Trouble in the Tails? What We Know about Earnings Nonresponse 30 Years after Lillard, Smith, and Welch

Abstract

Earnings nonresponse in household surveys is widespread, yet there is limited knowledge of how nonresponse biases earnings measures. We examine the consequences of nonresponse on earnings gaps and inequality using Current Population Survey individual records linked to administrative earnings data. The common assumption that earnings are missing at random is rejected. Nonresponse across the earnings distribution is U-shaped, highest in the left and right tails. Inequality measures differ between household and administrative data due in part to nonresponse. Nonresponse biases earnings differentials by race, gender, and education, particularly in the tails. Flexible copula-based models can account for nonrandom nonresponse.

kmr: A command to correct survey weights for unit nonresponse using groups’ response rates

“…Now, we try a different specification that adds the squared log of income per capita as a second regressor. This will help to capture the fact that high nonresponse rates may occur in both tails of the income distribution, not just among rich households, as documented in Bollinger et al (2019).…”

“…There is evidence that household income systematically affects survey response. Using the current population survey (CPS) of the United States, Bollinger et al (2019) show that nonresponse increases in the tails of the income distribution. This empirical evidence rejects the ignorability assumption (the fact that nonresponse is random within some arbitrary subgroup of the population).…”

Donative labor effect of the nonprofit pay: A multilevel explanation

“…It provides comprehensive information on individuals, industries, and occupations. Individuals with missing and imputed data on all relevant variables are dropped, as Census Bureau's imputation process does not differentiate the sector pay, which might level off the wage differentials between the two sectors (Bollinger, Hirsch, Hokayem, & Ziliak, 2019). I restricted to full-time workers aged 16-65 years (Hirsch et al, 2018) who work 35 hours or above per week and 50-52 weeks per year.…”

REDI for Binned Data: A Random Empirical Distribution Imputation method for estimating continuous incomes

“…Therefore, I treat top-coded income categories like all others. This approach might be further improved with the enrichment of top incomes in the reference dataset to better reflect the true distribution of high incomes (Fixler et al 2019), or adjustments to reflect nonresponse bias in the upper and lower tales of the income distribution (Bollinger et al 2019). Finally, I model artificial multivariate regressions using both the original continuous values from the reference dataset and the REDI-calculated income values.…”