Survey misreporting is known to be pervasive and bias common statistical analyses. In this paper, I first use administrative data on SNAP receipt and amounts linked to American Community Survey data from New York State to show that survey data can misrepresent the program in important ways. For example, more than 1.4 billion dollars received are not reported in New York State alone. 46 percent of dollars received by house- holds with annual income above the poverty line are not reported in the survey data, while only 19 percent are missing below the poverty line. Standard corrections for measurement error cannot remove these biases. I then develop a method to obtain consistent estimates by combining parameter estimates from the linked data with publicly available data. This conditional density method recovers the correct estimates using public use data only, which solves the problem that access to linked administrative data is usually restricted. I examine the degree to which this approach can be used to extrapolate across time and geography, in order to solve the problem that validation data is often based on a convenience sample. I present evidence from within New York State that the extent of heterogeneity is small enough to make extrapolation work well across both time and geography. Extrapolation to the entire U.S. yields substantive differences to survey data and reduces deviations from official aggregates by a factor of 4 to 9 compared to survey aggregates.
-
MISCLASSIFICATION IN BINARY CHOICE MODELS
May 2013
Working Paper Number:
CES-13-27
We derive the asymptotic bias from misclassification of the dependent variable in binary choice models. Measurement error is necessarily non-classical in this case, which leads to bias in linear and non-linear models even if only the dependent variable is mismeasured. A Monte Carlo study and an application to food stamp receipt show that the bias formulas are useful to analyze the sensitivity of substantive conclusions, to interpret biased coefficients and imply features of the estimates that are robust to misclassification. Using administrative records linked to survey data as validation data, we examine estimators that are consistent under misclassification. They can improve estimates if their assumptions hold, but can aggravate the problem if the assumptions are invalid. The estimators differ
in their robustness to such violations, which can be improved by incorporating additional information. We propose tests for the presence and nature of misclassification that can help to choose an estimator.
View Full
Paper PDF
-
Within and Across County Variation in SNAP Misreporting: Evidence from Linked ACS and Administrative Records
July 2014
Working Paper Number:
carra-2014-05
This paper examines sub-state spatial and temporal variation in misreporting of participation in the Supplemental Nutrition Assistance Program (SNAP) using several years of the American Community Survey linked to SNAP administrative records from New York (2008-2010) and Texas (2006-2009). I calculate county false-negative (FN) and false-positive (FP) rates for each year of observation and find that, within a given state and year, there is substantial heterogeneity in FN rates across counties. In addition, I find evidence that FN rates (but not FP rates) persist over time within counties. This persistence in FN rates is strongest among more populous counties, suggesting that when noise from sampling variation is not an issue, some counties have consistently high FN rates while others have consistently low FN rates. This finding is important for understanding how misreporting might bias estimates of sub-state SNAP participation rates, changes in those participation rates, and effects of program participation. This presentation was given at the CARRA Seminar, June 27, 2013
View Full
Paper PDF
-
Using Linked Survey and Administrative Data to Better Measure Income: Implications for Poverty, Program Effectiveness and Holes in the Safety Net
October 2015
Working Paper Number:
CES-15-35
We examine the consequences of underreporting of transfer programs in household survey data for several prototypical analyses of low-income populations. We focus on the Current Population Survey (CPS), the source of official poverty and inequality statistics, but provide evidence that our qualitative conclusions are likely to apply to other surveys. We link administrative data for food stamps, TANF, General Assistance, and subsidized housing from New York State to the CPS at the individual level. Program receipt in the CPS is missed for over one-third of housing assistance recipients, 40 percent of food stamp recipients and 60 percent of TANF and General Assistance recipients. Dollars of benefits are also undercounted for reporting recipients, particularly for TANF, General Assistance and housing assistance. We find that the survey data sharply understate the income of poor households, as conjectured in past work by one of the authors. Underreporting in the survey data also greatly understates the effects of anti-poverty programs and changes our understanding of program targeting, often making it seem that welfare programs are less targeted to both the very poorest and middle income households than they are. Using the combined data rather than survey data alone, the poverty reducing effect of all programs together is nearly doubled while the effect of housing assistance is tripled. We also re-examine the coverage of the safety net, specifically the share of people without work or program receipt. Using the administrative measures of program receipt rather than the survey ones often reduces the share of single mothers falling through the safety net by one-half or more.
View Full
Paper PDF
-
Errors in Survey Reporting and Imputation and Their Effects on Estimates of Food Stamp Program Participation
April 2011
Working Paper Number:
CES-11-14
Benefit receipt in major household surveys is often underreported. This misreporting leads to biased estimates of the economic circumstances of disadvantaged populations, program takeup, and the distributional effects of government programs, and other program effects. We use administrative data on Food Stamp Program (FSP) participation matched to American Community Survey (ACS) and Current Population Survey (CPS) household data. We show that nearly thirty-five percent of true recipient households do not report receipt in the ACS and fifty percent do not report receipt in the CPS. Misreporting, both false negatives and false positives, varies with individual characteristics, leading to complicated biases in FSP analyses. We then directly examine the determinants of program receipt using our combined administrative and survey data. The combined data allow us to examine accurate participation using individual characteristics missing in administrative data. Our results differ from conventional estimates using only survey data, as such estimates understate participation by single parents, non-whites, low income households, and other groups. To evaluate the use of Census Bureau imputed ACS and CPS data, we also examine whether our estimates using survey data alone are closer to those using the accurate combined data when imputed survey observations are excluded. Interestingly, excluding the imputed observations leads to worse ACS estimates, but has less effect on the CPS estimates.
