Can State Child Tax Credits Reduce Poverty Without Discouraging Work?

Unrath et al. (2026) asked whether unconditional state child tax credits can reduce child poverty without substantially reducing parental employment. They used American Community Survey data, Supplemental Poverty Measure data, tax simulations, and labor supply estimates from recent state child tax credit studies to model the effects of different credit amounts and income phaseouts. They found that a $1,000 credit per child, phased out beginning at $50,000 for single filers, would reduce poverty among children under age six by 6.3% (about 148,000 children) and deep poverty by 9.9%, while 0.6% of working parents would leave the labor force.

Why This Article Was Selected for The Policy Scientist

Child tax credits remain a central policy issue because they directly affect family resources, labor market incentives, and child poverty, making careful empirical evaluation essential. As more states adopt their own credits, evidence on their likely effects has become increasingly valuable. Unrath and colleagues have contributed extensively to this research agenda, and this article extends their earlier work by applying recent causal evidence on state child tax credits to national policy simulations. Published in AEA Papers and Proceedings, a highly respected economics journal, the study draws on high-quality national survey and tax data. Its projections are broadly applicable to states considering similar policies. The microsimulation is carefully executed and informed by recent regression discontinuity estimates, a strong causal inference design.

Full Citation and Link to Article

Unrath, M., Tollett, N., Goldin, J., Homonoff, T., Lal, N., & Michelmore, K. (2026). How do state child tax credits affect employment and poverty? AEA Papers and Proceedings, 116, 320–325. https://doi.org/10.1257/pandp.20261088

Central Research Question

This study asks whether unconditional state child tax credits can substantially reduce child poverty without producing meaningful reductions in parental employment. More specifically, the authors estimate how alternative state child tax credit designs would affect labor force participation, poverty, deep poverty, and government costs in states that do not currently offer unconditional credits. Rather than evaluating one existing program, they simulate the likely consequences of adopting new refundable state child tax credits that resemble those already implemented in several states. They also examine how two important policy choices—the size of the credit and the income level at which benefits begin to phase out—change the projected employment and poverty effects. The authors seek to inform policymakers about the tradeoffs associated with expanding state-level support for families with young children.

Previous Literature

The study builds upon a substantial literature examining child tax credits, child allowances, and labor supply. Earlier research has frequently relied on evidence from expansions of the federal Earned Income Tax Credit (EITC) to estimate how parents respond to financial incentives. The authors argue that these older estimates may overstate employment reductions because the EITC targets a different population and encourages work among very low-income adults. In contrast, unconditional child tax credits operate differently and primarily affect families with children through changes in after-tax income.

Instead of relying on decades-old labor supply estimates, the authors incorporate findings from their own recent work and related studies examining actual state child tax credit programs. Their earlier regression discontinuity analyses exploited birth-date eligibility rules to estimate maternal employment responses using administrative tax records. Those studies found extremely small employment effects, providing more appropriate parameters for the current simulations. Consequently, the present article extends the existing literature by combining recent causal evidence with national microsimulation techniques to estimate how similar policies might perform across states that have not yet adopted unconditional child tax credits.

Data

The analysis combines several high-quality national data sources. The primary population data come from the American Community Survey (ACS), which provides detailed demographic, household, employment, and income information for millions of Americans. The authors use these data to identify eligible households with children younger than six years of age and to estimate how many parents could potentially be affected by alternative tax credit designs. They report that similar analyses using the Current Population Survey Annual Social and Economic Supplement produced nearly identical results, increasing confidence in the robustness of the findings.

To estimate poverty outcomes, the researchers use the Supplemental Poverty Measure (SPM), which incorporates taxes, transfers, geographic differences in living costs, and other resources ignored by the official poverty measure. They also estimate taxes using the National Bureau of Economic Research’s TAXSIM program and incorporate Supplemental Nutrition Assistance Program (SNAP) benefits into household income calculations. These data allow the authors to estimate not only employment responses but also changes in poverty status after accounting for tax liabilities, transfer payments, and simulated labor market adjustments. Overall, the data sources are nationally representative, widely respected, and well suited for evaluating the likely effects of alternative state child tax credit policies.

