The myth of "evidence-based" rehabilitation programs

The campaign to pass leniency-for-federal-felons legislation — a campaign that culminated with the passage (thanks to Donald Trump) of the First Step Act, had to counter a big problem: Prisoners released from jail commit crimes at an alarmingly high rate.

According to one study, around 75 percent of prisoners are rearrested within five years of release. Around 60 percent are convicted. Around 50 percent return to prison. After ten years, the percentages are around 80, 70, and 60.

Thus, as Bill has said, shorter sentences and early release mean more crime, sooner.

To get around this reality, advocates of leniency touted prison rehabilitation programs as the answer to recidivism. Over-and-over, they told us that “evidence-based” rehabilitation programs succeed, and that therefore prisoners who take advantage of them can be released early with little risk to public safety. And, they added, early release provides just the incentive needed for prisoners to take advantage of these wonder-working programs.

But reformers have focused on how to rehabilitate criminals for 200 years without achieving much success. To believe that the current crop of do-gooders has finally cracked the code is the kind of hubris characteristic of liberals.

Strangely, though, many conservatives and conservative institutions (notably, the Heritage Foundation), also adopted this belief. And it was Donald Trump whose support pushed the First Step Act past the finish line.

Was there an evidentiary basis for putting faith in “evidence based” rehabilitation programs? Not much of one. Some of the alleged evidence came from state programs — Texas, in particular. That evidence has been debunked. It did not make an apples-to-apples comparison of recidivism rates.

Some of it came from Scandinavian countries, aka “the Nordic rehabilitation paradise.” In Norway, U.S. News tells us, “the prison system boasts a 20 percent recidivism rate – compared to the 76.6 percent recidivism rate in the U.S.” According to the same source, this Norwegian rehabilitation success story inspired North Dakota to adopt Norwegian practices.

And why not? If Norway has discovered the secret to keeping criminals from re-offending, why shouldn’t the U.S. follow its lead and adjust our sentencing practices accordingly?

The problem is that the Nordic success story appears to be a myth. So argues this evidence-based — indeed, evidence-packed — article.

The author contends that when one makes an apples-to-apples comparison of recidivism rates, Norway and other Scandinavian countries do not perform much better — and maybe not better at all — than the U.S.

The first step (so to speak) in the analysis is to measure recidivism the same way for the countries being compared. No good-faith analysis can do otherwise.

According to the author:

The American 76.6% figure was based on rearrest within 5 years (Durose et al. ,2014), whereas the Norwegian 20% figure described the number who received a new prison sentence or community sanction that became legally binding within 2 years (Kristoffersen, 2013).

Of the American recidivism statistics. . .the 28.8% incarceration figure [within two years] is arguably the most comparable in definition to that of the 20% Norwegian figure. Thus, when the comparison is closer to apples-to-apples, the difference between Norway and the United States is far more modest.

And that’s not all:

The distribution of crime types will affect the composition of sentenced offenders and therefore the reoffending rate. In general, observed recidivism rates are lower for sexual and traffic crimes, high for property crimes, and in the middle for violent and drug crimes (Yukhnenko et al., 2023).

Kristoffersen (2013) has shown how relevant this can be in the Nordic context. In all of Europe, Norway has been unique in the extent to which they have given prison sentences for crimes like speeding in traffic. Traffic crimes have low reoffending rates in all the Nordic countries, so the imprisonment of these individuals at low risk of reoffending artificially brings down the reoffending rate among released prisoners.

In contrast, (perhaps due to greater use of non-prison punishments) Norway had the smallest proportion of released offenders sentenced for thefts; and those sentenced for theft have the highest reoffending rates across the Nordic countries.

If all traffic offenders had instead been on probation rather than in prison, that alone would bring the Norwegian recidivism figure from 20% to 25%. That is even closer to the 28.8% American reincarceration figure cited above.

There’s more:

The base rate of crime and the size of the population who receive prison sentences in the country may also affect recidivism rates. Yukhnenko et al. (2023) provided evidence of this and they explain: “the more criminogenic a society is, the higher the recidivism rates (given other factors are held constant).” Thus, rather than reflecting rehabilitative differences, recidivism figures can also partly be ascribed to differences in the overall level of criminality. This is an important consideration because the United States has a far higher rate of serious violent crime than the Nordic countries.

In this connection, one of the comments to this article points out that the U.S. has about 30 percent more police per 100,000 people than Norway. More policing means that more ex-prisoners will be caught when they commit crimes, thus pushing recidivism our numbers up for reasons having nothing to do with not following Norwegian rehabilitation practices.

And consider this:

Foreign prisoners make up a disproportionate fraction of Norwegian prisoners. . . According to Aftenposten, 794 foreign convicts were deported and were formally barred from re-entry in 2011. The following year 1,019 foreign convicts were deported. For context, the nationwide prison population in Norway is only about 3,000-4,000 (give or take) at any given time, so this number of deportees is not inconsequential.

Deported individuals obviously cannot contribute to recidivism statistics any further, except if they manage to return despite being barred from re-entry, which can happen (and when they do, they are often discovered due to recidivism). Excluding deportees could plausibly move [Norway’s] recidivism rate a few additional percentage points.

Thus, an evidence-based analysis shows that the Nordic rehabilitation success story is, indeed, a myth.

The author also examines data regarding the efficacy of rehabilitation programs in Scandinavia. The only intervention that has been effective, according to the author’s review of the data and literature, is the use of medication for psychiatric disorders. But such treatment is only applicable in specific circumstances and thus, says the author, “would only have modest effects on national recidivism rates.”

It seems, therefore, that the laundry list of “EVIDENCE BASED RECIDIVISM REDUCTION (EBRR) PROGRAMS” contained in the Biden Justice Department’s First Step Act Approved Programs Guide isn’t likely to curb recidivism in any meaningful way. Prisoners might gain something from taking courses in “English-as-a-second-language,” “Hooked on Phonics,” or “Money Smart for Adults,” but there’s little reason to believe these courses will make them less likely to commit crimes when they are released.

In sum, the First Step Act was sold through a myth — the myth that “evidence-based” rehabilitation programs will reduce recidivism significantly. Indeed, the very phrase, “evidence based,” is false advertising. People will die as a result of the fraud.

The First Step Act was a mistake — arguably the biggest policy plunder Donald Trump committed during his presidency. Let’s hope there’s no Second Step Act.