Toward a 21st Century National Data Infrastructure: Enhancing Survey Programs by Using Multiple Data Sources (2023) / Chapter Skim
Currently Skimming:
5 Data Linkage to Improve Income Measurement
5 Data Linkage to Improve Income Measurement
Pages 105-124
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 105... ...
There is also a large literature in which researchers have used linked data to study issues of policy concern, such as poverty disparities, income inequality, earnings volatility, and effects of tax policies. This chapter, relying in part on presentations in the workshop ses sion Data Linkage for Income and Health Statistics, focuses on using record linkage to improve measurement and understanding of income and related concepts.
|
From page 106... ...
Census Bureau projects -- the Comprehensive Income Dataset project, which examines the effect of improved income measures on poverty estimates, and the National Experimental Well-being Statistics project, which proposes to blend administrative and survey data to create more accurate income estimates. 5.1 INCOME DATA COLLECTION ON SURVEYS Numerous federal statistical agency surveys collect data about income and poverty.
|
From page 107... ...
Current Population Survey Annual Social and Economic Supplement As discussed in Chapter 2, the CPS originated in a program established in 1940 to provide direct measurement of national unemployment each month, on a sample basis. The monthly CPS sample still serves that purpose but is periodically supplemented by additional data collections that concentrate on specific aspects of the nation's social or economic well-being.2 The oldest of these supplements, established in 1947, is now known as the Annual Social and Economic Supplement (ASEC)
|
From page 108... ...
108 ENHANCING SURVEY PROGRAMS BY USING MULTIPLE DATA SOURCES FIGURE 5-1 American Community Survey income questions, 2022. SOURCE: https://www2.census.gov/programs-surveys/acs/methodology/questionnaires/ 2022/quest22.pdf
|
From page 109... ...
. The SIPP captures aspects of income and program participation not measured by the ACS or the CPS ASEC, such as changes in household composition, periods of program participation, and detailed data on assets and liabilities (which play a role in determining program eligibility)
|
From page 110... ...
The surveys also ask about topics that would not be available from administrative records on income. For example, the ACS asks about education and disability; the CPS ASEC asks about health insurance and child care expenses; the SIPP asks about unpaid time away from work and adult and child well-being.
|
From page 111... ...
The CPS ASEC and SIPP have higher unit nonresponse than the ACS,6 and also have item nonresponse for income questions (Hokayem, Bollinger, & Ziliak, 2015)
|
From page 112... ...
The next section describes administrative data sources that might be used to supplement or study properties of income surveys. 5.2 ADMINISTRATIVE RECORDS SOURCES FOR INCOME DATA Many administrative data sources collect information about income.
|
From page 113... ...
82) , studying item nonresponse, found that "there are clear efficiency gains from using administrative data" in CPS ASEC imputation models.
|
From page 114... ...
. 5.3 USING ADMINISTRATIVE DATA WITH INCOME SURVEYS Administrative records have the potential to address some of the shortcomings of survey data, including small sample sizes that limit the groups for which disaggregated statistics can be produced, inaccuracies in reporting income, and possible bias from misreporting and nonresponse.
|
From page 115... ...
linked the 2011 CPS ASEC sampling frame to IRS records and found that survey respondents and nonrespondents differed in demographic characteristics such as marital status but had similar income distributions. Effects of item nonresponse in surveys can also be studied through linkage with administrative records.
|
From page 116... ...
We could then use this information to adjust the probability that a given individual is asked par ticular income questions. This could reduce respondent burden on those that are more likely to have administrative data, maintain the questions for those likely not to have administrative data, and preserve a sample of each group with survey responses for modeling and imputation.
|
From page 117... ...
, and persons in certain occupational categories. • Using linked CPS ASEC/SSA records from 1998–2009, Hokayem, Bollinger, and Ziliak (2015)
|
From page 118... ...
• By linking 2013 CPS ASEC records for persons aged 65 or older to administrative data records supplied by the SSA, Bee and Mitchell (2017) were able to examine discrepancies at the individual record level (previous studies had compared aggregated statistics from separate sources)
|
From page 119... ...
Self-employment income may be inaccurate in all sources. 5.5 USING LINKED INCOME DATA TO IMPROVE INCOME STATISTICS While many of the studies using linked income data from surveys and administrative records have addressed undercoverage, nonresponse, and reporting differences, other studies have looked at the effects of adjusting survey data on outcomes of interest, such as poverty or income
|
From page 120... ...
6) offered anticipated outcomes from the CID Project, "including improving the Census Bureau's household surveys, becoming a critical resource for policymakers to evaluate policies, programs and taxes, and offering better evidence for researchers investigating a diverse range of topics." Furthermore, linkage to Social Security and Supplemental Security Income files might permit actual benefit amounts to be substituted for the questions normally asked on income surveys, which could reduce respondent burden and increase accuracy.
|
From page 121... ...
, data sources for the NEWS Project include surveys such as the ACS and the CPS ASEC as well as the decennial census. The administrative data include IRS and SSA data, the U.S.
|
From page 122... ...
on SNAP. In addition, NEWS researchers are using linked administrative data to adjust CPS ASEC weights for unit nonresponse (Rothbaum & Bee, 2021)
|
From page 123... ...
capturing income for the bottom and top 10 percent of the income distribution where survey data is less useful due to more complex income payments."13 Income measurement remains an active area for survey and administrative data development. With the development of data linkages, there is great potential for improving federal statistics.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.