Unemployment Compensation Supplement Sample Notes
The Unemployment Compensation Supplement contains information on respondents' participation, or lack of participation, in US Government programs for persons unemployed for reasons beyond their control. The terms “unemployment compensation,” “unemployment benefits,” and “unemployment insurance” are used interchangeably in the original Census Bureau documentation. Most recently, the supplement is referred to as the "Unemployment Insurance Nonfilers Supplement." IPUMS CPS unites all of these concepts under the term “unemployment compensation.” The supplement includes questions about whether a respondent has applied for unemployment compensation, whether they have received the compensation, or if they have not applied for or received compensation, reasons that compensation was not sought or received.
The Unemployment Compensation Supplement was first fielded from 1989 to 1990, and again in 2005 and 2018. Due to small sample sizes, the Unemployment Compensation Supplement is asked in multiple months in a single year (or pair of years, in the case of 1989 and 1990), with the results intended to be used together as one dataset. To correctly use Unemployment Compensation data, all months including Unemployment Compensation supplement data from a given year (or pair of years, in the case of 1989 and 1990) must be included in an extract in order to replicate the Census Bureau sampling. UCSUPPWT, the Unemployment Compensation weight variable, must be used with pooled data rather than individual months to yield correct tallies. Weighted totals for the Unemployment Compensation Supplement are larger than expected in 1989-90 and 2005. We have been unable to explain this discrepancy.
The universe of the Unemployment Compensation Supplement has changed over time. The supplement universes found in the original documentation are not reflective of universes found in the data empirically. For all Unemployment Compensation Supplements and supplement variables IPUMS CPS uses the empirically determined universe that best fits the data.
In 1989 and 1990, the universe is civilians age 15 or older in month-in-samples (MISH) 4 or 8 who are unemployed and have ever worked.
The 2005 supplement universe is civilians age 16 or older in month-in-samples 4 or 8 who are either on layoff, unemployed and looking but have worked in the past, or not in the labor force for reasons other than retirement or disability and have worked in the past 12 months.
In 2018 and 2022 the universe is civilians age 16 or older who were on layoff, unemployed and looking for work, or not in the labor force and had not stated they did no work in the past 12 months.
While these empirically derived universes reflect the data well, a small number of records in 2005 and 2018 do not exactly reflect supplement response criteria.These responses have been left unmodified in IPUMS CPS data. In all Unemployment Compensation supplements from these years there are some respondents who either do not meet supplement universe criteria but have responses, or meet all supplement universe criteria but have no responses and are designated “not in universe” for the supplement response flag UCSUPFLG. The following table shows the number of such responses in each supplement.
| Year | Month | Met criteria but have no responses | Did not meet criteria but have responses |
|---|---|---|---|
| 2005 | January | 149 | 20 |
| 2005 | May | 141 | 16 |
| 2005 | July | 135 | 21 |
| 2005 | November | 126 | 15 |
| 2018 | May | 2 | 336 |
| 2018 | September | 1 | 329 |
| 2022 | February | -- | -- |
| 2022 | May | -- | -- |
Supported By
