1. Since 2003, federal surveys have allowed respondents to report more than one race. Therefore, two basic ways of defining a race group are possible. A group, such as Asian, may be defined as those who reported Asian and no other race (the race-alone or single-race concept) or as those who reported Asian regardless of whether they also reported another race (the race-alone-or-in-combination concept). This table shows data using both approaches. Since Hispanic individuals may be any race, data for the Hispanic population overlap with data for race groups. Data users should exercise caution when interpreting aggregate results for the Hispanic population or for race groups because these populations consist of many distinct groups that differ in socioeconomic characteristics, culture, and nativity. Data on Hispanic origin were first collected in 1972. Data on Asian and Pacific Islander origin and American Indian and Alaska Native origin were first collected in 1987. Estimates for Native Hawaiians and Other Pacific Islanders are not shown separately due to sample size.

2. Implementation of 2020 Census-based population controls.

3. Estimates reflect the implementation of an updated processing system and should be used to make comparisons to 2018 and subsequent years.

4. The 2014 CPS ASEC included redesigned questions for income and health insurance coverage. All of the approximately 98,000 addresses were eligible to receive the redesigned set of health insurance coverage questions. The redesigned income questions were implemented to a subsample of the 98,000 addresses using a probability split panel design. Approximately 68,000 addresses were eligible to receive a set of income questions similar to those used in the 2013 CPS ASEC, and the remaining 30,000 addresses were eligible to receive the redesigned income questions. The source of these 2013 estimates is the portion of the CPS ASEC sample that received the redesigned income questions, approximately 30,000 addresses.

5. The source of these 2013 estimates is the portion of the CPS ASEC sample that received the income questions consistent with the 2013 CPS ASEC, approximately 68,000 addresses.

6. Implementation of 2010 Census-based population controls.

7. Data have been revised to reflect a correction to the weights in the 2005 CPS ASEC.

8. Implementation of a 28,000-household expansion.

9. Implementation of 2000 Census-based population controls.

10. Full implementation of 1990 Census-based sample design and metropolitan definitions, 7,000-household sample reduction, and revised editing of responses on race.

11. Introduction of 1990 Census-based sample design.

12. Data collection method changed from paper and pencil to computer-assisted interviewing. In addition, the 1994 CPS ASEC was revised to allow for the coding of different income amounts on selected questionnaire items. Limits either increased or decreased in the following categories: earnings limits increased to $999,999; Social Security limits increased to $49,999; Supplemental Security Income and public assistance limits increased to $24,999; veterans' benefits limits increased to $99,999; child support and alimony limits decreased to $49,999.

13. Implementation of 1990 Census-based population controls.

14. Estimates are revised to correct for nine omitted weights from the original 1992 CPS ASEC. See "Money Income of Households, Families, and Persons in the United States: 1992" P60-184.

15. Estimates reflect the implementation of a new CPS ASEC processing system and are also revised to reflect corrections to the files after publication of the 1988 advance report "Money Income and Poverty Status in the United States: 1988" P60-166.

16. Full implementation of 1980 Census-based sample design.

17. Implementation of Hispanic population weighting controls and introduction of 1980 Census-based sample design.

18. Implemented three technical changes to the poverty definition. See "Characteristics of the Population Below the Poverty Level: 1980" P60-133.

19. Implementation of 1980 Census-based population controls. Questionnaire expanded to show 27 possible values from 51 possible sources of income.

20. Implementation of a new CPS ASEC processing system. Questionnaire expanded to ask 11 income questions.

21. Full implementation of 1970 Census-based sample design.

22. Introduction of 1970 Census-based sample design and population controls.

23. Implementation of a new CPS ASEC processing system.

24. Beginning in January 1978, the Bureau of Labor Statistics introduced a new price index for all urban consumers (CPI-U) that forms a continuous series with the earlier index for urban wage earners and for clerical workers as of December 1997.

25. The "Outside metropolitan statistical areas" category includes both micropolitan statistical areas and territory outside of metropolitan and micropolitan statistical areas. For more information, see "About Metropolitan and Micropolitan Statistical Areas" at www.census.gov/programs-surveys/metro-micro/about/glossary.html

26. The 2015 and 2016 estimates presented for residence may not match the previously published estimates due to a correction in the assignments of principal city status for a small number of households.

27. In 2005, the CPS ASEC sample transitioned from a Census 1990-based sample design to a Census 2000-based sample design, which used different definitions of metropolitan areas. As a result, estimates for 2004 that rely on metropolitan definitions are not directly comparable with either the Census 1990 or Census 2000 definitions and are thus suppressed.

28. Work experience: Refers to the longest job held in the previous calendar year. The work experience categories are based on the number of weeks worked, and the number of hours worked per week. Full-time year-round: Worked at least 35 hours per week, for at least 50 weeks last year (including paid sick leave and vacations). Not full-time year-round: Worked for at least 1 week last year, but for less than 50 weeks, or less than 35 hours per week, or both.

29. A margin of error (MOE) is a measure of an estimate’s variability. The larger the MOE in relation to the size of the estimate, the less reliable the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in this table for years through 2003 are based on standard errors calculated using generalized variance function (GVF) parameters. MOEs for 2004 and beyond are based on standard errors calculated using replicate weights.

30. Beginning in 2003, CPS ASEC offered respondents the option of choosing more than one race. The 2002 and 2001 CPS ASEC recorded only one race for each respondent. The 3-year averages for 2002 are based on combining the 2003 CPS ASEC race categories shown in the stub with the relevant single race categories of White, Black, American Indian or Alaska Native, or Asian and Pacific Islander recorded in the 2002 and 2001 CPS ASEC.

31. These estimates were calculated using the 2020 production file, the 2019 production file, and the 2018 Bridge file.

32. These estimates were calculated using the 2019 production file, the 2018 Bridge file, and the 2017 Research file (which was not publicly released).

33. Estimates are unavailable due to a questionnaire change. Footnotes 4 and 5 provide more information.