However irrational you think it may be to commit a crime, it differs no less than from other theories of human behaviour: "“[it] does not require ad hoc concepts of differential association, anomie, and the like, nor does it assume perfect knowledge.”

This quote comes from a research paper by Gary Becker (I think he is quite famous) linked here. That research paper is what this article is based on. What Becker is saying, is that whilst we don’t assume people always have perfect knowledge or always make rational decisions, people do consider the costs and benefits of their actions that usually “result in an optimal level of benefit or utility” (Hayes, 2024),. And Becker believes this is even the same with criminals. Criminals weigh the cost and benefits of their crimes - “some persons become criminals, therefore, not because their basic motivation differs from that of other persons, but because their benefits and costs differ.” (Becker, 1974)

What an interesting point of view. So, we can think of crime like any other human behaviour – an individual considers their costs, benefits, and chooses the one with the greatest utility (utility means satisfaction/benefit, remember this). This is not to say criminals always make rational decisions, sometimes they compute it incorrectly and it leads to an overproduction of crimes.

The biggest factors Becker believes that affect the costs and benefits of a crime is the probability of being caught (p) and the severity of punishment (f). When these increase, the EU from a crime decreases, and the same goes inversely. This makes sense… and it is nothing groundbreaking. This is in general too, it affects every kind of person who has some or no tolerance for risk.

But what isn’t in general is u - in the formula, u is a “portmanteau” of things.. it’s a function. It changes for each individual. An obvious example might be some individuals have greater utility from crime due to having a lower income, or lower education (thus less “law-abidingness”), so they get greater utility from illegal activity versus legal activity…. but what one contributor he touched on was u being risk preference type.

What is your risk preference? Are you risk neutral (linear u), averse (concave u) or risk preferred (convex u)? This shapes the curve (shown below). From that, people have different expected utilities from committing a crime, and they also evaluate the risky payoff differently. Therefore, the police are not just up against reducing p or f, but also crime pays depending on how you value uncertainty.

I did run the numbers (they are below) and for a risk averse individual, their expected utility (satisfaction) is lower than risk neutral and risk preferred - makes sense, they don’t get a good thrill from committing crimes. Additionally, their certainty equivalent (CE) < the crimes expected income. The intuition behind this is, the risk averse individual would rather take the CE - the guaranteed return that they can have now, instead of the crimes expected income. They would take less just to avoid the crime! (5.89 CE < 5.9 E(Y))

It’s the opposite case for a risk preferred individual, who’s CE > the crimes expected income. This means they’d rather take a risky bet and try to gain more than the expected income. (5.91 CE > 5.9 E(Y))

For a risk neutral individual, they are indifferent - they don’t care about risk. Their CE is equal to expected income of 5.9.