is a mandatory survey. Telephone and personal-visit interviews for the 50 states and the District of Columbia. The CPS is a voluntary survey. Population Universe The 2005 ACS included only the household population. This universe includes both the civilian and military population in households and excludes the group quarters population. The group quarters population consists of the institutionalized (such as people in correctional institutions or nursing homes) and the noninstitutionalized (most of whom are in college dormitories). The ACS began a group quarters data collection effort in 2006 and released the first total population estimates in 2007. The CPS includes the civilian noninstitutionalized population. This universe includes civilians in households, people in noninstitutionalized group quarters (other than military barracks) and military in households living off post or with their families on post (as long as at least one household member is a civilian adult). The universe excludes other military in households and in group quarters (barracks), and people living in institutions. Time Period Covered ACS collects data continuously throughout the year and the reference period for select questions (like income and the number of weeks worked) is the 12-month period prior to the response month. This yields a total time span covered by responses of 23 months. The survey’s 12-month estimates are centered on December 15, 2004. Monthly interviews conducted from February to April 2006 ask about calendar year 2005 income. The survey’s 12-month estimates are centered on July 1, 2005. Length and Detail of Questions Using a series of eight questions, the ACS asks about money income, plus one type of noncash benefit (food stamps) during the previous 12 months. CPS asks a series of questions about more than 50 sources of income, including questions about the amount of several noncash benefits such as food stamps and employment-related health insurance, during the previous calendar year.

10. ACS – PUMS DATASETS Sample comparison of median earnings for young adults ages 25 to 34 by state using ACS 1-year, ACS 3-year, and CPS (reference Excel handout) –U.S. total=$49,910 (ACS 1-year), $50,000 (ACS 3-year), $47,990 (CPS) –Estimates seem similar enough BUT cell sizes vary widely 177,364 records (ACS 1-year), 319,660 records (ACS 3-year), 16,325 records (CPS) SOURCE: 2008-2010 ACS 3-year PUMS Accuracy of the Data

11. Special Focus: Calculating the Median Income Elementary statistics: median is the value in the middle Not the same with Census sample data Steps in the sample algorithm: –Accumulate the weights of all records with the same income –Sum these accumulated totals until you exceed 50% of the sum of all weights –Interpolate between the associated incomes of prior and current subtotals to arrive at the approximated 50 th percentile SOURCE: 2008-2010 ACS 3-year PUMS Accuracy of the Data

12. Selected 2010 Findings on Employment and Median Annual Earnings for Young Adults Ages 25 to 34 with a Bachelor’s or Higher Degree SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

13. Bachelor’s or higher degree attainment by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

14. Bachelor’s or higher degree attainment by nativity status SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

15. Bachelor’s or higher degree attainment by citizenship status SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

16. Employment for bachelor’s degree holders by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

17. Employment for bachelor’s degree holders by nativity status SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

18. Employment for bachelor’s degree holders by citizenship status SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

19. Employment in STEM occupations for employed bachelor’s degree holders with a STEM BA by race/ethnicity and sex Any STEM occupation SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010. Race/ethnicity

20. Employment in STEM occupations for employed bachelor’s degree holders with a STEM BA by citizenship status Any STEM occupation SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

21. Employment in computer science occupations for employed bachelor’s degree holders with a STEM BA by race/ethnicity and sex Computer scientists SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

22. Employment in engineering and architecture occupations for employed bachelor’s degree holders with a STEM BA by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

23. Employment in scientist occupations for employed bachelor’s degree holders with a STEM BA by race/ethnicity, and sex Scientists SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

24. Employment in medical professional occupations for employed bachelor’s degree holders with a STEM BA by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

25. Employment in business and management occupations for employed bachelor’s degree holders with a STEM BA by race/ethnicity and sex Business workers/managers SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

26. Employment in education occupations for employed bachelor’s degree holders with a STEM BA by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

27. Employment in various occupations for employed bachelor’s degree holders with a STEM BA by citizenship status SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

28. Median earnings by highest level of educational attainment and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

29. Median earnings for computer science majors by race/ethnicity and sex Earnings SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010. Computer and information sciences

30. Median earnings for engineering majors by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

31. Median earnings for biology/biomedical science majors by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

32. Median earnings for mathematics/statistics majors by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

33. Median earnings for physical sciences majors by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

34. Median earnings for health/clinical science majors by race/ethnicity and sex SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010.

35. Median earnings by field of study and citizenship status Earnings Engineering/ engineering technologies Mathematics/ statistics Physical sciences Health professions/ clinical sciences SOURCE: U.S. Department of Commerce, Census Bureau, American Community Survey, 2010. Field of study

36. GAPS Report (EXPECTED August 2012 Release) Terris Ross, Ph.D. National Center for Education Statistics Terris.Ross@ed.gov 202.502.7302 http://nces.ed.gov/ For More Information & Additional Resources Uses of ACS Data Bill Sonnenberg National Center for Education Statistics William.Sonnenberg@ed.gov 202.502.7453 Questions about Presentation Grace Kena National Center for Education Statistics Grace.Kena@ed.gov 202.502.7888