Literature Review

The welfare effects of tax rate uncertainty connect several strands of economic literature, from optimal taxation theory to behavioral responses under uncertainty. This section situates the current analysis within this broader context, highlighting how the paper builds on and extends existing work.

Tax Uncertainty and Welfare¶

The foundational work on tax uncertainty’s welfare effects comes from Skinner (1988), who developed a two-period model showing that uncertain tax policy could reduce welfare by 0.4% of national income annually. This seminal contribution established that tax uncertainty represents more than a minor friction—it imposes first-order welfare costs comparable to major tax distortions. Skinner’s analysis focused primarily on savings decisions and used a representative agent framework with limited empirical calibration to actual tax rate distributions.

Alm (1988) extended this analysis by examining how uncertainty affects the difficulty of individual decision-making, demonstrating that tax complexity and uncertainty interact to amplify welfare losses. When agents cannot easily determine their marginal tax rates, they make systematically suboptimal choices even in the absence of policy uncertainty. This insight motivates the current paper’s focus on information provision as a policy tool.

More recent work by Baker et al. (2016) has documented the macroeconomic effects of policy uncertainty more broadly, showing that uncertainty shocks can reduce investment and employment. While their analysis encompasses various types of policy uncertainty beyond taxation, they find that tax policy uncertainty accounts for a substantial share of overall policy uncertainty effects. Their work provides empirical support for the large welfare costs found in theoretical models.

Hassett and Metcalf (1999) examined how tax rate uncertainty affects investment decisions, finding that uncertainty can either increase or decrease investment depending on the structure of adjustment costs and irreversibility. Their analysis highlights that uncertainty effects are not uniform across different types of economic decisions, a theme that emerges in the current paper’s finding of heterogeneous effects across the income distribution.

Optimal Taxation Under Information Constraints¶

The optimal taxation literature, beginning with Mirrlees (1971) and extended by Diamond and Mirrlees (1971), traditionally assumes that agents have perfect information about tax schedules when making economic decisions. This assumption, while analytically convenient, abstracts from a crucial real-world friction. Several papers have begun to relax this assumption in various ways.

Chetty et al. (2009) show that salience matters for tax responses—agents respond more strongly to taxes that are more visible or easier to understand. Their experimental and quasi-experimental evidence demonstrates that information presentation affects behavior even when the underlying economic incentives remain constant. This work suggests that complexity-induced uncertainty may have effects similar to actual policy uncertainty.

Benzarti et al. (2020) find that firms and consumers have asymmetric responses to tax changes, partly due to differential awareness and understanding of tax provisions. Their work emphasizes that information frictions can fundamentally alter tax incidence, a consideration that extends to the uncertainty context studied here.

Feldstein (1999) and Chetty (2009) have shown how behavioral responses to taxation depend on the elasticity of taxable income, which itself may be affected by uncertainty. When agents are uncertain about future tax rates, they may exhibit different elasticities than under certainty, complicating the design of optimal tax systems.

Labor Supply and Uncertainty¶

The labor supply literature has long recognized that uncertainty affects work decisions, though most analyses focus on wage uncertainty rather than tax uncertainty. Block et al. (1980) examine labor supply under wage uncertainty, finding that risk-averse workers generally supply less labor when wages are uncertain. However, they note that with Cobb-Douglas preferences—the case examined in detail in this paper—labor supply is invariant to wage uncertainty due to exactly offsetting income and substitution effects.

This special property of Cobb-Douglas preferences, while knife-edge in theory, provides a useful benchmark for isolating the pure effect of tax rate uncertainty from risk aversion. Ham and Reilly (2002) provide empirical evidence that actual labor supply responses to uncertainty fall between the extremes of complete invariance (Cobb-Douglas) and strong negative responses (high risk aversion), supporting the use of Cobb-Douglas as a reasonable approximation for policy analysis.

Blundell and MaCurdy (1999) survey the vast literature on labor supply elasticities, noting that estimates vary widely depending on the population studied and the source of variation used for identification. Their work suggests that uncertainty about the relevant elasticity parameters themselves represents another source of welfare loss not captured in standard models.

Information Provision and Economic Outcomes¶

A growing literature examines how information provision affects economic decisions and welfare. Hastings and Weinstein (2008) show that providing information about school quality significantly affects school choice, with larger effects for disadvantaged families. Duflo and Saez (2003) find that information about retirement savings options substantially increases participation rates. These studies suggest that information interventions can have large welfare effects at low cost.

In the tax context, Chetty & Saez (2013) conducted a field experiment showing that providing information about the Earned Income Tax Credit (EITC) to tax preparers increased EITC claims and changed reported income. Their work demonstrates that even when tax rules are stable, information frictions prevent optimal responses.

Recent work by Hoopes, Reck, and Slemrod (2015) uses data on searches for tax information to show that taxpayers have limited knowledge of tax provisions and actively seek information primarily around filing time. This pattern suggests that tax uncertainty is partly endogenous to the complexity of the tax system and the cost of acquiring information.

Rees-Jones & Taubinsky (2020) provide crucial evidence on how taxpayers mentally approximate complex tax schedules. They find that 43% of taxpayers use an “ironing” heuristic that linearizes the tax schedule around their average tax rate, leading to systematic underestimation of marginal tax rates. This mental model helps explain why taxpayers struggle to optimize their responses to tax incentives, even when tax policy itself is stable and knowable.

Heterogeneous Effects and Distributional Considerations¶

The distributional effects of tax uncertainty have received limited attention in the literature, despite their importance for policy design. Batchelder (2003) argues that tax complexity disproportionately burdens low-income households who cannot afford professional tax assistance. However, the current paper’s finding that middle-income households face the highest uncertainty costs suggests a more nuanced distributional pattern.

Goldin (2018) examines how complexity affects the take-up of tax benefits, finding that complexity acts as an ordeal mechanism that screens out some eligible recipients. While his focus is on benefit programs rather than tax rates, the analysis highlights how information frictions can have unintended distributional consequences.

Contributions of This Paper¶

This paper makes several contributions to these literatures. First, it provides a unified framework for analyzing tax rate uncertainty that incorporates both policy uncertainty and complexity-induced uncertainty. Second, it uses comprehensive administrative data from PolicyEngine-US to calibrate the model to actual tax rate distributions, providing more reliable welfare estimates than previous studies. Third, it explicitly analyzes the heterogeneous effects of uncertainty across the income distribution, revealing that middle-income households bear the highest costs.

The paper also extends the optimal taxation literature by showing how a social planner should adjust tax rates when accounting for uncertainty costs. This normative analysis provides practical guidance for tax reform efforts and highlights the value of policy transparency and advance notice of tax changes.

Finally, the emphasis on information provision as a costless tool for welfare improvement connects the tax uncertainty literature with the broader information economics literature, suggesting new avenues for policy intervention that do not require traditional fiscal tradeoffs.