Skip Main Navigation Skip To Navigation Content

Labor Force Statistics






Changes to the American Community Survey between 2007 and 2008 and the Effect on the Estimates of Employment and Unemployment

Background

For the 2008 American Community Survey (ACS), changes were made to the questionnaire that modified and improved existing questions for several subject areas.1 In particular, an improved series of labor force questions was introduced to better capture data on employment status. Prior research and analysis of employment data from the ACS and Census 2000 revealed that employment levels were underestimated and unemployment levels were overestimated relative to benchmark data from the Current Population Survey (CPS) or from the Local Area Unemployment Statistics (LAUS) program at the Bureau of Labor Statistics. The CPS and LAUS are used to produce the official estimates of employment, unemployment, and the unemployment rate for the nation and states. We believe the modifications and improvements to the existing series of labor force questions had the effect of increasing the number of employed persons captured in the 2008 ACS data relative to CPS and LAUS estimates, and as a result, contributed to the lack of significant change in ACS unemployment rates between 2007 and 2008.

2008 Questionnaire Changes to the Labor Force Question Series

In January through March of 2006, the ACS conducted the first test of new and modified content since the ACS reached full implementation levels of data collection, hereafter referred to as the 2006 ACS Content Test. The evaluation and results of this test helped to determine the content for the 2008 ACS.

The primary objective of the 2006 ACS Content Test work on the employment status question series, specifically, the questions about worked last week, temporarily absent, and looking for work, was to improve the measurement of employment status by addressing several limitations that previous research suggested were present in the ACS question wording prior to 2008. Past research and analysis of employment data from the ACS and Census 2000 revealed that employment levels were underestimated and unemployment levels were overestimated relative to benchmark data from the CPS or LAUS program at the Bureau of Labor Statistics. For more information, see the reports “Census 2000 Evaluation B.7 – Accuracy of Data for Employment Status as Measured by the CPS-Census 2000 Match" at: http://www.census.gov/pred/www/rpts/B.7%20Final%20Report.pdf and “Census 2000 Auxiliary Evaluation B.8 – Comparing Employment, Income, and Poverty: Census 2000 and the Current Population Survey" at http://www.census.gov/hhes/www/laborfor/final2_b8_nov6.pdf.

Three of the ACS employment status questions were modified for the test panel (Person Questions 28a-28b, 34b, and 35). The worked last week (28a-28b) and temporarily absent (34b) questions are key components in the measurement of employed people and people who are not in the labor force, while the looking for work question (35) is a component in the measurement of unemployed people. These changes were done with an overall goal of better matching CPS labor force estimates by increasing the estimate of employed people, reducing the estimate of unemployed people, and reducing response inconsistencies in the individual categories of the employment status concept.

The 2006 ACS Content Test compared two versions of the employment status series question set. The control version replicated the 2006 ACS questions.2 The test version modified the employment status series questions by

The 2006 ACS Content Test findings overall showed that the test questions produced a higher estimate of employed people compared to the control. While the test questions did not produce a lower estimate of unemployed people, the overall unemployment rate was lower for the test panel. The unemployment rate is the more useful measure because it excludes people who are not in the labor force. Empirically, the test version of the questions performed better than the existing ACS questions in terms of more favorable estimates. For more information, see the report “2006 American Community Survey Content Test Report P.6.a. – Evaluation Report Covering Employment Status" at: http://www.census.gov/acs/www/AdvMeth/content_test/P6a_Employment_Status.pdf.

Effect on 2008 Employment and Unemployment Estimates

Tables 1 and 2 present ACS and CPS/LAUS employment status data for the U.S. civilian noninstitutionalized population by state for 2007 and 2008. One notable difference between the ACS and CPS survey universes relates to the group quarters (GQ) population. The ACS includes both institutionalized (e.g., prisons, residential treatment centers, and nursing facilities) and noninstitutionalized (e.g., college/university housing and military quarters) GQ in its survey universe, while the CPS includes only noninstitutionalized GQ. This survey sample difference affects total population counts and, likewise, employment-population ratio comparisons. However, it does not impact employment and unemployment estimates since those in the institutionalized GQ population are not considered part of the labor force. To provide data users with additional context regarding the impact of the labor force question changes, employment-population ratios are presented in this note. Furthermore, for comparison purposes, all individuals living in institutionalized group quarters are excluded from the ACS estimates, and as a result, ACS estimates presented in this note may not match published ACS data on American Fact Finder.

Table 1 presents ACS and CPS/LAUS employment-population ratios for the U.S. civilian noninstitutionalized population by state for 2007 and 2008.3 In 2008, the ACS and CPS U.S. employment-population ratios were 62.7 and 62.2, respectively.4 Between 2007 and 2008, the ACS U.S. employment-population ratio increased from 61.6 to 62.7, while for the CPS, the U.S. employment-population ratio decreased 0.8 percentage points, from 63.0 to 62.2. The difference between the ACS and CPS U.S. employment-population ratios was 1.3 percentage points in 2007 and 0.5 percentage points in 2008.

