The Color of Crime, a booklet by white separatist Jared Taylor, uses faulty analysis in its claims that whites are overwhelmingly victimized by blacks.
Around the nation, white supremacists and their fellow travelers are brandishing copies of a 1999 booklet that purports to show that whites have every reason to be terrified of blacks. For people from former Klansman David Duke to an array of neo-Confederates, The Color of Crime: Race, Crime and Violence in America has become a kind of Bible that shows them that they were right all along.
Sponsored by the New Century Foundation, an organ of white separatist author Jared Taylor, The Color of Crime is being circulated in hard copy and via the Web site of Taylor's magazine American Renaissance.
This dubious report, in slightly modified form, also has appeared in the Journal of Social, Political and Economic Studies, a periodical that is home to the writings of many "intellectuals" whose views about race are similar to Taylor's.
In this latter publication, Taylor's co-author is Glayde Whitney, a Florida professor who recently wrote a positively glowing introduction to Duke's racist and anti-Semitic autobiography.
Based on a cursory examination of 1994 data about interracial crimes between whites and blacks — less than a sixth of all crimes committed that year — Taylor comes to a series of what he describes as "startling conclusions" about black criminality. Blacks, he claims, are vastly more likely to attack whites than vice versa and, in fact, are far more prone to criminality in general. The reason, Taylor suggests amidst a blizzard of misleading statistics, is their blackness — something about black genes, or perhaps black culture, that endows African-Americans with a natural proclivity to criminality.
Mistaking Poverty for Race
But Taylor is plainly, demonstrably wrong.
Taylor uses an incredibly simplistic analytical method that flatly ignores the fundamental conclusion of decades of serious criminology: Crime is intimately related to poverty. In fact, when multivariate statistical methods such as regression analysis are used, study after study has shown that race has little, if any, predictive power.
This basic fact is so well understood among scholars of criminal justice that the preface to Minnesota's official crime data reports carries this caveat: "Racial and ethnic data must be treated with caution. ... Existing research on crime has generally shown that racial or ethnic identity is not predictive of criminal behavior with data which has been controlled for social and economic factors."
When more sophisticated methodology is employed, socioeconomic factors including poverty, education, social status and urban residence account far better for criminal behavior than race. Above all, income counts.
It is precisely because being black in America is closely correlated with being poor, suffering from high unemployment and having low levels of education that the black community has relatively high crime rates.
In 1994, the same year that Taylor's data comes from, the poverty rate among blacks was three times that of whites. In addition, nearly 40 percent of black children grew up in poverty.
So while it is true, for instance, that blacks rob whites far more than vice versa, that is hardly a surprise — whites, after all, own nearly 10 times the wealth that blacks do on average. They also own far more businesses. Thus, it is only natural that any rational robber would select whites over blacks as victims.
It would truly have been a "startling conclusion" if the facts had shown that whites attacked blacks more than the other way around. That poor people are more prone to criminality at the expense of the wealthy is utterly unsurprising.
Selection Bias and Reality
Taylor's decision to simply ignore these well-documented criminological findings is not his report's only flaw. Another major error — a cardinal sin in the science of statistics — is "selection bias." Although Taylor wants crime patterns to be explained by the mere presence of black people, only a contorted analysis based on a small subset of crime data is able to produce such "evidence."
Selection bias may best be understood by example. Say a researcher wants to identify the leading causes of death, then decides to answer that question by examining data on drowning deaths — even though drowning accounts for only a small percentage of all deaths.
By looking only at drowning, the researcher might conclude that children are at very high risk of this type of death and that swimming in pools is the highest-risk activity one could engage in.
Ignored in this analysis would be the impact of disease and traffic accidents, for instance, which rank far higher as causes of death in the general population than drowning. Similarly, by concentrating only on interracial crime, Jared Taylor paints a severely distorted picture of crime and victimization patterns in the United States today.
What Taylor actually does is consider only a subset of data on crime — statistics on interracial crimes between blacks and whites from the National Crime Victimization Survey (NCVS).
For crimes of violence — the crimes Taylor focuses on — that data covers just 16 percent of the crimes committed in 1994. The result is a skewed view of the impact of race on crime that suggests that whites ought to be terrified of blacks who, in Taylor's view, present a serious threat to society.
Missing the Forest for the Trees
But this analysis completely overlooks the larger — and far more scientifically defensible — pattern in the data: Most crime is intra-racial (black-on-black and white-on-white), not interracial.
In fact, the NCVS data show that 73 percent of white violent crime victims were attacked by whites, and 80 percent of black victims were targeted by blacks. This pattern is even clearer in the category of murder.
