“Personal anecdotes don’t prove anything. The justice system isn’t racist. Black people are arrested more often because they commit more crime. Period. End of story.”
So read the message in my inbox this morning, sent by someone who had happened across my essay about Ferguson, the grand jury decision in the Darren Wilson case, and the history of police misconduct in black communities. To the writer of said missive, that history didn’t matter; even if true, in his mind it was no longer relevant—a trifle of an earlier less enlightened era when compared to the present. Yes, racism may once have plagued the nation’s legal apparatus and those charged with enforcing its rules, but today if black youth die at the hands of police or find themselves in jail or stopped and searched on the streets, it is either the result of their own wrongdoing, or the wrongdoing of others like them. To the extent crime rates are higher in black communities than white ones, according to this logic, any and all black people (especially males) will simply have to accept the possibility that regardless of their own criminality or lack thereof, they may be subjected to suspicion, profiling, search, harassment, even violence at the hands of the cops. Life as a walking contagion is simply the price that must be paid for wearing the epidermal uniform of the team with the higher rate of offending. Innocent until proven guilty, on this rendering, is but a theoretical contrivance with no applicability to those who are dark; a legal standard meant for show, and for no more serious reason than that. Because of the odds represented by the y-axis on some social science graph, the black people represented on the x-axis can be stripped of their humanity and reduced to walking actuarial tables.
This is America. Welcome to it.
I suppose that at some abstract level, such statistical discrimination—simply comparing crime rates to arrest rates and deciding there is no real unfairness operating—may make mathematical sense. And yet, there are still two significant problems with such an analysis, both of which demonstrate quite splendidly why we do not, as a general rule, leave matters of social morality to mathematicians. First, this kind of argument ignores the inherent injustice of using aggregate data to justify mistreatment of individuals who are not represented by that data. In short, even if black crime rates are disproportionate in certain categories, there will be millions of individual African Americans (indeed a majority of them) who are not criminals. To consign millions of innocent people to potential mistreatment because of the actions of a clear minority of one’s group—however statistically disproportionate that minority may still be—is intrinsically unfair. And second, such an analysis is highly selective. The data one chooses to examine (and to ignore) when making such a case as this is instructive.
So on the one hand it is true that the arrest rates for the most serious offenses—murder, rape, aggravated assault and robbery—do closely mirror the offending rates as reported by victims to the Bureau of Justice Statistics each year in the annual crime victimization surveys. There is no disputing that rates of offending in these categories are racially imbalanced, with African Americans committing these crimes at higher rates than whites. Don’t misunderstand: this fact still doesn’t justify generalized suspicion of black folks; after all, the percentage of blacks committing these crimes each year is still a small minority of all black folks in a given community. But so far as this argument goes, the data doesn’t lie: rates of arrest and incarceration for crimes such as these are racially-imbalanced less for reasons of racial bias and more because African Americans commit these crimes at higher rates; not because they are black of course, but because these crimes are highly correlated with various socioeconomic conditions disproportionately experienced by black folks in America.
And yet, the higher black offending rate for these crimes hardly acquits the justice system of the charge of racial bias. First, because, as already noted, for individuals who are innocent of wrongdoing to be suspected of criminality because of the actions of entirely different black people is to experience racialized unfairness, no matter what the larger data says. And second, because unequal treatment can still enter into the system for other, lesser offenses. And indeed, it is there—especially regarding the so-called war on drugs—where disparities have been the most apparent.
As I noted in a second essay following the grand jury decision, there are millions of us—myself very much included—who have broken an assortment of laws, and in particular those related to drugs, without worrying about arrest, prosecution or incarceration. Indeed, for most whites our drug use is entirely risk-free. We are not the ones suspected of possessing or using narcotics (even though whites use drugs at rates that are equal to or even higher than for people of color), and so we can avoid detection and punishment, irrespective of our behaviors. This is among the largest categories of white privilege—a concept that most whites still reject, no matter the clear and convincing evidence of its existence.
