Post#302
1000 words; 4 minutes to read
By: Emma Scott, MA Student at the School of Criminology at Simon Fraser University
In a 1998 aticle, American scholars John J. Donohue and Peter Siegelman proposed shifting resources from prisons to preschool programming. They suggested that it may be possible to reduce incarceration and use the resultant savings to invest in preschool enrichment programs to lower crime rates in the United States. Based on the evidence at the time, the authors concluded that preschool programming could be as effective in reducing crime as further increases in incarceration.
In a recent paper, Donohue and Siegelman revisited their 1998 article to see if their conclusions held up two decades later. Indeed, they found that their original conclusion, that preschool programming is a more cost-effective crime fighter than incarceration, is even stronger today. It is important to acknowledge that this data is based in the U.S.; however, it still offers a fruitful conversation about resource allocation and crime reduction in Canada. Let’s break it all down.
The Costs of Incarceration
Over time, the monetary cost of incarceration has grown substantially while its effectiveness has shrunk. In 1998, Donohue and Siegelman estimated that the annual cost per prisoner was $65,100 (in 2020 dollars). In 2024, the authors estimated the average annual per-inmate cost to be $83,950 (in 2020 dollars) – a 30% increase since the 1998 estimate. (Costs of imprisonment in Canada are generally higher.)
Incarceration also has indirect effects on society and the economy. For instance, the authors discussed lost productivity as a social burden. People who are incarcerated cannot contribute to the legitimate economy, and society loses out on the goods, services, and economic value they would have produced. The authors estimated that for every person in prison, the economy loses between $9,000 and $28,000 worth of productive work per year.
The authors also highlighted the potential for incarceration to increase recidivism, rather than deter it. Some scholars have found that jurisdictions which prioritize rehabilitation policies may see reductions in post-incarceration recidivism, whereas those that don’t may see increased recidivism. A recent meta-analysis makes a strong case for this as well, concluding that incarceration either has no effect on recidivism, or slightly increases it (see past blog post for a more detailed summary).
The Elasticity of Crime
Donohue and Siegelman used the concept of ‘elasticity’ to explain why incarceration has become less effective. In simple terms, elasticity is a mathematical measure of how much crime changes for each percentage change in the prison population. In 1998, the authors estimated an elasticity of -0.15, which meant that if you increased the prison population by 100% (meaning several million more people), you would expect a 15% drop in crime rates. The estimate has since dropped to -0.05, meaning society gets much less crime reduction for imprisoning people than it did in the past.
This decline is fueled by ‘diminishing returns’ – the idea that as you do more of something, each additional unit provides less benefit than the one before it. As the prison population expanded, more non-dangerous offenders were locked up for smaller property or drug-related offences. Removing these individuals from society had very little effect on the overall crime rate. Locking up non-dangerous people provided a significantly smaller public safety benefit even though the cost per prisoner remained the same or increased. Ultimately, incarcerating less dangerous people has proven to be ineffective and inefficient.
Preschool Interventions
Research over the last few decades has generally aligned with Donohue and Siegelman’s conclusion that preschool interventions can meaningfully reduce crime. It may seem strange to target crime prevention as early as preschool because these programs are potentially pricey and the effects aren’t immediately clear. However, preschool is not just about learning the ABCs. It is a critical period for children to build prosocial character traits to keep them out of trouble later in life.
Since the authors’ original 1998 article, evidence shows that large-scale programs have reduced subsequent criminal behaviour. Head Start, for example, is a government-funded program that provides education, nutrition, health, and family support services for children aged 3-5 in the U.S. It has been the focus of many studies looking into its effects on crime prevention and reduction. Access to Head Start has been shown to lower the probability of adult incarceration for low socioeconomic status children by anywhere from 25% to 100%.
Much Canadian research points to the same message. In 2021, Evan McCuish, Patrick Lussier, and Raymond Corrado published an analysis of a 20-year longitudinal study of ~1,700 young people incarcerated in British Columbia. This analysis has many important findings (see summary here). One notable takeaway is that we should promote crime desistance through assistance in childhood and adolescence. The authors call for allocating resources to both formal and informal intervention strategies to improve the chances of desistance and quality of life, which will impact youths’ prosocial skills and likelihood of crime involvement.
The Choice Becomes Clear
In their newer paper, Donohue and Siegelman drew 3 main conclusions that strengthened the findings of their original work. First, they confirmed that at least some model preschool programs have lasting effects on crime reduction in the U.S. Second, Head Start and other large-scale programs likely have similar crime-reducing effects. Lastly, preschool interventions, both small and large, have other positive effects on a child’s outcomes, beyond crime reduction. These include positive consequences related to adult smoking, employment, income, marriage, and adult mental health.
The bottom line is that shifting resources to preschool programming instead of prisons can be more cost-effective and cost-efficient. Prisons are more expensive and less effective than they used to be, while preschool interventions are shown to be more effective and economically sensible. The authors also observed that policy may finally be catching up to the evidence, because prison populations have declined and preschool enrichment programs have grown. In a society where tough-on-crime narratives continue to threaten and dismiss evidence-based crime prevention strategies, shifting resources becomes even more critical.
About this blog: The John Howard Canada blog is intended to support greater public understanding of criminal justice issues. Blog content does not necessarily represent the views of John Howard Canada. All blog material may be reproduced freely for any non-profit purpose as long as the source is acknowledged. We welcome comments (moderated).
Winner of 2025 Clawbies award for law blogs.
Back