THE DETERRENT EFFECT OF CRIMINAL SANCTIONS ON AFDC FRAUD: AN ECONOMETRIC STUDY

<p>This research measures the general deterrent effect of the risk of conviction and severity of punishment on the incidence of AFDC fraud in the State of Maryland. This study also determines the impact of law-enforcement expenditures and various socioeconomic factors on fraud rates. The enforcement of laws against welfare cheating and the treatment of fraudulent cases resemble the operation of the broader Criminal Justice System. Thus, this inquiry considers fraud as a specific offense which can be studied as an application of general economic theories of crime deterrence. A three-equation simultaneous econometric system is taken from one of such theories and adjusted to reflect the welfare environment. The model uses pooled cross-section and time-series aggregate data and the Maryland counties as the units of analysis. The results indicate that the expected cost of punishment (i.e., the product of the probability of conviction and the severity of punishment) produces a significant deterrent effect. However, the generated effect is not necessarily significant from a policy viewpoint. The impact of the size of the defrauded AFDC grant and the unemployment rate were found to outweigh the deterrent effect of the cost of punishment and expected earnings from employment. Also, the results show that increases in fraud rates and crime rates, other than AFDC fraud, reduce the magnitude of the deterrent effect of the cost of punishment. On the other hand, increases in investigative expenditures and productivity of investigators/prosecutors/judges were found to augment the effect of the cost of punishment. Similarly, the data analysis detected high costs of punishment in wealthy counties. Finally, the study found investigative resources to be allocated in an incremental manner and to be insensitive to prevailing rates and social losses from fraud. This inquiry concludes that the expected cost of punishment imposed on potential AFDC cheaters in the State of Maryland does not lead to fraud control and prevention. This will be achieved only when the expected cost of punishment at least equals the expected illegal gains from fraud. Strategies to achieve this goal are suggested by the findings of this study.</p>