View Full
Paper PDF
-
BIAS IN FOOD STAMPS PARTICIPATION ESTIMATES IN THE PRESENCE OF MISREPORTING ERROR
March 2013
Working Paper Number:
CES-13-13
This paper focuses on how survey misreporting of food stamp receipt can bias demographic estimation of program participation. Food stamps is a federally funded program which subsidizes the nutrition of low-income households. In order to improve the reach of this program, studies on how program participation varies by demographic groups have been conducted using census data. Census data are subject to a lot of misreporting error, both underreporting and over-reporting, which can bias the estimates. The impact of misreporting error on estimate bias is examined by calculating food stamp participation rates, misreporting rates, and bias for select household characteristics (covariates).
View Full
Paper PDF
-
Estimation and Inference in Regression Discontinuity Designs with Clustered Sampling
August 2015
Working Paper Number:
carra-2015-06
Regression Discontinuity (RD) designs have become popular in empirical studies due to their attractive properties for estimating causal effects under transparent assumptions. Nonetheless, most popular procedures assume i.i.d. data, which is not reasonable in many common applications. To relax this assumption, we derive the properties of traditional non-parametric estimators in a setting that incorporates potential clustering at the level of the running variable, and propose an accompanying optimal-MSE bandwidth selection rule. Simulation results demonstrate that falsely assuming data are i.i.d. when selecting the bandwidth may lead to the choice of bandwidths that are too small relative to the optimal-MSE bandwidth. Last, we apply our procedure using person-level microdata that exhibits clustering at the census tract level to analyze the impact of the Low-Income Housing Tax Credit program on neighborhood characteristics and low-income housing supply.
View Full
Paper PDF
-
Labor Market Effects of the Affordable Care Act: Evidence from a Tax Notch
July 2017
Working Paper Number:
carra-2017-07
States that declined to raise their Medicaid income eligibility cutoffs to 138 percent of the federal poverty level (FPL) under the Affordable Care Act (ACA) created a "coverage gap'' between their existing, often much lower Medicaid eligibility cutoffs and the FPL, the lowest level of income at which the ACA provides refundable, advanceable "premium tax credits'' to subsidize the purchase of private insurance. Lacking access to any form of subsidized health insurance, residents of those states with income in that range face a strong incentive, in the form of a large, discrete increase in post-tax income (i.e. an upward notch) at the FPL, to increase their earnings and obtain the premium tax credit. We investigate the extent to which they respond to that incentive. Using the universe of tax returns, we document excess mass, or bunching, in the income distribution surrounding this notch. Consistent with Saez (2010), we find that bunching occurs only among filers with self-employment income. Specifically, filers without children and married filers with three or fewer children exhibit significant bunching. Analysis of tax data linked to labor supply measures from the American Community Survey, however, suggests that this bunching likely reflects a change in reported income rather than a change in true labor supply. We find no evidence that wage and salary workers adjust their labor supply in response to increased availability of directly purchased health insurance.
View Full
Paper PDF
-
The Work Disincentive Effects of the Disability Insurance Program in the 1990s
February 2006
Working Paper Number:
CES-06-05
In this paper we evaluate the work disincentive effects of the Disability Insurance program during the 1990s. To accomplish this we construct a new large data set with detailed information on DI application and award decisions and use two different econometric evaluation methods. First, we apply a comparison group approach proposed by John Bound to estimate an upper bound for the work disincentive effect of the current DI program. Second, we adopt a Regression-Discontinuity approach that exploits a particular feature of the DI eligibility determination process to provide a credible point estimate of the impact of the DI program on labor supply for an important subset of DI applicants. Our estimates indicate that during the 1990s the labor force participation rate of DI beneficiaries would have been at most 20 percentage points higher had none received benefits. In addition, we find even smaller labor supply responses for the subset of 'marginal' applicants whose disability determination is based on vocational factors.
View Full
Paper PDF
-
An Economist's Primer on Survey Samples
September 2000
Working Paper Number:
CES-00-15
Survey data underlie most empirical work in economics, yet economists typically have little familiarity with survey sample design and its effects on inference. This paper describes how sample designs depart from the simple random sampling model implicit in most econometrics textbooks, points out where the effects of this departure are likely to be greatest, and describes the relationship between design-based estimators developed by survey statisticians and related econometric methods for regression. Its intent is to provide empirical economists with enough background in survey methods to make informed use of design-based estimators. It emphasizes surveys of households (the source of most public-use files), but also considers how surveys of businesses differ. Examples from the National Longitudinal Survey of Youth of 1979 and the Current Population Survey illustrate practical aspects of design-based estimation.
View Full
Paper PDF
-
The Distributional Effects of Minimum Wages: Evidence from Linked Survey and Administrative Data
March 2018
Working Paper Number:
carra-2018-02
States and localities are increasingly experimenting with higher minimum wages in response to rising income inequality and stagnant economic mobility, but commonly used public datasets offer limited opportunities to evaluate the extent to which such changes affect earnings growth. We use administrative earnings data from the Social Security Administration linked to the Current Population Survey to overcome important limitations of public data and estimate effects of the minimum wage on growth incidence curves and income mobility profiles, providing insight into how cross-sectional effects of the minimum wage on earnings persist over time. Under both approaches, we find that raising the minimum wage increases earnings growth at the bottom of the distribution, and those effects persist and indeed grow in magnitude over several years. This finding is robust to a variety of specifications, including alternatives commonly used in the literature on employment effects of the minimum wage. Instrumental variables and subsample analyses indicate that geographic mobility likely contributes to the effects we identify. Extrapolating from our estimates suggests that a minimum wage increase comparable in magnitude to the increase experienced in Seattle between 2013 and 2016 would have blunted some, but not nearly all, of the worst income losses suffered at the bottom of the income distribution during the Great Recession.
View Full
Paper PDF