Do recessions increase crime?

Recessions typically lead to an increase in youth unemployment rates, leaving young people to face more difficulties in finding jobs. Concerns about long-term impacts on the youth arise when youth unemployment rates are very high. One prominent example is the current policy debate about the lost generation in the context of the Great Recession. We know, for instance, that unemployment at young ages can have persistent negative effects on future wages (e.g. Oreopoulos et al. 2012) and career progression (e.g. Oyer 2008). In recent research (Bell et al. 2015), we show that recessions have a more disturbing and substantial impact on initiating and forming criminal careers:

Young people who leave school during recessions are significantly more likely to become involved in crime than those who leave school while labour markets are more buoyant.

Why do recessions at labour market entry matter for crime?

So, why is it that youth who graduate during recessions are more likely to engage in crime? Those who leave school during a recession, when youth unemployment rates are particularly high, struggle to find a job but do not yet have financial insurance. Hence, following the notion from Gary Becker’s seminal work on criminal choices (Becker 1968), low expectations on returns to legal activity may lead to initial involvement in crime and subsequently to a first encounter with the criminal justice system. Knock-on effects can then lead to criminal careers for the young. On the one hand, those who initially get involved in criminal activity learn the ‘criminal know-how’ (Mocan et al. 2005). On the other hand, those who have criminal records early on in their career may reduce their job opportunities and expected returns in the legal labour market (see, for example, Baert and Verhofstadt 2015).

There is a substantial body of criminological evidence that illustrates the importance of the youths’ experience for understanding crime. In particular, research has established a strong age-crime pattern with crime rates typically peaking during the late teenage years. Yet, crime is not only a feature of the teenage years – crime rates decrease with age but do not disappear subsequently. That suggests that there is an initial effect but criminal activity is somewhat persistent over the life cycle.

What is the empirical evidence?

Can that persistence be explained by the long-term impact of recessions? A typical recession leads to a 5 percentage points higher than normal unemployment rate. What is the long-term impact of graduating into such conditions? Our empirical analysis of the link between crime and unemployment at labour market entry is based on a variety of US and UK cohort and individual level data sources.

We exploit cohort level data for both countries to estimate the average effect of initial labour market conditions on criminal activity of cohorts that enter the labour market at different points in time, taking into account differences in cohort composition. For the US, we use data on arrest rates from the FBI Uniform Crime Reports and from the Current Population Survey.

We find that the average arrest rate for a cohort entering the labour market during a recession is 10.2% higher than for an otherwise similar cohort entering a more buoyant labour market.

For the UK, we use data on conviction rates from the Offenders Index Database and the Police National Computer.

We find that the average conviction rate for a cohort entering the labour market during a recession is 4% higher than for an otherwise similar cohort. These effects are statistically significant in both cases.

Moreover, when we allow subsequent unemployment rates over the cohort’s life cycle to enter the regression, we find that what matters is indeed the unemployment rate at labour market entry. That finding contributes to understanding a puzzle in the literature: The overall link between crime and unemployment has been found to be relatively weak. Our findings suggest that the key impact of unemployment on crime is the early experience of unemployment rather than the average unemployment experienced over the cohort’s life cycle.

Interpreting the empirical evidence

How should we understand these results? The above interpretations refer to the effect on the average cohort arrest and conviction rates. Presumably, a large share of individuals in a cohort never gets involved in crime – independent of initial labour market conditions. A much smaller share is actually at the margin of becoming criminal, and in that respect affected by economic conditions when first entering the labour market. The average effect outlined above is a combination of the effects for both groups, meaning that the effect on individuals at the threshold to criminal activity may be even more substantial than the average effect suggests.

Studying individual level data adds to that picture. Using data from the US decennial Census and the American Community Survey, we find that entering the labour market during a recession increases the probability of being incarcerated at some point over the next two decades by 5.5% for the US case. Using data on self-reported arrests from the British Crime Survey and taking into account an extensive set of individual characteristics, we find that entering the labour market during a recession is associated with a 5.7% increase in the probability of ever being arrested in life. These effects are statistically significant. In both cases, the effects are even stronger for individuals who leave education early and are less qualified for the labour market, which confirms the notion of the average effect being a conservative estimate.

What about the persistence of the effect of recessions and the forming of criminal careers? Figures 1a and 1b show the estimated effects by year since labour market entry both for the US and the UK. Albeit slightly decreasing, the effect of entering the labour market during a recession is indeed persistent. The initial unemployment effect pushes some individuals towards a criminal career, implying long-lasting sizeable effects.

Figure 1a. Entry unemployment effects by experience, US

Note: The chart shows the estimated coefficients and 95% confidence intervals when we allow the coefficient on initial unemployment to vary by years of potential experience. Years of potential experience are years since labour market entry.

Figure 1b. Entry unemployment effects by experience, UK

Note: The chart shows the estimated coefficients and 95% confidence intervals when we allow the coefficient on initial unemployment to vary by years of potential experience. Years of potential experience are years since labour market entry.

Concluding remarks

Our results have important implications that contribute to the current policy debate about the long-term impact of recessions on youth. We demonstrate a disconcerting and long-run effect of economic downturns. Recessions not only lead to short-term negative outcomes on the labour market but can indeed produce career criminals. We find robust evidence of an initially strong and eventually long-lasting detrimental effect of entering the labour market during a recession for individuals at the threshold of criminal activity. These effects are economically substantial and potentially more disturbing than short-run effects.

This article is published in collaboration with VoxEU. Publication does not imply endorsement of views by the World Economic Forum.

To keep up with the Agenda subscribe to our weekly newsletter.

Author: Brian Bell is a Research Fellow, LSE Centre for Economic Performance. Anna Bindler is a PhD candidate in Economics, University College London. Stephen Machin is a Professor of Economics, University College London; Research Director at the Centre for Economic Performance, LSE; CEPR Research Fellow.

Image: Scales of Justice. REUTERS/Stephane Mahe