Match Bias from Earnings Imputation in the Current Population Survey: The Case of Imperfect Matching

By C. Bollinger, B. Hirsch. Journal of Labor Economics, 2006

194 citations recorded by the source corpus.

This article examines match bias arising from earnings imputation. Wage equation parameters are estimated from mixed samples of workers reporting and not reporting earnings, the latter assigned earnings of donors. Regressions including attributes not used as imputation match criteria (e.g., union) are severely biased. Match bias also arises with attributes used as match criteria but matched imperfectly. Imperfect matching on schooling (age) flattens earnings profiles within education (age) groups and creates jumps across groups. Assuming conditional missing at random, a general analytic expression correcting match bias is derived and compared to alternatives. Reweighting a respondent‐only sample proves an attractive approach.

Subfields: Economics and Econometrics, Sociology and Political Science

View Paper: https://www.semanticscholar.org/paper/ffe3f4ba00fe1c371e76926e932e0dba33288cce

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Sociology and Political Science

Citations in the corpus

The Benefits of Attending Community College: A Review of the Evidence (2011, background): “Using North Carolina data, Bowen et al. (2009) found that of all students who were “presumptively eligible” to attend a selective college (i.e., their grade point averages or Scholastic Aptitude Test scores were high enough), 40% did not (p. 102). Moreover, these “under-matched” students had lower g”
A Review of Hot Deck Imputation for Survey Non‐response (2010, methodology): “An example is imputation of income in the Current Population Survey Outgoing Rotation Group (CPS-ORG), which uses seven variables leading to 11,520 adjustment cells (Bollinger & Hirsch, 2006), some of which contain non-respondents but no matching respondents.”
Decomposing Trends in Income Volatility: The 'Wild Ride' at the Top and Bottom (2014, background): “Bollinger and Hirsch (2006) recommend that imputed data should be dropped owing to potential bias, and many papers in the inequality literature drop these observations (Autor, Katz, and Kearney 2008; Lemieux 2006).”
The Medium-Term Labor Market Returns to Community College Awards: Evidence From North Carolina (2014, background): “Unlike sample survey data, UI earnings data are not affected by biases due to imputation, self-reporting, and nonresponse (Bollinger & Hirsch, 2006; Schmitt & Baker, 2006).”
Household Surveys in Crisis (2015, background): “Nonresponse rates are typically low for most questions (Bollinger and Hirsch, 2006), but they can be quite high for questions related to labor and nonlabor income.”
Economic Analysis and Statistical Disclosure Limitation (2016, methodology): “Our analysis draws on the work of Hirsch and Schumacher (2004) and Bollinger and Hirsch (2006), who study the closely related problem of bias from missing data imputation in the CPS. Respondents to the CPS commonly fail to provide answers to certain questions.”
Earnings volatility in America: Evidence from matched CPS (2011, methodology): “We also present a detailed decomposition of earnings volatility to account for changing shares in employment probabilities, conditional variances of continuous workers, and conditional mean variances from employment entry and exit.”
Financial literacy and its consequences: Evidence from Russia during the financial crisis (2013, methodology): “…As an alternative to the imputation and bootstrapping, and based on the analysis of ‘‘match bias’’ during imputation in Hirsch and Schumacher (2004) and Bollinger and Hirsch (2006) for the US Current Population Survey (CPS), we have also conducted the analysis dropping the observations with…”