Misleading Statistics
Kevin J. Mullen
According to a just released study [1996]from the Center on Juvenile and Criminal Justice, a San Francisco non-profit policy group, African Americans are being imprisoned in California under the "three strikes" law at a rate 13 times that of whites. The reason for the disparity, its authors claim, is racial bias on the part of prosecutors who charge and try cases. Every few days or so, it seems, we are treated to the results of yet another "study" showing that white folks continue to beat up on people of color. It's all the work of a group of over educated and underemployed academics who have made a cottage industry funded by public money to dissect the racist white mainstream society.
The "three strikes" finding is the second shoe to be let drop from a study published by the Center a few weeks ago which found that youthful African American offenders in San Francisco were imprisoned "at 10 times the rate of blacks in [Apartheid era] South Africa." Then there was the 1994 study showing that despite the best efforts of the Clinton Administration to correct the situation, mortgage lenders were twice as likely to reject black loan applicants than whites from the same income levels.
There was another study released the same year which found that blacks and Hispanics constituted nearly all the accused murderers against whom the Justice Department sought the death penalty. Other studies showed that blacks who killed whites were far more likely to receive the death penalty than whites who killed blacks. A study prepared by a Brown University sociologist and released by the Office of Personnel Management in 1995 reported that African American federal employees were more than twice as likely as whites, Asians or Latinos to be dismissed. While the report's author stopped short of attributing the phenomenon strictly to discrimination, she did say that "We should proceed as if there might be racial bias involved."
No wonder there is so much free floating anger in our world today. Standing alone, the figures produced by the studies -- or at least the story headlines they produce in the press -- constitute a damning indictment of the mainstream society. But, as is often the case with selectively chosen statistics, there is more to the story than is at first evident.
When reports showing the disparity in loan approval rates began appearing the early 1990s, the Governor of the Federal Reserve Board cautioned that "while its findings might suggest discriminatory practices, the data reviewed did not include information on the applicants' credit history or other information that play into lending decisions," by any standards important factors in loan approval rates. Four years later, when similar disparate figures were reported, a representative of the American Bankers Association again pointed out that the report did not "take into account such factors as applicants' previous indebtedness and credit record."
Now there's nothing in evidence to demonstrate any group necessarily has a worse credit history than another, but you'd think that in the four year period between the two reports someone would check to find out whether that had anything to do with the disparities -- unless, of course, the conclusion is already foregone. Buried in the story about the disparity of "three strikes" sentencing of African Americans is the comment that "the report does not break them [the sentences] down by specific crimes such as murder. robbery, or sale of narcotics." Why -- one is compelled to ask -- not?
The authors of the study might have profited from a look at the U.S. Justice Department's "Bureau of Justice Statistics Sourcebook," but I doubt it. There they would have found that part of the reason blacks were incarcerated at a higher rate than whites had to do with the higher proportionate rate of violent crime committed by African Americans in today's world. Similarly, part of the explanation for the disparity between black and white death sentences lies in the fact that murders of whites by blacks are more likely to be accompanied by "special circumstances," e.g. during the commission of a robbery, than the reverse.
A look beyond the headline in the story about disparate dismissal rates of black federal employees reveals that race was only one of the five predictors explain the disparity. The principal area where the dismissal disparity occurred was for "misconduct" including "theft, embezzlement and insubordination and acts such as striking or throwing something at a supervisor." When these factors were removed, the dismissal rates were similar among the various groups.
In an effort to understand what is happening in our society, it would be reasonable to study why any group would be more be more inclined than another to throw things at supervisors. But to drive a wedge between groups by introducing distorted statistical analyses into a volatile racial mix is more than intellectually dishonest. It's downright dangerous.