The Price of Getting Caught

MOST of us assume the answer to crime is obvious: more police and longer prison sentences. Dr. Emiliano Tealde, an associate professor in the Department of Social Sciences at Universidad Católica del Uruguay in Montevideo, tests assumptions like these with data. He uses causal evaluation methods, statistical tools for telling cause from coincidence, on real-world, non-experimental data to study reoffending, policing and deterrence across Latin America.

We asked him what the evidence says about light, jobs, police and punishment, and where he would point Uruguay's policymakers if he could choose.

The Confluence: How did you first get drawn to the economics of crime and criminal justice?

Dr. Tealde: For three reasons. First, the importance of security in our societies. When we cannot control crime, particularly organised and violent crime like drug trafficking, we face a host of negative consequences, ranging from being unable to walk the streets without fear to democratic instability.

Second, I was drawn by the fact that the public policies used to combat it generally take a very short-term or very narrow approach, or both. It is an area where we could still make significant progress if we achieved a better connection between academia and the state. Third, it is a field that allows me to use causal evaluation methods with observational data, that is, non-experimental data, which I enjoy working with.

The Confluence: Your work on reoffending after prison looks at how job opportunities affect whether someone returns to crime. What did you find, and why does something as simple as information about jobs matter so much?

Dr. Tealde: In my research on recidivism, I focus on how employment in low-skill sectors affects reoffending rates. The availability of jobs in sectors like construction, traditionally a major employer of people with criminal records, has a significant impact. However, this impact is strong but fades over time: it leads to a drop in recidivism during the first few months after release from prison, but after a year the rates are unaffected.

Evidence from various countries points to similar results, revealing a common pattern. When job opportunities in low-skill sectors improve, recidivism decreases, but only for economic crimes such as theft and robbery. More violent crimes, such as homicide or gender-based violence, are far less likely to be influenced by changes in the labour market.

There are two sides to this reality. On the one hand, incentives matter, and we can shape behaviour by offering the right economic incentives. On the other hand, there are serious crimes over which we have far less influence.

The body of research linking job-related information to recidivism is a fascinating area of study. To understand it, we must briefly put ourselves in the shoes of those leaving prison. This group faces hardships that many of us would find difficult to grasp fully. For certain sections of society, access to information is a privilege that cannot be taken for granted, and accessing that information is the first step towards entering the labour market.

The Confluence: You've also studied how street lighting affects crime rates. What's the intuition behind why something as basic as light changes people's behaviour?

Dr. Tealde: Criminals are always looking for ways to commit a crime without getting caught. The street criminal's main adversary is the witness who can provide evidence against them. Street lighting can make it easier for witnesses to see someone who may be committing a crime, which deters criminals.

This idea stems from the Becker model of crime, named after the economist Gary Becker, which proposes that a criminal weighs the expected costs and benefits [allowing for the chance of being caught] of committing a crime. If the expected benefit is greater, they will commit the crime; otherwise, they will not. Street lighting increases the expected cost of committing a crime by making it more likely that a witness will provide evidence against the criminal.

The Confluence: On policing: does putting more police in one area actually reduce crime overall, or does it just push it into neighbouring areas?

Dr. Tealde: Increasing police presence in a specific location typically reduces crime in the area where officers are deployed, while increasing it in adjacent areas. Evidence also suggests that a heavier police presence leads to the temporary displacement of crime; for instance, increasing police presence during the week may shift criminal activity towards the weekend.

Overall, crime levels tend to drop with a greater police presence, as the displacement of crime rarely fully offsets the reduction achieved in the areas with more officers. Criminals had chosen to operate in particular places because that suited their activities best; being displaced results in losses for them.

However, this effect is neither universal nor applicable to all types of crime. In countries with severely flawed judicial systems, police struggle far more to reduce crime, because offenders are less likely to face consequences for their actions. Furthermore, police are effective at reducing crime when the activity is not highly organised. Highly organised criminals invariably find ways to evade the police, and different kinds of measures are needed to tackle them.

The Confluence: Of the three levers governments usually reach for (harsher punishment, more deterrence through policing, and better legal opportunities such as jobs and support for reintegration), what does the evidence say works best, and what is overrated?

Dr. Tealde: Harsher penalties work for certain types of offender. For instance, there is evidence that they have a slight effect on adolescents and on adults accused of gender-based violence. For most common crimes, however, they generally have little impact. In general, the public tends to overestimate the effect of increasing penalties. When committing a crime, offenders are not thinking about getting caught; they do not consider whether getting caught would mean one, two or five years in prison.

The deterrent effect the police can provide is limited to specific, less serious crimes, and relies on police action being backed by a robust judicial system. While it is a useful tool, it does not address every security problem.

Greater opportunities in legal markets have a significant impact on crime rates. As I mentioned, more and better jobs in low-skill sectors reduce crime. When the labour market successfully absorbs people at risk of turning to crime, or of reoffending, crime rates drop.

However, I do believe the impact of support programmes for released prisoners, whether run by the state or by other organisations, is often overestimated. The same applies to in-prison programmes aimed at training or educating inmates. Of course, there are valid reasons to offer these programmes, especially given the harsh conditions faced by current and former inmates. But it is a common mistake to assume they are enough to lower recidivism; the evidence often shows they have little or no effect. Unless they are paired with genuine integration into the legal labour market, they are not enough.

The Confluence: What's the biggest misconception people, including policymakers, have about what reduces crime?

Dr. Tealde: The idea that simply increasing police numbers will lower crime rates, and that this applies to any crime, is flawed. Police presence acts as a deterrent for crimes such as street-level property offences; for instance, having more officers on the beat makes it harder for pickpockets to operate, as they actively seek out victims in public spaces. Deterrence is effective when the offender does not know the victim beforehand.

The situation changes completely, however, when the victim and the offender already know each other. In Uruguay, as in many countries across the Americas and, more recently, in Europe, the crimes causing the greatest public alarm are murders linked to drug trafficking: specifically, executions planned and carried out by contract killers. Such crimes cannot be prevented merely by putting more police on the streets.

The Confluence: If you could get one policymaker in Uruguay to change one thing based on your research, what would it be?

Dr. Tealde: I would shift the focus away from placing such heavy emphasis on reducing reoffending. Of course that is important, but it is not the only thing that matters.

I would focus instead on young people who are at risk of falling into a life of crime but have not done so yet: those whose circumstances put them at risk, or make it difficult for them to enter the labour market on favourable terms. I would design policies to ensure that this group does not see crime as a way out of their current reality.