Does a Rising Tide Always Lift All Boats? How Public Financing Affects Racial Disparities in Labor and Education Outcomes

This paper investigates whether increases in local government spending benefit all racial groups equally, focusing on labor market and educational outcomes. Using a natural experiment—Moody’s 2010 municipal bond rating recalibration—the study examines labor and test score data from the American Community Survey and the Stanford Education Data Archive. A one-standard-deviation increase in spending raises White employment by 0.13 percentage points and annual earnings by $137 but has no significant effect on Black or Hispanic individuals. Similarly, White students’ test scores rise by 0.48% of a standard deviation, while minority students show no gain, widening racial disparities.

Full Citation and Link to Article

Qiu, Tian. “Public Financing and Racial Disparities: Does A Rising Tide Always Lift All Boats?” American Economic Journal. https://tinyurl.com/27hf863z

Extended Summary

Central Research Question

This paper examines whether increases in local government spending—enabled through improved municipal bond financing—benefit racial and ethnic groups equally. Specifically, it asks whether White, Black, and Hispanic households experience similar improvements in labor market and educational outcomes following exogenous financial shocks that increase local government spending. The study explores not only the immediate economic effects on adults but also the intergenerational consequences for children’s academic achievement, highlighting how a well-intentioned “rising tide” may not lift all boats equally.

Previous Literature

Previous research has extensively documented the positive effects of public spending on socioeconomic outcomes. Studies by Adelino, Cunha, and Ferreira (2017) and Moretti (2010) found that improved local government financing boosts employment and household income. Other work by Dahl and Lochner (2012) and Johnson (2020) established that gains in parental income improve children’s academic performance. Additionally, literature on school finance (e.g., Jackson et al., 2016) has shown that increases in school-specific spending lead to better educational outcomes. However, the distribution of these benefits by race has received less attention. This paper addresses that gap by analyzing whether gains from increased public financing are shared equally among racial groups.

Data

The analysis uses several datasets:

1. Moody’s Municipal Bond Ratings Recalibration (2010): A nationwide, exogenous event that upgraded credit ratings on nearly 70,000 municipal bonds, improving the borrowing capacity of affected local governments.

2. Annual Survey of State and Local Government Finances: Used to measure changes in local government spending before and after the recalibration.

3. American Community Survey (ACS) Microdata (2006–2014): Used to examine individual labor market outcomes across racial groups, including employment, earnings, hours worked, and SNAP usage.

4. Stanford Education Data Archive (SEDA): Provides standardized test scores in math and English for students in grades 3–8, disaggregated by race at the county level.

5. Implicit Association Test (IAT): Used to classify states by levels of implicit racial bias, to assess how local attitudes may moderate the effects of public financing.

Methods

The study uses a difference-in-differences (DiD) estimation strategy that exploits variation in the intensity of Moody’s recalibration across counties. Counties where a larger fraction of local governments received bond upgrades are considered to have experienced a stronger treatment. The intensity of treatment is measured as the proportion of a county’s local governments affected by the recalibration. Key regression models compare labor and education outcomes before and after the recalibration, controlling for individual demographics, county fixed effects, and state-by-year fixed effects.

To ensure validity, the analysis accounts for potential confounders such as migration (by excluding movers in robustness checks) and differential effects of the Great Recession. Event-time estimations are also used to visually assess trends in outcomes over time, helping to confirm the parallel trends assumption required for DiD inference.

Findings/Size Effects

Labor Market Outcomes:

For White individuals, a one-standard-deviation increase in recalibration intensity leads to a 0.13 percentage point increase in employment and a $137 rise in annual earnings. Weekly hours worked also increase by about 0.09 hours, and SNAP participation falls by 0.18 percentage points. These findings are statistically significant and economically meaningful. However, the same treatment has no significant effect on Black or Hispanic individuals, leading to a widening racial gap in labor market outcomes.

A principal component analysis was used to generate a composite labor market index, which showed an overall improvement for White residents but no change for minorities. Subgroup analyses reveal that low-SES minorities, especially in high-bias states (as measured by the IAT), experience worse outcomes post-treatment, possibly due to substitution effects in hiring.

Educational Outcomes:

Consistent with findings in the labor market, the improvements in economic outcomes for White families translated into better test scores for their children. A one-standard-deviation increase in treatment intensity is associated with a 0.48% of a standard deviation increase in test scores for White students. Black and Hispanic students, however, show no statistically significant changes.

The study further finds that the effects on education are persistent and emerge with a one-year lag, consistent with theories of income-induced educational investment. Event-time estimations indicate no pre-treatment differences between treated and control counties, supporting the validity of the DiD design.

Achievement Gaps:

The paper finds that the White-Black academic achievement gap widened by 1.02% of a standard deviation and the White-Hispanic gap by 0.53% following increased local government spending. These widening gaps are more pronounced in smaller counties and are robust to controls for pre-existing differences in county-level socioeconomic status (SES).

A decomposition model shows that over 90% of the widening test score gaps are attributable to racial disparities rather than differences in household income. This implies that the effects are not merely reflections of class-based inequalities but are deeply tied to race-specific dynamics in how public investments are distributed and internalized.

Conclusion

The findings challenge the conventional wisdom that public investment automatically reduces inequality. Instead, this study shows that while improved municipal financing can raise aggregate outcomes, the benefits are unequally distributed across racial lines. White households and their children are significantly more likely to experience economic and educational gains, while Black and Hispanic households often see no change—or in some cases, relative decline.

The study’s contributions are twofold. First, it demonstrates that the “last hired, first fired” phenomenon extends into the domain of public investment, not just economic downturns. Second, it identifies an important intergenerational mechanism: uneven gains in adult employment and income translate into widening educational disparities among children.

These findings have critical implications for policymakers. Without targeted efforts to ensure that public funds are equitably distributed and effectively reach underrepresented communities, well-meaning fiscal policies may unintentionally exacerbate existing racial inequalities. As governments pursue stimulus and recovery plans—especially in the wake of COVID-19—it is essential to incorporate equity considerations into public finance decisions.