Income Imputation and Child Support Outcomes

• In Event: Policies Affecting Separated Families Around the World

Abstract

The child support system is charged with assuring that parents who live apart from their children provide appropriate economic support, and determining the amount of child support to be ordered is key. While guidelines suggest order amounts tied to parents’ incomes, it is particularly challenging to determine the right amount of child support when reliable information about obligor income is unavailable, or judged not to be a good indicator of actual earnings potential. Recent Federal regulation requires courts to reconsider the practice of imputing income in such cases. In this paper we use unique data drawn from court records to examine the frequency of income imputation, and leverage differential imputation propensities across judges to evaluate the causal impact on child support outcomes.

In recent years policymakers, practitioners, and researchers have called for setting child support orders to align with obligors’ actual financial situations. Many states allow the court, or other decision-maker, to impute income when actual earnings are unknown. Previous research suggests that imputed income amounts often do not align with actual earnings, particularly for those with low incomes. This mismatch may lead to compliance issues, as large orders relative to actual income may discourage payment, while low orders may reduce payments and leave custodial parents with insufficient resources. Moreover, imputed income orders can also feel unfair to both parents, affecting their engagement with the child support system. In response to these concerns, the federal Office of Child Support Enforcement’s 2016 Final Rule included directives intended to limit the use of imputed orders. However, the causal impact of imputed income orders on child support outcomes remains underexplored, and is the focus of this study.

Our primary data source is the Wisconsin Court Record Database (CRD), which includes data from a sample of divorce and paternity cases from 21 of the 72 counties in Wisconsin. We focus on cases that came to court in selected periods between 2007 and 2019. We supplement these court records with additional administrative data from the Wisconsin Administrative Data Core (WADC), including demographic information, child support data from the Kids Information Data System (KIDS) (such as order and payment amounts), and earnings and employment history from the state Unemployment Insurance (UI) records.

In addition to documenting the prevalence of income imputation we examine the association between imputed-income and case characteristics and the relationship between income imputation and child support outcomes (i.e., amount ordered, likelihood of payment, amount paid, and ratio of paid to ordered amount). To address potential unobservable differences in cases with and without income imputation, we employ two-stage Instrumental Variable techniques using judge-level propensity of imputed-income orders to estimate the causal effect of imputed-income orders on child support outcomes. We find important heterogeneity in the types of cases with imputation, and the implications for order amounts, payments, and compliance. 

Authors

• Yoona Kim, University of Wisconsin-Madison
  Presenting Author

• Molly A Costanzo, University of Wisconsin-Madison
  Non-Presenting Co-Author