80% of the worst of the worst will not reoffend

Sex offender recidivism rates below expectations: more than 80% of sex offenders who underwent treatment did not reoffend. (15-Year Prospective Study)

Author: Finn, Robert

More than 80% of sex offenders who have undergone treatment do not reoffend within 15 years, according to preliminary results of a 15-year prospective study of 626 individuals reported at the annual meeting of the American Academy of Psychiatry and the Law.

The study did not examine the efficacy of various forms of treatment, but it did find that sexual offenders who were compliant with their treatment were less likely to reoffend than those who were noncompliant.

The results of the study were reported by Dr. Fred Berlin, who is director of the National Institute for the Study, Prevention, and Treatment of Sexual Trauma, Baltimore, and by Gerard J. McGlone. Ph.D., a clinical fellow at Johns Hopkins University, Baltimore.

The results are "a far cry from what people often think is occurring, particularly with a group of patients who are this serious, people with multiple prior offenses, people with pedophilia," Dr. Berlin said. "The notion that most of these people would quickly be back into trouble was simply not true."

The preliminary 15-year data update the group's 5-year study of the same individuals (Am. J. Forensic Psychiatry 12[3]:5-28, 1991). That study found an overall 9.7% recidivism rate.

The estimated rate of additional recidivism for the following 10 years was 9.3%, for a total of 19% over 15 years.

Both studies followed 626 individuals treated at the Johns Hopkins sexual disorders clinic between 1978 and 1990. Of those, 406 were pedophiles, 111 were exhibitionists, and 109 were characterized as "sexually aggressive," with most diagnosed as "paraphilic disorder not otherwise specified" or antisocial personality disorder.

The patients underwent a variety of treatments at the clinic, including individual, group, and family therapy. Approximately 40% of the patients received an-tiandrogenic drugs in an effort to lessen their sex drive.

Investigators examined three local and national databases for evidence of recidivism in the 5-year study.

Anyone who was charged with a sexual crime was included; conviction was not necessary.

For the 15-year follow-up, investigators have so far examined only Maryland state databases and have not yet examined national databases.

The 5-year study showed that 7.4% of pedophiles, 4.6% of sexual aggressives, and 23.4% of exhibitionists reoffended sexually. The 15-year study added an additional 8.6% of pedophiles (total 16%), 10% of sexual aggressives (total 14.6%), and 10.8% of exhibitionists, (total 34.2%), the investigators said.

For the 15-year study, investigators have not yet reported recidivism rates for compliant vs. noncompliant patients. These data were reported at the 5-year follow-up.

For pedophiles, 2.9% of the treatment compliant and 11.1% of the noncompliant reoffended sexually. For sexual aggressives, 2.8% of the compliant and 8.1% of the noncompliant reoffended sexually

And for exhibitionists, 12.5% of the compliant and 32.6% of the noncompliant reoffended sexually

On the issue of exhibitionists, Dr. Berlin observed, "Even with treatment, this is a very driven, compulsive disorder. The good news, though, is that in almost every instance where there was recidivism, it was a non-hands-on offense.

"These people were not recidivating by doing other things. With very few exceptions they were exposing. ... The more dangerous patients, the ones who had been sexually aggressive or who were pedophiles, had the lowest recidivism rates.

Nonsexual recidivism rates were considerably higher in all groups. At the 5-year follow-up, 9.2% of compliant pedophiles and 26.2% of noncompliant pedophiles were charged with a nonsexual offense such as burglary. Among sexual aggressives, the rates were 14% of compliant patients and 29.7% of non-compliant patients. And among exhibitionists the rates were 10.4% of compliant patients and 4 1.9% of non compliant patients.

The study's results have important public policy implications, Dr. Berlin said.

Over the years, for example, legislatures in all 50 states and the District of Columbia have passed numerous laws involving the civil commitment, registration, and community notification of sexual offenders.

In addition, a federal law that took effect recently requires the states to notify the public about sexual offenders who are enrolled or working at college campuses.

The idea behind these registries and public notifications is the notion that sexual offenders are difficult to rehabilitate and have a particularly high rate of recidivism. (See related story page 5.)

But the data show that sexual offenders actually have a lower rate of recidivism that those who commit other kinds of crimes.

And some experts speculate that these new laws just might have the unintended consequence of actually increasing the risk of recidivism.

"We think that many of the people who seem to be doing well in our treatment program have done well because they could get jobs," Dr. Berlin said. "They could fit into their community.

"We now have these registries. Are [sexual offenders] who have been punished and released going to be able to get housing? Are they going to be able to get work? Are we inadvertently actually going to make it more difficult for some of these people?"

The Safe Return of Offenders to the Community Statistical Overview April 2005

The Safe Return of Offenders to the Community Statistical Overview April 2005 (Canada) (HTML) (PDF)

B1a. Homicide Offences

* Homicides account for less than 1% of all violent crime in Canada and have decreased significantly since 1991 - there were 754 homicides in 1991 versus 548 in 2003.6

Homicide in Canadian and American Cities

Despite having incarceration rates that are 5 to 6 times higher, American cities are much more dangerous than comparable Canadian cities. Seven U.S. and Canadian cities were selected to compare homicide and violent crime rates. Cities were matched for general similarity in size and/or geographical location.

* In all cases, the homicide rates of the American cities are all higher than their Canadian neighbors, and this would generally be true for any sample of larger American versus Canadian cities.
* Between the matched pairs in the study, the difference in homicide rates is usually quite large (e.g., between Chicago - 20.6 per 100,000 - and Toronto - 1.8 per 100,000).

Homicide rates (per 100,000),2003.......Population......Rate
Vancouver...............................................2,126,111.......2.1
Seattle....................................................576,296.........5.9
Calgary..................................................1,023,666.......1.1
Denver...................................................565,905.........11.1
Winnipeg................................................688,746.........2.6
Minneapolis............................................378,602.........12.1
Toronto..................................................5,118,992.......1.8
Chicago..................................................2,898,374.......20.6
Ottawa-Carleton......................................866,621.........1.1
Whashington DC......................................563,384.........44.0
Montreal..................................................3,586,221.......1.6
Philidelphia..............................................1,495,903.......23.3
Halifax.....................................................377,932.........0.8
Norfolk.....................................................242,077.........16.5
Canada....................................................31,629,677......1.7
United States............................................290,809,777.....5.7

Source: Uniform Crime Reporting Survey, Canadian Centre for Justice Statistics, Statistics Canada. Uniform Crime Reports, Bureau of Justice Statistics, Federal Bureau of Investigation.
Note: Crime data are based on reports for municipal police forces, not the wider Census Metropolitan Area populations. (p

As a proportion of all crimes reported in the 2003 Uniform Crime Reporting survey, released federal offenders re-admitted with a new conviction were therefore responsible for just over 1 of every 1,000 federal statute offences reported to police in 2003, including:

* 1.3 of every 1,000 violent offences;
* 0.6 of every 1,000 sexual offences;
* 1.0 of every 1,000 drug offences;
* 0.9 of every 1,000 property or other federal statute offences.

As a proportion of all convictions, released federal offenders re-admitted with a new conviction were responsible for about one percent of criminal convictions in Canada.