According to a 1997 government report, 94 percent of black murder victims, as well as 85 percent of white murder victims, were slain by members of their own race. Thus, the larger reality, that danger comes mainly from one's own race, is utterly ignored by Taylor, who for reasons of his own is interested only in interracial crime.
So while it is true that blacks victimize whites in interracial violent crime more than vice versa, that is, for the reasons described above, no surprise.
As the authors of a recent book, The Color of Justice: Race, Ethnicity and Crime in America, point out: "Some researchers have challenged the assertion that crime is predominantly intra-racial, pointing to the fact that a white person has a greater likelihood of being victimized by an African American offender than an African American has of being victimized by a white offender.
Although this is true, it does not logically challenge the assertion that crime is predominantly an intra-racial event. Remember the NCVS reveals that the typical offender is white."
Throughout the report, Taylor makes similar errors in analysis. Another instructive example is his treatment of interracial hate crime.
Misinterpreting Hate Crimes
Taylor looks to statistics on hate crime to make the point that blacks are far more likely to attack whites for racial reasons than the other way around — even though, as Taylor himself acknowledges, hate crime statistics are widely known to be seriously flawed because of reporting errors.
Through a combination of strange methodology and mathematical miscalculation, he ends up overstating his case.
First, Taylor excludes hate crimes based on religion, sexual orientation and disability. Then, using the remaining motivation categories of race and ethnicity, he says that 63 percent of these crimes were committed by whites, less than their 72 percent proportion of the population would suggest; and 19 percent were the work of blacks, even though blacks account for only 12 percent of Americans.
Thus, Taylor concludes, blacks are more likely to commit hate crimes while whites are less so.
These numbers are deceptive. If one looks at all hate crimes and all ethnic groups, the data show that whites are responsible for 75 percent of all hate crimes — higher than their proportion in the population — while the black rate remains at 19 percent.
Corrected for population, these numbers mean that blacks are 1.37 times more likely to commit hate crimes than all other races combined — a far cry from the 1.99 rate that Taylor advances. Whites, too, commit more hate crimes than all other races combined, but only slightly more so.
This type of exaggeration is typical of the poor methodology Taylor employs in his analysis.
Taylor also asserts that "millions of ordinary interracial crimes" should really be considered hate crimes — an insupportable conclusion. Hate crimes are not simply crimes committed between persons of different races.
They are crimes that are motivated by the race, ethnicity, religion, sexual orientation or other group characteristic of the victim. A black man's robbery of a store owned by a white businessman is not a hate crime unless the offender, motivated by animus toward whites, chose the store simply because the owner was white.
Racial Profiling Defended
Ultimately, Taylor's article concludes blacks are so much more likely to commit crimes than whites — "blacks are as much more violent than whites as men are more violent than women" — that police officers are justified in stopping them more often as they drive.
It's just good police work, Taylor argues, to check up on members of such a crime-prone race. Of course, using this logic, police should routinely stop men under the age of 35; those whose license plates indicate they come from high-crime states; and anyone in a downtown area.
Quite apart from the glaring Constitutional violations involved in such a practice, Taylor is simply wrong about the usefulness of racial profiling. A study of profiling released by the nonpartisan U.S. General Accounting Office earlier this year shows that stopping individuals based on race, gender or state of origin does not increase the likelihood of discovering contraband or illegal activities.
That is because most members of minority communities — just like most young people, most whites, most males and most people overall — are law-abiding citizens.
This fact has been accepted by major law enforcement agencies around the country. The National Association of Police Organizations, for instance, opposes the practice. So does the International Association of Police Chiefs.
Failing the Test
In the end, it is simply poor analysis to claim that one factor is responsible for a phenomenon as complex as crime. Researchers have been exploring the causes of crime for decades, and there is wide agreement that there is no answer as monolithic as that which Taylor offers: race, and specifically the black race.
Although it is essential to respond to the arguments advanced in The Color of Crime, it is equally important to remember the politics of the man who produced it. Taylor is the editor of American Renaissance, a magazine that focuses on dubious theories about the relationship between race, IQ, sexuality and crime.
He recently appeared at the National Press Club with three other "researchers" to hawk the idea that crime is highest among blacks worldwide, and that that is correlated to their allegedly smaller brains. (These ideas are roundly dismissed by mainstream researchers.)
One of Taylor's co-presenters was J. Phillippe Rushton, a college professor who has posited an inverse relationship between brain and penis size.
Clearly, Jared Taylor's goal is to offer an "academic" justification for the pre-existing belief that African-Americans are a menace to society and a serious threat to white America.
His booklet is simply the latest effort of racial ideologues to demonstrate black America's hatred for whites and to encourage whites to "take back the country" as a matter of survival. Although there are many inaccuracies in Taylor's analysis, those already detailed should give a sense of the quality of his research.
As social science, The Color of Crime fails the test.