Does Enforcement Reduce Compliance?

Shu-Yi Oei 

Boston College Law School held its first Tax Policy Workshop of the semester last Thursday and the speaker was Surly Blogger Leandra Lederman. Leandra presented a draft paper entitled “To What Extent Does Enforcement Crowd Out Voluntary Tax Compliance?” The draft isn’t publicly available yet, but you can email Leandra for a copy.

So, what’s the paper about? The standard economic model tells us that a taxpayer will weigh the magnitude of the penalty and the likelihood of audit to reach an “expected” cost of punishment for tax evasion. Allingham & Sandmo (1972). So, if the audit rate is low (which it is), the expected cost of evasion also remains low, absent draconian penalties. Yet, we see relatively high voluntary compliance rates in the U.S. Some scholars claim that this is a “puzzle” and theorize that there is some sort of “intrinsic motivation” to comply with tax obligations, regardless of the low expected costs of punishment. Leandra has pointed out in several articles that this simple comparison presents a false puzzle because it ignores information reporting (and withholding), which IRS voluntary compliance statistics show is highly effective. She argues that information reporting is akin to an invisible audit. Nonetheless, some scholars suggest that enforcement and deterrence action are “extrinsic motivators” that might actually reduce compliance by displacing (i.e., “crowding out”) preexisting internal motivations to comply.

Leandra’s paper synthesizes the empirical literature on the effects of audit threats and fines as well as the growing tax and non-tax literature on contexts in which enforcement can lead to reduced compliance. In brief, the paper finds:

• That enforcement is generally effective, and that deterrence (once information reporting is taken into account) largely explains observed voluntary compliance;
• That an audit tends to increase compliance in subsequent periods for taxpayers who are found to owe more tax on audit;
• But that emerging evidence suggests that audits may decrease future compliance by those for whom no adjustment is made on audit.

Leandra recommends that more research be undertaken in order to see whether additional studies find decreased compliance by taxpayers with no audit adjustment, and to determine whether those taxpayers actually were compliant in the earlier year.

One interesting point that (as Leandra observes) has been raised by the Taxpayer Advocate, and that was debated in the workshop is that there could be other reasons for reduced compliance after audit by those found compliant. Explanations may apply that, strictly speaking, don’t fall within the rubric of “motivational crowding out” as narrowly defined to mean external motivators displacing preexisting internal motivations. It’s possible that an observed decrease in compliance could result from a number of different factors. For example, as others have pointed out, compliance may decrease after audit due to the “bomb crater” effect (taxpayers misperceiving the risk of future audit). Or perhaps audited taxpayers who were overly compliant before an audit found their audit experience not as unpleasant as anticipated (or learned in the audit that they were overcomplying) and thus stopped overcomplying. Another possibility that has been floated is that taxpayers learn ways to evade through experiencing the audit process. All these reasons—none of which implicate changed internal motivations—might possibly explain reduced compliance after audit.

More broadly, effects on tax compliance behaviors aside, I would be interested to know if enforcement action affects taxpayer behavior in other spheres (for example, in voting behavior, labor supply choices, consumption and investment behavior, and other social behaviors). This is obviously a separate issue from tax compliance, but one that deserves some consideration.

In the tax context, Leandra’s paper stands out as an invaluable, comprehensive, review of the literature on enforcement and how enforcement may crowd out compliant behavior. It sounds a cautionary note against claims that audits and enforcements should be reduced due to fears about lowered tax morale. Leandra has indicated an interest in doing empirical work on these questions, and I’d be interested in seeing what she finds.