Taxpayer compliance under uncertainty

Paul J. Beck, Woon Oh Jung

Research output: Contribution to journalArticlepeer-review


The present study investigates taxpayers' economic incentives to comply under conditions of uncertainty induced by tax law complexity and audit outcome uncertainty. In particular, previous research is extended by modeling an environment in which taxpayers have uncertainty about their tax liability and by simultaneously considering different tax rate structures and risk-taking attitudes. The major theoretical finding is that greater tax liability uncertainty creates incentives for taxpayers to increase reported income under certain conditions, but to decrease reported income under other conditions. For ranges of parameter conditions that commonly occur in the United States, however, we show that, unless taxpayers are highly risk-averse, an increase in income uncertainty is much more likely to lead to a reduction in reported income. Taxpayers' incentives to obtain tax advisory services as a means of reducing uncertainty also are investigated and shown to depend upon the audit probability, the penalty rate, and the tax rate. We find that increasing the audit probability will not necessarily increase taxpayers' incentives to purchase advisory services. Another finding is that the effects of tax rate changes depend upon taxpayers' risk-taking attitudes and can either increase or have no effect upon reported income. Furthermore, we find that taxpayers' reporting decisions under a progressive rate structure are the same as under a proportionate rate structure. Several policy implications of our analysis for taxpayer compliance and experimental studies are identified.

Original languageEnglish (US)
Pages (from-to)1-27
Number of pages27
JournalJournal of Accounting and Public Policy
Issue number1
StatePublished - 1989

ASJC Scopus subject areas

  • Accounting
  • Sociology and Political Science


Dive into the research topics of 'Taxpayer compliance under uncertainty'. Together they form a unique fingerprint.

Cite this