Rational choice theory holds that individuals have preferences among the available alternatives, based on an analysis of the benefits to be gained, which are balanced against offsetting costs. The individual is expected to take into account all available information, the probabilities of outcomes, and best estimates of costs and benefits. Thus, if the incremental gain from cheating exceeds the expected incremental cost of doing so, then a rational person is more likely to cheat. Part of this calculation is an estimate of both the probability of being detected and the severity of the punishment to be expected if the person is detected. Thus, the cheating calculation is:
Engage in cheating when the expected incremental benefits > (Probability of detection
× Severity of punishment)
At a high level, the model would appear to explain cheating behavior. For example, a student is much less likely to cheat on the CPA exam when he knows that a monitoring camera is positioned in the ceiling directly over his head. Similarly, a cash receipts clerk is unlikely to steal a customer payment when she knows that the incoming cash was already logged in the mailroom, and her final results will be compared to that log. In these cases, the probability of detection is high, so it would take an inordinate benefit to offset this risk. Similarly, if a country has imposed the death penalty for anyone convicted of theft, it would be reasonable to assume that such a massive punishment would be pondered by perpetrators before they try to steal $50 from the petty cash box.
Nonetheless, there are several issues with rational choice theory, which are noted in the following bullet points:
Impulsive behavior. Many people who cheat do so in the moment, not running through the rational choice theory (which presupposes that a person is something of a cold-blooded, calculating machine). Instead, a sales clerk might suddenly decide to miscount the cash in the cash drawer and stuff the excess amount in his pocket, or a warehouse clerk might impulsively grab an expensive consumer goods item from the shelf, stick in his lunch box, and walk out with it. It never occurs to these people to consider the probability of being caught, or of the penalties that might be imposed on them.
Non-cash effect. People are more likely to cheat when the object being stolen is not cash. The reason is that there is a strong moral effect associated with the theft of cash, and a much lower effect with the theft of any other type of property. For example, an employee of a candy factory might wander into the company store and routinely take a candy bar from a retail display for lunch; the theft barely registers with the person. However, would that same employee go into the company store and extract the equivalent value of the candy bar from the cash register? Probably not. This non-cash effect applies to a great many instances of cheating, such as when employees steal goods from the warehouse, manipulate the exercise dates on their stock options to generate more profits, or claim extra benefits on their medical insurance to which they are not legally entitled.
Ethical override. In many instances, a person’s sense of moral right and wrong will prevent him from cheating, even when there is essentially no risk of being caught. For example, a truck carrying large amounts of cash flips over on the highway, scattering cash all over the road. Certainly, some drivers will scoop up all the cash they can and drive off with the loot. However, others will retrieve the cash and hand it back to the driver, despite the essentially zero risk of being caught. The same situation arises in a business, where an accountant might be fully aware of the complete absence of one or more key controls that would otherwise detect a theft, and yet persists in not taking advantage of the situation. The same moral override may occur when the amount of punishment expected is quite low, or zero. Thus, having a strong sense of ethics can completely offset rational choice theory.
In essence, the rational choice model assumes that the average person is an emotionless computer, who is constantly weighing costs and benefits, and who will cheat if the balance shifts in his favor. This is clearly not the case for most people.