Methods

The study employs microsimulation rather than estimating the effects of a single existing policy. The authors simulate child tax credits worth $500, $1,000, and $2,000 per child while varying the income threshold at which benefits begin to phase out ($25,000, $50,000, and $100,000 for single parents). These policy combinations allow them to estimate how different program designs influence employment, poverty, deep poverty, and fiscal costs.

The employment model distinguishes between two mechanisms. First, an income effect occurs because families receive additional resources even if they do not work. Second, a substitution effect arises because receiving the credit may reduce the financial return from remaining employed once households enter the phaseout range. The authors calculate changes in both work incentives and after-tax income for individual households before estimating aggregate labor force responses. They intentionally treat some assumptions as upper bounds, recognizing that high-income parents may be less likely to leave employment than the model technically permits.

An important strength of the analysis is that the labor supply elasticities are not borrowed from older EITC research. Instead, they come from recent regression discontinuity studies of actual state child tax credits. Regression discontinuity is a well-established causal inference design because eligibility is determined by an arbitrary cutoff—in this case, children’s birth dates. This provides more credible estimates of behavioral responses than conventional multivariate regression models. Although the present article itself is a simulation rather than a new causal evaluation, it is grounded in recent causal evidence and therefore provides more credible projections than studies relying solely on historical labor supply estimates.

Findings/Size Effects

The simulations indicate that unconditional state child tax credits could substantially reduce child poverty while producing relatively small reductions in employment. Under the benchmark scenario—a $1,000 credit per child beginning to phase out at $50,000 of income for single parents—the model predicts a 6.3 percent reduction in poverty among children younger than six years of age. This translates into approximately 148,000 children being lifted above the Supplemental Poverty Measure poverty threshold if the policy were adopted across eligible states. Deep poverty declines even more, falling by 9.9 percent.

The projected employment effects are comparatively modest. The authors estimate that only 0.6 percent of eligible working parents—approximately 80,600 out of 14.4 million workers—would leave the labor force. Annual fiscal costs are estimated at roughly $11 billion nationwide under the benchmark scenario, with costs varying considerably across states because of differences in population size and eligibility.

Policy design substantially influences outcomes. Doubling the credit approximately doubles both poverty reduction and employment effects. Likewise, increasing the income threshold at which benefits phase out increases projected labor force exits because more households experience reduced returns to work. However, raising the phaseout threshold has only a limited effect on poverty because most eligible low-income families already qualify for the benefit under the lower thresholds. These findings suggest that benefit size has a stronger influence on poverty reduction than expanding eligibility further into the income distribution.

The results also suggest that earlier concerns regarding large employment reductions may have been overstated. Because the simulations rely on recent causal estimates from actual state child tax credit programs, they predict much smaller labor supply responses than studies based primarily on historical EITC expansions. Consequently, the article contributes updated evidence that better reflects contemporary labor markets and current policy environments.

Conclusion

This article provides a careful projection of how unconditional state child tax credits may influence employment, poverty, and government spending. By combining nationally representative survey data with microsimulation methods and labor supply estimates derived from recent regression discontinuity studies, the authors produce projections that are grounded in contemporary causal evidence rather than older policy environments. The analysis consistently indicates that relatively modest employment reductions accompany substantially larger reductions in child poverty and deep poverty.

The study also demonstrates that policy design matters. Larger credits produce greater poverty reduction but also somewhat larger employment effects, while broader eligibility has comparatively modest additional effects on poverty. Although the article models hypothetical programs rather than evaluating newly implemented policies directly, its methodology provides policymakers with realistic projections based on the best currently available evidence. As additional states adopt unconditional child tax credits, future quasi-experimental evaluations will be valuable for validating and refining these projections. Overall, the study represents an important extension of the growing literature on state child tax credits and provides a rigorous empirical framework for evaluating future policy proposals.