Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published September 20, 2017 | Submitted
Report Open

An Equilibrium Model of Tax Compliance with a Bayesian Auditor and some 'Honest' Taxpayers

Abstract

Empirical work on tax compliance has yielded conservative estimates of unreported taxable income in the U.S. that average 10 to 15 percent of total taxable income for recent years. Moreover, it is held by many that the rate of noncompliance has been growing dramatically. This problem is widely perceived as one of eroding ethics—more and more people are ceasing to comply voluntarily and are instead acting "strategically" in response to the structure of the U.S. income tax laws. We propose a simple model of tax compliance in which an exogenously given fraction of taxpayers comply voluntarily, while the remainder behave strategically. We distinguish between a general decision to act strategically and a specific decision not to report honestly. This is done in an equilibrium setting where the IRS is allowed to adjust its audit policy in response to taxpayer behavior. Because the audit policy of the IRS is endogenous and thus co-determined with the reporting behavior of potential noncompliers, several non-intuitive results emerge. In particular, we find that an increase in the fraction of strategic taxpayers decreases the likelihood that a given strategic taxpayer fails to comply. In fact, the decrease in the likelihood of underreporting exactly offsets the increase in the fraction of strategic taxpayers, so that aggregate compliance (and net tax revenues) are unaffected.

Additional Information

Revised. Original dated to December 1983. We would like to thank Kim Border and members of the Caltech Theory Workshop for helpful comments. The financial support of National Science Foundation Grant No. SES-8315422 is gratefully acknowledged.

Attached Files

Submitted - sswp506_-_revised.pdf

Files

sswp506_-_revised.pdf
Files (353.8 kB)
Name Size Download all
md5:cdc3fed8e854372dedc57e2768cac7a2
353.8 kB Preview Download

Additional details

Created:
August 19, 2023
Modified:
January 14, 2024