Boning et al. (2024) examine the returns to expanding IRS tax audits on individuals across the income distribution. The paper combines a series of comprehensive administrative datasets to estimate the breakdown of costs and revenues associated with a marginal increase in audit enforcement spending. To measure the welfare impacts of audits, the paper also estimates the individual deterrence effect of audits on future tax collection as well as the monetized burden associated with being audited.
The paper uses these results to estimate the MVPF of auditing low-income (20-30th percentile) individuals. The paper separately estimates the MVPF of auditing high-income (90-99th percentile) individuals.
MVPF = 1.3
The net cost includes the mechanical cost of expanding audits, the direct increase in government revenue from the audit, and future revenue from individual deterrence. The mechanical cost of an additional dollar spent on audit enforcement is $1.
The paper estimates that audits of individuals in the 20-30th income percentile generates $0.83 per $1 of cost. Using business unit accounting information from the IRS, the paper estimates that roughly 27% of audit costs are fixed. Therefore, an additional $1 spent on auditing low-income individuals generates $1.14 of revenue (0.83 / (1 – 0.27)).
Audits also raise revenue through individual deterrence in which auditing an individual in one year increases the taxes paid by that individual in subsequent years. Using randomly selected audits from the IRS’ National Research Program, the paper estimates that, in present discounted value, these additional tax payments raise 3.2 times as much revenue as the initial audit. For low-income individuals, this results in $3.63 of deterrence revenue.
The net cost is then -$1 + $1.14 + $3.63 = $3.78.
Individuals are willing to pay to avoid the increase in spending on audits. Following the approach of calculating the net cost, low-income individuals are willing to pay $4.77 (1.14 + 3.63) to avoid paying the upfront increase in their tax liability ($1.14) and the future increase ($3.63) from deterrence.
Individuals have an additional willingness to pay to avoid the audit apart from the increased tax liability. Using survey data, the IRS constructs a measure of the total monetized burden associated with being audited (e.g. time spent, payments to lawyers). The paper estimates that a $1 increase in audit spending leads to a $0.14 burden to low-income individuals.
The total willingness to pay is then $1.14 + $3.63 + $0.14 = $4.92.
The MVPF of raising revenue from a marginal expansion of audits on low-income individuals is $4.92/$3.78 = 1.30.
Boning, William C., Nathaniel Hendren, Ben Sprung-Keyser, and Ellen Stuart (2024). “A Welfare Analysis of Tax Audits Across the Income Distribution.” Forthcoming in The Quarterly Journal of Economics. DOI: https://doi.org/10.1093/qje/qjae037