Electronic Monitoring in Sweden
Introduction
Grenet et al. (2024) studies the nationwide expansion of electronic monitoring (EM) in Sweden in 1997, which allowed offenders sentenced to up to three months in prison to substitute incarceration with EM. Under the EM program, offenders are required to stay at home unless pre-approved to participate in school, work, or treatment programs. Participants were required to have suitable accommodation, stable employment (or commitment to seek employment), and were subject to home visits and alcohol and drug testing. The program was voluntary, with offenders needing to apply to the Swedish Prison and Probation Service following their conviction.
The analysis employs a difference-in-differences strategy to compare the outcomes of offenders who received prison sentences of up to three months (treatment group) to offenders sentenced to 4-12 months (control group) before and after the 1997 reform. The data comes from Swedish administrative registers covering the entire Swedish population aged 15 and above from 1990 to 2016, linked to conviction and crime suspicion registers.
For the purposes of the MVPF calculation, the paper reported figures in 2023 prices and converted to US dollars using the then-current exchange rate of 10.22 SEK/USD.
Pays for Itself
Net Cost
The net cost of electronic monitoring includes the direct fiscal savings, additional income tax due to earnings gains, and savings from reduced recidivism.
According to the Swedish Prison and Probation Service, the daily cost per client in the EM program was $113 in 1998, compared to $184 for placement in a minimum-security prison. With the average prison sentence for the treatment group being approximately 50 days, the cost of an offender placed on EM was $5,650 (50 × $113), while incarceration would have cost $9,200 (50 × $184). This results in direct fiscal savings of $3,550 per EM admission.
The pre-tax earnings gains are $13,509 per offender and $5,283 for offenders’ children. The paper approximates income tax as 30 percent of gross earnings. Thus, the government earns an additional 0.3*$18,792 = $5,637.6 in income tax revenues per EM admission.
In addition, the paper calculates further cost savings from reduced recidivism. This policy prevents 0.047 re-convictions per EM admission which translates to 0.29 crimes when adjusted for the 16% crime clear-up rate. The cost of an additional crime is modelled as the cost of incarceration as well as the cost of police, fire, and victim services.
The costs of police, fire and victim services are taken from Miller et al. (1996). For a conservative estimate, the paper assumes violent crimes are assaults and property crimes are larceny. The police, fire, and victim services costs sum to $160, $171, and $84 for assault, larceny, and drunk driving respectively (adjusted to 2023 dollars). Each type of crime is then weighted by the fraction of crimes in the treatment group that fall under each category. Thus, the savings from police, fire and victim services is .29*(.214*$160 + .187*$171 +.348*$84) = $95. Hence, the net cost component from reduced recidivism is .29*($9,200 + $95) = $2,695.
When accounting for direct fiscal savings, income tax gains, and indirect benefits from reduced recidivism, the net cost for the government is -$3,550 – $5,637.6 – $2,695 = $11,883.
Willingness to Pay
The EM program led to a total earnings increase of $13,509 per offender over the three-year follow-up. For children of offenders (aged 11-14 at the time of their parent’s trial), the program increased their earnings at age 25 by an estimated $11,739 annually. Assuming these increased earnings are observed for three years between ages 25-28, and that the average number of children aged 11-14 per offender is 0.15, this adds $5,283 ($11,739 × 3 × 0.15) per EM admission. The paper approximates income tax as 30 percent of gross earnings. Thus, the total after-tax earnings gains are 0.7*($13,509 + $5,283) = $13,154 per EM admission.
In addition, the calculation includes the willingness-to-pay of potential victims to avoid crimes. This policy prevents 0.047 re-convictions per EM admission which translates to 0.29 crimes when adjusted for the 16% crime clear-up rate. For a conservative monetary value, the paper assumes violent crimes are assaults and property crimes are larceny. Miller et al. (1996) provides estimates of various costs associated with assault, larceny, and drunk driving. The willingness-to-pay estimate includes the lost victim productivity, medical care, mental health, and quality-of-life costs. These sum to $19,513, $30, and $34,132 for assault, larceny, and drunk driving respectively (adjusted to 2023 dollars). Each type of crime is then weighted by the fraction of crimes in the treatment group that fall under each category. Thus, the total willingness-to-pay of potential victims for reduced recidivism from violent crimes, property crimes, and driving while intoxicated is 0.29*(0.214*$19,513 + 0.187*$30 +0.348*$34,132) = $4,657.21.
Summing increased earnings for offenders and their children and avoided victim costs due to reduced recidivism leads to a total willingness-to-pay of $13,154 + $4657 = $17,811.
MVPF Summary
As the electronic monitoring program has a positive willingness to pay and negative net costs to the government, the MVPF is infinite.
The exact construction of the MVPF presented here is modified from the calculation presented in the published paper. Despite the differences, both methods lead to the conclusion that the MVPF is infinite, and the policy pays for itself.
References
Grenet, Julien, Hans Grönqvist, and Susan Niknami (2024). “The effects of electronic monitoring on offenders and their families.” Journal of Public Economics, 230:105051. DOI: https://doi.org/10.1016/j.jpubeco.2023.105051
Miller, Ted R., Mark A. Cohen, and Brian Wiersema (1996). “Victim Costs and Consequences: A New Look.” National Institute of Justice Research Report, NCJ-155282, U.S. Department of Justice. https://www.ojp.gov/pdffiles/victcost.pdf
Policy Features
- Category
- Criminal Justice
- Sub-Category
- Recidivism Programs/Counseling
- Beneficiary Type(s)
- Adults
- Average Age
- 34
- Country of Implementation
- Sweden
- Year of Implementation
- 1997
- Empirical Method
- Difference in Differences
- Research Type
- Primary
- Peer Reviewed
- Yes
- MVPF Publication Link
- doi.org/10.1016/j.jpubeco.2023.105051