Over the past three decades much empirical work has appeared on the effect of the criminal law on crime rates.  Usefully summarized in review articles by, among others, Professors Daniel Nagin of Carnegie Mellon University and Michael Tonry of the University of Minnesota (click here and here for examples), this research offers several insights for those engaged in the fight against corruption.

The first is that the criminal justice system can make a difference.  Save for acts committed in the heat of the moment, crime is a cost-benefit proposition.  Would-be criminals tote up the (usually) monetary gains of violating the law against the risks of being caught and punished and, when the benefits exceed the costs, commit an offense.  Thus policies that drive up the cost of crime by increasing the chances an offender will be caught, prosecuted, and appropriately punished reduce the crime rate.  Recent studies confirm that corruption crimes are no exception.  Putting more resources into to prosecuting corruption in the United States and ensuring corruption in the construction of roads is detected both reduced corruption.

But if the good news from the deterrence literature is that the criminal law can make a difference, the bad news is that that the difference is not easy to realize and that the logic of deterrence can lead policymakers astray.  One of the more striking findings is that what would seem to be the easiest way to enhance deterrence, sharp increases in the penalties for corruption crimes, may actually lead to more corruption. 

A number of studies have found that draconian punishments for drug offenses, mandatory sentences for repeat offenders, and enhanced penalties for crimes involving firearms have either had no effect on crime rates or increased them.  The reason is that police, prosecutors, and judges all enjoy enormous discretion in arresting, charging, and sentencing offenders, and when they feel the punishment is disproportionate to the offense, they use their discretion to blunt the law’s effect.  Those doing a cost-benefit analysis before deciding to commit a crime are fully aware of the reluctance of criminal justice personnel to observe these laws and factor it into their calculations.

That stiffening the penalties for corruption offenses could reduce the deterrent effect of the criminal law is one of many conclusions that flows from a basic finding of the deterrence research:  for more sophisticated crimes, would-be offenders have a well-honed sense of the costs and benefits of crime.  The evidence from studies of tax evasion in the United States provides an illustration.  While a common form of evasion is to under-report one’s income, American tax evaders are very careful about what income they report and what income they do not.  Because the lnternal Revenue Service can easily verify if income received by check or bank transfer is reported, tax payers rarely fail to report money received these ways.  But income paid in cash is very hard for the Service to verify.  Taxpayers know it, and as a result, upwards of 80 percent of income Americans receive in cash appears to go untaxed.

This finding explain why cross-checking land and motor vehicle registries and other public records showing asset ownership against the financial declarations public officials in many countries must make so rarely reveals an inconsistency.  Just as American tax evaders know the Internal Revenue Service will compare the income they report against the record of checks and bank transfers they received, corrupt officials know anticorruption agencies will check their financial declarations against public records.  Hence it is the exceptional case when a corrupt official is exposed in this way.

What the tax evasion research does suggest, though, is the deterrent value of regularly conducted audits of officials’ lifestyles by comparing the value of the official’s house, amounts spent on vacations and education, and the like against declared income.  This has a much greater likelihood of exposing wealth acquired illicitly.  Officials tempted by corruption will know it and as a result audtis can be a powerful deterrent to corrupt behavior.

Two other findings of note for anticorruption policy.  One is that deterrence calculations are affected by the capacity of the system to actually punish crime.  If those contemplating the commission of a crime know the prisons are overcrowded or the courts overwhelmed by delays, the deterrent value of the criminal justice system will be close to nil.  Thus without a well functioning judiciary and corrections system it makes little difference how stiff the penalties for bribery, collusion, and other corruption crimes are.  Second, in some countries the stigma of being revealed a criminal is a strong deterrent.  The fear that one would be treated as an outcast if caught provides a more powerful incentive to obey the law than the threat of prison.  Public campaigns stressing how shameful corruption is can therefore be a valuable part of an anticorruption effort.