DNA Registration In Denmark Seems To Greatly Reduce Revidivism

I'm skeptical that this result would replicate to the very different society and criminal justice system found in the U.S. from the results in Denmark. But even a much smaller effect size from the same strategy of creating  a DNA database of felons would still be worth it.

This paper studies the effects of adding criminal offenders to a DNA database. Using a large expansion of Denmark’s DNA database, we find that DNA registration reduces recidivism within the following year by up to 42%. It also increases the probability that offenders are identified if they recidivate, which we use to estimate the elasticity of crime with respect to the detection probability and find that a 1% higher detection probability reduces crime by more than 2%. We also find that DNA registration increases the likelihood that offenders find employment, enroll in education, and live in a more stable family environment.

Anne Sofie Tegner Anker, Jennifer L. Doleac, and Rasmus Landersø, "The effects of DNA databases on the deterrence and detection of offenders" (April 2020).

Hat tip to Yglesias via Fully Myelinated.

From the body text of the paper:

[W]e measure the effects of a 2005 Danish reform that increased offenders’ probability of being added to the DNA database from 4% to almost 40%. The change allowed police to add anyone charged with what is roughly equivalent to a felony in the U.S. (which is the relevant policy margin for most U.S. states considering database expansions), increasing offenders’ average probability of being included in the DNA database dramatically. Using the database expansion as an exogenous shock to the likelihood of DNA registration, we estimate that being added to the DNA database reduces recidivism by 6.5 percentage points (42%) in the first year (p < 0.01) – a deterrence effect persisting for at least three years. 

Using the rich Danish register data, we are further able to explore heterogeneity in effects of DNA registration by previous criminal history, age, and family structure. We find statistically significant deterrence effects for all groups except older offenders. The effects of DNA registration are larger for first time offenders, offenders with children, and offenders initially charged with violent crime, while DNA databases prevent subsequent property, weapon, and violent crime, which supports the hypothesis that offenders frequently commit multiple types of crime instead of specializing in one specific type. 

In addition, we find that DNA registration has beneficial effects on subsequent employment, education, and family life. Young offenders are more likely to enroll in education while older offenders are more likely to be employed if they are in the DNA database. Also, first-time offenders are more likely to be married after they are added to the DNA database, and recidivists are more likely to be with the same partner and to live with their children, at least in the short run. These findings are consistent with the hypothesis that keeping people out of trouble (and out of prison) can put their lives on a more positive track. 

Table 1 compares the characteristics of felons before and after the 2005 reforms in Denmark (which are generally speaking similar before and after the reforms).

Table 6 sets forth the core conclusions of the paper.

 

Table A-2 shows that the more serious the crime, the more inclusion in the DNA registry discouraged felons from reoffending by committing it.

Recidivism was extremely low, however, for federal prisoners who received early release due to COVID-19. As the Washington Post reports:

To protect those most vulnerable to covid-19 during the pandemic, the Cares Act allowed the Justice Department to order the release of people in federal prisons and place them on home confinement. More than 11,000 people were eventually released. Of those, the Bureau of Prisons (BOP) reported that only 17 of them committed new crimes.

That’s not a typo. Seventeen. That’s a 0.15 percent recidivism rate in a country where it’s normal for 30 to 65 percent of people coming home from prison to reoffend within three years of release.

Of those 17 people, most new offenses were for possessing or selling drugs or other minor offenses. Of the 17 new crimes, only one was violent (an aggravated assault), and none were sex offenses.

This extremely low recidivism rate shows there are many, many people in prison we can safely release to the community. These 11,000 releases were not random. People in low- and minimum-security prisons or at high risk of complications from covid were prioritized for consideration for release.

Better distinguishing between offenders with high and low risks of recidivism and adjusting sentences accordingly could greatly reduce mass incarceration in the United States. The most important factor was that the early released prisoners were mostly old or sick.