Research

This article provides evidence that exposure to gang control during incarceration pushes individuals onto a path of organised criminal involvement. We use an IV design that exploits short-run variation in exits from prison galleries to identify the effect of initial assignment to gang-controlled environments. When relatively more vacancies open in galleries controlled by criminal organisations, newly incarcerated young men are 9.8% more likely to be assigned to them. Leveraging intelligence-based mapping of prison galleries and linked administrative data for 2017-2025, we find that initial assignment to a gang-controlled gallery increases the share of incarceration time spent under gang control by 76%, and raises the probability of re-incarceration within two years of release by 61% relative to individuals first assigned to neutral galleries. The effects are stronger for individuals with weak or no prior links to organised crime networks. Mechanism evidence shows that exposed inmates are less likely to work, more likely to be involved in in-prison incidents and criminal occurrences, and become progressively more disconnected from their families through fewer visits while in prison. These findings suggest that gang-controlled prison environments reduce rehabilitative opportunities and strengthen the conditions for continued criminal involvement during and after incarceration. 

Can criminal law prevent domestic violence and change how victims use the state? We study Brazil’s 2015 recognition of femicide as qualified homicide using ten years of linked population, police, hospital, and prison records for 2.6 million women. Exploiting predetermined municipal exposure to men with histories of violence against women, we find that police-recorded non-fatal domestic violence fell disproportionately in more exposed areas. Conditional on an incident reaching police, requests for protective measures rose, while pursuit of charges did not. Legal recognition at the lethal endpoint can therefore affect both violence below that margin and survivors’ choice of state remedy.

When minimum wage increases impose a cost shock on employers of low wage workers, there are a variety of ways in which firms can adjust. Rather than study the main focal point of much minimum wage research, possible labour demand adjustment, this paper considers a more understudied angle. It examines whether firms can offset the cost shock through changing non-wage aspects of work related to the nature of work. This includes altering employment composition in the workplace, the use of alternative work arrangements and redefining job contracts. Whether these alter in response to minimum wages is studied through the lens of the UK’s 2016 National Living Wage (NLW) introduction. In terms of traditionally studied outcomes, the NLW boosted worker wages, but with no change in total employment. Instead, firms did indeed adjust operations through changes in employment composition and by altering employment contracts. These non-labour demand adjustments of employment relations show how employment stability can be maintained in response to minimum wages as employers can restructure work through within-firm job composition and contracts. 

We study the role of job-to-job transitions in shaping wage dynamics in the United Kingdom. We show that job moves have continued to generate substantial wage gains for movers even during a period of real wage stagnation and declining labour market mobility. Using an econometric model that accounts for worker, firm, and match heterogeneity, we decompose the job mobility premium and find a limited role for worker and firm effects, with most of the return explained by match-specific factors. We also examine the determinants of job mobility and show that much of its recent decline reflects compositional changes in the workforce, particularly ageing and increasing occupational professionalisation. We conclude by discussing the policy challenges of promoting job mobility as a driver of wage growth and more efficient labour market reallocation.

This paper studies consequences of the large exchange rate depreciation occurring when the UK electorate unexpectedly voted to leave the European Union. Sterling plummeted, recording the biggest one-day depreciation of any of the world's four major currencies since Bretton Woods. The prospect of Brexit happening generated sizable differences in how much sterling depreciated against different currencies. Coupled with pre-referendum cross-country trade patterns, this generated variations in exchange rates facing businesses in different industries. The paper offers evidence of a cost shock from the prices of intermediate imports rising by more in higher depreciation industries, but with no revenue offset from exports. Workers were impacted by these increased cost pressures, not in terms of job loss but through relative real wage declines in higher depreciation, larger cost shock industries. This resulted in an aggregate fall in real wage growth of 3 to 3.6% cumulatively over the three years after the referendum.

We provide a unifying empirical framework to study why crime reductions occurred due to a sequence of state-level dropout age reforms enacted between 1980 and 2010 in the United States. Because the reforms changed the shape of crime-age profiles, they generate both a short-term incapacitation effect and a more sustained crime-reducing effect. In contrast to previous research looking at earlier US education reforms, we find that reform-induced crime reduction does not arise primarily from education improvements. Decomposing short- and long-run effects, the observed longer-run effect for the post-1980 education reforms is primarily attributed to dynamic incapacitation.

Do compulsory schooling laws reduce crime? Previous evidence for the US from the 1960s and 1970s suggests they do, primarily working through their effect on educational attainment to generate a causal impact on crime. In this paper, we consider whether more recent experience replicates this. There are two key findings. First, there is a strong and consistent negative effect on crime from stricter compulsory schooling laws. Second, there is a weaker and sometimes non-existent link between such laws and educational attainment. As a result, credible causal estimates of the education–crime relationship cannot in general be identified for the more recent period, though they can for some groups with lower education levels (in particular, for blacks).