Income |
|
Changes in Methodology for the March Current Population Survey
(This chart is produced by the Income Surveys Branch as a convenience to our users. If you have any questions, please Call 301-763-3243)
Income Year | Change |
1947 | Data based on 1940 census population controls. |
1949 | Implementation of expanded income questions to show wage and salary, farm self-employment, nonfarm self-employment and all other nonearned income separately. |
1952 | Implementation of 1950 census population controls. |
1961 | Implementation of first hotdeck procedure to impute missing income entries (all income data imputed if any missing). Introduction of 1960 census-based sample design. |
1962 | Full implementation of 1960 census-based sample design and population controls. |
1965 | Implementation of new procedures to impute missing data only. |
1966 | Questionnaire expanded to ask eight income questions. |
1967 | Implementation of a new March CPS processing system. |
1971 | Introduction of 1970 census-based sample design and population controls. |
1972 | Full implementation of 1970 census-based sample design. |
1974 | Implementation of a new March CPS processing system. Questionnaire expanded to ask 11 income questions. |
1975 | These estimates were derived using Pareto interpolation and may differ from published data which were derived using linear interpolation. |
1976 | First year medians are derived using both Pareto and linear interpolation. Prior to this year all medians were derived using linear interpolation. |
1979 | Implementation of 1980 census population controls. Questionnaire expanded to show 27 possible values from 51 possible sources of income. |
1983 | Implementation of Hispanic population weighting controls and introduction of 1980 census-based sample design. |
1985 | Recording of amounts for earnings from longest job increased to $299,999. Full implementation of 1980 census-based sample design. |
1987 | Implementation of a new March CPS processing system. |
1992 | Implementation of 1990 census population controls. |
1993 | Data collection method changed from paper and pencil to computer-assisted interviewing. In addition, the March 1994 income supplement 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 increased to $999,999; social security increased to $49,999; supplemental security income and public assistance increased to $24,999; veterans' benefits increased to $99,999; child support and alimony decreased to $49,999. |
1994 | Introduction of 1990 census-based sample design. |
1995 | Full implementation of the 1990 census-based sample design and metropolitan definitions, 7,000 household sample reduction, and revised race edits. |
1999 | Starting in 1999, alternative income definition 7 includes federal EIC and EIC for the nine states that use federal eligibility rules to compute the state credit as a percentage of the federal EIC. The nine states are: Iowa, Kansas, Massachusetts, Maryland, New York, Oregon, Rhode Island, Vermont, and Wisconsin. Also starting in 1999, when looking at the quintiles in the historical income series, you will notice that the 50th percentile is based on micro-sorted data and may differ from the median published in the income report which is based on linearly interpolated grouped data. The Census Bureau started using the Bureau of Labor Statistics' CPI-U-RS series as an inflation factor in income year 2000. You may access information on the CPI-U-RS series and the differences between this series and the CPI-U-X1 series at /hhes/income/income01/cpiurstxt.html |
2000 | There are two versions of the 2000 income data available. One version is based on the traditional sample of about 50,000 households and reflects the use of 1990 census population controls. The second version is based on a sample of 78,000 households, reflecting a 28,000 household sample expansion and the use of Census 2000 population controls. Please check the table footnotes and headnotes to ascertain which data is being displayed. |