Does Knowing Where Your Taxes Go Make You More Willing to Pay Them?

This study explores whether perceptions about how tax dollars are spent influence taxpayers’ willingness to comply, focusing on property taxes funding public schools. Using a natural field experiment in Dallas County, Texas, involving 2,110 households, the authors randomly provided accurate information about the share of property taxes going to schools. Households with children reduced their probability of filing a tax appeal by 4.09 percentage points, while households without children increased it by 2.78 percentage points following a 10 percentage point perceived increase in school share. The findings support the reciprocal motivation hypothesis: perceived personal benefit increases tax morale and compliance.

Full Citation and Link to Article

Giaccobasso, M., Nathan, B., Perez-Truglia, R., & Zentner, A. (2024). Where do my tax dollars go? Tax morale effects of government spending. American Economic Journal: Applied Economics. Advance online publication. https://doi.org/10.1257/app.20220251

Expanded Summary

Central Research Question

This study examines whether individuals’ perceptions about how their tax dollars are spent affect their willingness to pay taxes. Specifically, it tests the “reciprocal motivation” hypothesis, which suggests that taxpayers are more willing to comply with tax obligations if they believe the resulting government services benefit them personally. The authors use the case of property taxes funding public schools in Dallas County, Texas, to explore whether correcting misperceptions about government spending alters behavior—in this case, the likelihood of filing a tax appeal. The study seeks to understand whether transparent, targeted information about tax allocation can increase tax morale and compliance.

Previous Literature

Two main perspectives dominate the literature on tax compliance: institutional enforcement and tax morale. While strong institutions and third-party reporting have been shown to improve tax compliance (e.g., Slemrod, 2019; Bagchi and Dušek, 2021), evidence for tax morale—the intrinsic motivation to comply with taxes—is more limited. Luttmer and Singhal (2014) proposed that tax morale can be strengthened when individuals feel they benefit from the services funded by their taxes.

Earlier research often relied on surveys or lab experiments, which faced criticism for social desirability bias and lack of real-world stakes. More recent field experiments in developing countries (Carrillo et al., 2021; Krause, 2020) show that public service delivery (e.g., infrastructure improvements) can increase compliance. However, these settings differ significantly from high-enforcement contexts like the U.S., where compliance is already high due to institutional mechanisms like escrow systems and tax lien sales. This paper contributes to the literature by using a high-stakes, real-world setting in a developed country and by isolating the effect of perceived spending on actual tax behavior—property tax appeals—rather than on stated preferences or hypothetical choices.

Data

The study draws on a rich combination of administrative and survey data:

1. Administrative data: Property-level records from the Dallas County Appraisal District (DCAD), including assessed home values, tax amounts, and tax appeals from over 400,000 owner-occupied properties.

2. Survey data: A targeted survey of 2,110 homeowners invited from a sample of 78,128 households, with oversampling in high-income districts and among those with rising tax bills.

3. Linked data: Survey responses were linked to administrative data to measure prior tax behavior and post-survey outcomes.

The dataset allows for detailed measurement of household characteristics (e.g., home value, exemption status), tax details (e.g., the share going to schools), and real behavioral outcomes (i.e., filing a tax appeal).

Methods

The authors conducted a natural field experiment embedded in a survey distributed shortly before the annual property tax appeal deadline. The sample includes 2,110 households that completed the survey and passed quality checks.

The experiment unfolded in three steps:

1. Eliciting prior beliefs: Respondents were asked to estimate what proportion of their property taxes went to fund public schools (the “school share”).

2. Information treatment: Half of the respondents were randomly provided with the actual school share, based on administrative data (average = 49.78%).

3. Eliciting posterior beliefs: All respondents were then asked again to estimate the school share for the following year, allowing measurement of belief updating.

Randomization ensured that the only systematic difference between treatment and control groups was access to the correct information. This setup creates an exogenous shock to perceived tax benefit and allows causal inference.

The main outcome is whether the household filed a direct tax protest (appeal) in 2021. The authors also collected intention-to-protest responses and linked all subjects to actual appeal behavior in administrative records.

A Two-Stage Least Squares (2SLS) model was used to estimate the causal effect of posterior beliefs about school share on protest behavior. This approach addresses the endogeneity of beliefs by instrumenting for posterior beliefs with the randomized information shock (the difference between true and prior beliefs, interacted with treatment assignment).

The analysis distinguishes between two types of households:

• Households with children enrolled in local public schools (direct beneficiaries of school taxes).

• Households without children, including those with no children or children in private, charter, or home schools (not direct beneficiaries).

Findings/Size Effects

The study uncovers several important findings:

1. Widespread misperceptions: Only about one-third of households correctly estimated their school share within ±5 percentage points. On average, households underestimated the school share by 13 percentage points, suggesting low salience of existing public information.

2. Belief updating: Treated households significantly updated their beliefs. The average revision was 10.9 percentage points for households with children and 10.6 for those without. Bayesian updating estimates suggest that subjects adjusted their beliefs by about 76% of the gap between prior belief and actual value.

3. Behavioral effects:

   
   

   • For households with children, a 10-percentage point increase in the perceived school share reduced the likelihood of filing a protest by 4.09 percentage points, or 12.1% of the baseline protest rate (33.8%).

   • For households without children, the same belief shift increased the protest rate by 2.78 percentage points, or 9.6% of their baseline rate (29.0%).

   
   

   These effects are economically meaningful and statistically significant, confirming the reciprocal motivation hypothesis: perceived personal benefit from taxes increases willingness to comply.

4. Heterogeneous effects: The magnitude and direction of the effect depended on whether the household benefited directly from school spending. This heterogeneity underscores the importance of targeting public communications to taxpayer subgroups.

5. Tax appeals as high-stakes behavior: The protest rate among survey respondents was over 30%, and most successful protests resulted in substantial property tax reductions, often hundreds of dollars. Thus, behavior changes triggered by information had real financial consequences for households and fiscal implications for local government revenue.

6. Robustness: The results are robust to alternative specifications, adjustments for belief outliers, and falsification tests (e.g., no effect observed among subjects who had already appealed before the survey).

Conclusion

This study provides the first field-experimental evidence that perceptions about government spending causally affect tax compliance behavior in a high-enforcement, developed-country context. The authors show that when households are informed that their taxes support services they value—such as public education—they are significantly less likely to protest their tax bill. Conversely, when households learn that they benefit less than expected, they are more likely to challenge their tax burden.

The findings have substantial implications for tax policy and public finance. First, they suggest that transparent and personalized communication about how tax dollars are used can enhance tax morale and reduce costly legal appeals. Second, the results lend support to the broader idea that reciprocity and perceived fairness are central to voluntary tax compliance, especially in settings where traditional enforcement mechanisms are already strong.

This study also cautions that aggregate effects may obscure important subgroup dynamics: the same information can simultaneously increase and decrease compliance across different segments of the population. Future work should explore how these insights apply to other margins of tax behavior, such as underreporting income or paying taxes late, and in different policy domains (e.g., health care, transportation).

In sum, the paper adds rigor and nuance to the literature on tax morale by identifying a specific, measurable channel—perceived benefit from government spending—and providing causal evidence that even modest informational nudges can shift real-world tax behavior. It opens new avenues for policy interventions that complement enforcement with behavioral insights.