Skip to Navigation
University of Pittsburgh
Print This Page Print this pages

July 6, 2006

Taking a look at race, crime & communities:

Why are there racial differences in convictions?

What accounts for the racial imbalance in the number of arrests and convictions in the U.S. criminal justice system?

What do recent trends in crime rates say about the effectiveness of incarceration as a deterrent to serious crime? What do those trends predict for the near future?

Two prominent local researchers weighed in on these and related questions as part of a day-long discussion June 29 titled “Race, Crime and Communities,” sponsored by the School of Social Work’s Center on Race and Social Problems.

Rolf Loeber, Distinguished Professor of Psychiatry at Western Psychiatric Institute and Clinic, is the principal investigator on the Pittsburgh Youth Study, a longitudinal assessment study of 1,500 boys for risk factors and mitigating “promotive” factors in the progression of violent behavior. He discussed what that study, begun in 1987, and other longitudinal studies show about the relationship of risk factors to serious delinquency and violence.

Alfred Blumstein, J. Erik Jonsson Professor of Urban Systems and Operations Research at CMU’s Heinz School of Public Policy and Management and director of the National Consortium on Violence Research, presented some thoughts on recent national violent crime trends and offered some policy considerations to combat a potential new rise in violent crime.


“As a psychologist I’m interested in social interactions that actually give rise to crime,” said Rolf Loeber, who also is a professor of epidemiology in the Graduate School of Public Health and a professor in the Department of Psychology. “I’m also interested in the issue of disproportionate minority contact, or DMC, in the justice system, and what accounts for that.”

According to Loeber, the key questions are:

• Which risk factors are shared between the races and which are not?

• Is there a “dose-response relationship” between the number of risk factors and delinquency, and is that relationship the same for different racial groups?

• To what extent do risk factors explain the disproportionate minority contact with the criminal system?

To focus in on these questions, Loeber presented data comparing studies of young males.

He compared the middle cohort of the Pittsburgh Youth Study, that is, a random sample of 508 boys ages 10 & 11, 53 percent of whom were African American, with a group of 411 boys 8 and 9 years old in a low socio-economic area of London who were almost all Caucasian.

In both studies, data were collected at six-month intervals — more than a dozen times over nearly 20 years — from the boys, their mothers, teachers and peers, and from official juvenile court records.

Both groups showed some comparable risk factors that were predictive of violence and criminal activity, including hyperactive or impulsive behavior; low academic achievement; poor supervision by the parent, low socio-economic status and coming from a single-parent family.

“But we found large ethnic differences in delinquency rates between African-American boys and the Caucasian boys in both Pittsburgh and London,” Loeber said. “Among African Americans, 53 percent had been convicted of a crime between age 10 and 17, compared to 26 percent of the Caucasian boys in Pittsburgh and 25 percent of all the boys in London.”

The comparison revealed that there is a dose-response relationship between the number of risk factors among the 20 factors shared by the studies and the probability of later conviction.

“In the London sample, those with between 0-9 factors had relatively few convictions, while having 10 or more raises the rate of convictions to 55 percent,” Loeber said. “For Caucasian boys in the Pittsburgh sample the results were very similar.”

There also is a strong indication of a dose-response relationship in the African-American group, “but the picture is quite different,” he said. “The rate of convictions goes up higher as risk factors are added, but the beginning point is also much higher. So even at low levels of risk, there are higher levels of conviction.”

Accounting for those racial differences, in part, is that fact that in Pittsburgh, there is a geographic concentration of African Americans in neighborhoods marked by low-income public housing and low socio-economic status. “Using the whole Pittsburgh sample, dose-response risk factors rise as you move from higher socio-economic strata areas to lower ones. The higher the degree of social disadvantage, the higher the likelihood of violence and crime,” Loeber said.

Similarly, protective or promotive factors that are associated with a decreased probability of crime or violence — no domestic violence in the home; good supervision by parents; a boy’s involvement in family activities; low parental stress, and parental reinforcement of pro-social behavior outside the home — are more common in high socio-economic neighborhoods.

So the next question is, How many risk factors are there in principally African-American neighborhoods?

“Black boys in Pittsburgh experience 7.3 risk factors on average compared to 4.4 on average for Caucasian boys,” Loeber said. “These risk factors in their direct environment are increased both in terms of their family conditions and the neighborhoods themselves and also in their school performance factors. Also, African-American boys are three times more likely to become victims as Caucasian boys.”

While the dose-response relationship of risk factors to violent crime is one piece of the puzzle, it does not explain completely the disproportionate numbers of minorities in contact with the justice system, he noted.

Related explanations include:

• That higher conviction rates are a function of racial bias in the justice system.

“It is plausible, but there is very little research on that subject,” Loeber said.

• A second explanation is that minorities also have a higher rate of self-reported crime.

“So if one racial group commits more crime and admits doing so, the probability of arrest and judicial processing will be higher. There is some data to support that,” he said.

• Third, minorities are exposed to a much higher level of crime-causing risk factors.

“So, both the higher rate of self-reported delinquency and the higher exposure of risk factors actually can explain this huge difference in arrests and justice processing,” Loeber said. “Once these differences are taken into account the disproportionate minority contact with the system actually disappears.”

In conclusion, he offered a provocative speculation: “Is DMC — disproportionate minority contact — the result of double jeopardy?” he asked.

“First, we see that the higher exposure to risk factors produces a high probability of delinquency and violence, that there is a dose-response relationship. So, the higher the number of risk factors the more we can explain the DMC. But one way of looking at these findings is saying that the same risk factors that lead to more crime also influence the decision-processing in the court, so the same factors actually operate and create this overrepresentation in the system.”

Imagine, he said, a young African-American man who comes into court having been in a serious fight. Compare that to a young Caucasian man coming to the same court.

“The chances are that there are large differences in the support system between these two men, that is, the Caucasian man may come from a two-parent family with a reasonable income, who can afford a lawyer up front and who can offer treatment in a facility outside the court as an alternative option to court processing. These options are not as likely to be available to the African-American man. The risk factors may reverberate twice through the system and contribute to DMC.”

—Peter Hart

Leave a Reply