For the ACS between 2007 and 2008, 13 states experienced no significant change in their employment-population ratio, while 37 states and the District of Columbia experienced an increase and no states experienced a decrease in their employment-population ratio. For the LAUS between 2007 and 2008, 40 states and the District of Columbia experienced no significant change in their employment-population ratio, while no states experienced an increase and 10 states experienced a decrease in their employment-population ratio. Twenty-eight states and the District of Columbia had ACS and LAUS employment-population ratios that were statistically different from each other in 2007, while the number of statistically different states decreased to 11 in 2008.

Table 2 presents ACS and CPS/LAUS unemployment rates for the U.S. civilian noninstitutionalized population by state for 2007 and 2008.5 In 2008, the ACS and CPS U.S. unemployment rates were 6.4 percent and 5.8 percent, respectively. Between 2007 and 2008, the ACS U.S. unemployment rate increased 0.1 percentage points, from 6.3 percent to 6.4 percent, while for the CPS, the U.S. unemployment rate increased 1.2 percentage points, from 4.6 percent to 5.8 percent. The difference between the ACS and CPS U.S. unemployment rates was 1.7 percentage points in 2007 and 0.6 percentage points in 2008.

For the ACS between 2007 and 2008, 32 states and the District of Columbia had no statistically significant change in their unemployment rate, while six states (California, Delaware, Florida, Nevada, Rhode Island, and South Carolina) experienced an increase and twelve states experienced a decrease in their unemployment rate. For the LAUS between 2007 and 2008, 15 states experienced no significant change in their unemployment rate, while 35 states and the District of Columbia experienced an increase and no states experienced a decrease. Forty-nine states and the District of Columbia had ACS and LAUS unemployment rates that were statistically different from each other in 2007, while the number of statistically different states decreased to 26 in 2008.

Summary

The changes to the employment status series of questions in the 2008 ACS will make ACS labor force data more consistent with benchmark data from the CPS and the LAUS program. The CPS and LAUS are used to produce the official estimates of employment, unemployment, and the unemployment rate for the nation and states. We believe the modifications and improvements to the existing series of labor force questions had the effect of increasing the number of employed persons captured in the 2008 ACS data relative to CPS and LAUS estimates, and as a result, contributed to the lack of significant change in ACS unemployment rates between 2007 and 2008. Users should use caution in making comparisons between 2008 and earlier years.

Even though the 2008 ACS data did not capture as many significant increases in state unemployment rates between 2007 and 2008 as the LAUS data, the 2008 ACS did capture an increase in unemployment rates for those states (California, Florida, and Nevada) most affected by the economic slowdown that occurred over the course of 2008.6 Given this and the decreased prevalence of statistical differences between ACS and CPS/LAUS employment-population ratio and unemployment rate estimates, we believe that the 2008 ACS data represents an improvement in the estimates of employment and unemployment data compared to prior ACS data.

Footnotes:

  1. A summary of the changes made to the 2008 ACS can be found at http://www.census.gov/acs/www/AdvMeth/content_test/summary_results.htm. The 2007 and 2008 ACS questionnaires are available at http://www.census.gov/acs/www/Downloads/SQuest07.pdf and http://www.census.gov/acs/www/Downloads/SQuest08.pdf, respectively.
  2. The 2006 ACS questionnaire is available at http://www.census.gov/acs/www/Downloads/SQuest05.pdf.
  3. The employment-population ratio is the percentage of all working-age civilians who are employed. It is important to note that while both surveys collect data on a monthly basis, ACS data is reported annually while CPS data is reported monthly. To address this disparity, CPS estimates presented in this note (i.e., the total U.S. estimates) are based on an average of the monthly employment status data across the 12 months of 2007 and 2008. The 2007 and 2008 LAUS estimates are obtained from published Bureau of Labor Statistics data that can be found at http://www.bls.gov/lau/.
  4. All statistical comparisons are made at the 90 percent confidence level.
  5. The labor force is the sum of employed and unemployed persons. The unemployment rate is the number of unemployed persons divided by the labor force.
  6. The Business Cycle Dating Committee of the National Bureau of Economic Research (NBER) maintains a chronology of the beginning and ending dates (months and quarters) of U.S. recessions. The NBER was founded in 1920, and published its first business cycle dates in 1929. Please see http://www.nber.org/cycles/dec2008.html for more information concerning the dating of the most recent U.S. recession.

Download complete Research Note report with tables. [PDF - 1.2M]


[PDF] or PDF denotes a file in Adobe’s Portable Document Format. To view the file, you will need the Adobe® Reader® Off Site available free from Adobe. [Excel] or the letters [xls] indicate a document is in the Microsoft® Excel® Spreadsheet Format (XLS). To view the file, you will need the Microsoft® Excel® Viewer Off Site available for free from Microsoft®.
Contact the Demographic Call Center Staff at 301-763-2422 or 1-866-758-1060 (toll free) or visit ask.census.gov for further information on Labor Force Data.

Source: U.S. Census Bureau, Housing and Household Economic Statistics Division