But since the e-mail this morning insisted that “anecdotes don’t prove anything,” perhaps it would do us well to actually examine the data. And when it comes to the drug war, it is incredibly easy to ascertain the extent of racial injustice and white privilege. Well beyond mere anecdote, the numbers prove that racism in the justice system is so severe that it literally shapes everything about life in America: from who is and is not in jail or prison, to who is or is not employable, to whose families are financially stable, to the way media discusses matters of crime and deviance.
In order to determine the extent to which blacks, for instance, are over-arrested for drugs and whites under-arrested for the same, there are several things one needs to know. They are:
1) How many whites and how many blacks are arrested each year for drugs (and specifically for drug possession, since it is possession and use rather than drug dealing for which most drug arrestees of all races are busted)?
2) How many whites and how many blacks are current drug users, thus in violation of drug laws and potentially capable of being arrested for a drug offense?; and,
3) How much of a difference is there between those relative numbers, and thus, how many more whites and how many fewer blacks would be arrested each year if arrest rates comported with rates of drug use and/or possession violations?
The first number is easy enough to come by: According to the FBI, there were approximately 1.2 million arrests for drug violations in 2013. Roughly 815,000 of those arrested were white, while 366,000 were black. But these numbers are a bit deceptive. As it turns out—and this is common for most government data, unless noted otherwise—racial categories include persons deemed ethnically Hispanic, since Hispanics are not considered a race unto themselves. Approximately 88 percent of all Hispanics are classified racially as white in the data (thus they are included in white data tables), while a little less than five percent are classified as racially black, about three percent are classified as American Indian, a little more than one percent as Asian or Pacific Islander, and three percent as mixed race.* So to determine the number of “real whites” (those who are not socially seen as people of color) who were arrested for drugs in 2013, one must back out those among the 815,000 in the overall white totals who are Hispanic.
That too is a relatively quick and easy calculation. The FBI (in the table referenced above) provides data on the number of Hispanics arrested for drug violations, and in 2013 that number was about 119,000. However, this is an artificially low number, because the ethnicity data is only provided for those arrestees whose ethnicity is known, and as it turns out, that information was only available for 52 percent of all persons arrested for drug violations. Of those, Latinos represented nineteen percent of persons arrested. If we presume (and this seems reasonable) that Hispanics also represented nineteen percent of those whose ethnicity was not recorded, rather than Latinos representing 119,000 drug arrests in 2013, the actual number would have been more like 228,000. If 88 percent of these were “Hispanic whites,” this means that in 2013 there were 200,000 Hispanics arrested for drugs who appear in the data for white arrests, and would need to be backed out of the white arrest totals in order to ascertain the number of “real whites” arrested for drugs that year. Once this adjustment is made, rather than whites representing 815,000 drug arrests, the actual number is more like 615,000.
So reviewing now, this means that in 2013, when it comes to arrest for drugs there were 615,000 whites and 366,000 blacks arrested for drug violations.
Of these, the FBI notes that 82.3 percent were arrested for possession only as opposed to dealing. Assuming this percentage applies roughly equally to each racial group (and it very likely does), this would mean that there were roughly 500,000 whites and 300,000 blacks arrested for drug possession in 2013. If, as noted previously, there were 1.2 million drug arrests in all, and 82.3 percent were for simple possession, this would translate to about 990,000 possession arrests in all that year, with whites representing 51 percent of possession arrests and blacks constituting about thirty percent of possession arrests.
To determine whether these numbers suggest racial disparity and likely bias in law enforcement—and if so, how much bias—we now need to look at the data on drug usage. Here, we need information from two sources: Census data that will give us the total number of whites and blacks who are age twelve and older (thus, eligible for consideration in drug use and arrest data), and data on relative percentages of drug use by whites and blacks—data which is provided by the Substance Abuse and Mental Health Services Administration. Once we have the first numbers we can multiply them by the percentages in the second set of numbers, to come up with overall numbers of white and black drug law violators.
According to Census data (from the interactive tables found in the American FactFinder section of the Census Bureau website), there were 173.1 million non-Hispanic whites and 34.1 million blacks age twelve and older in 2013. According to the SAMHSA data on drug use, there were 24.6 million drug users in the U.S. in 2013. Data on racial usage rates indicates that 9.5 percent of whites and a very similar 10.5 percent of blacks used drugs that year. Multiplying these usage rates by the population numbers produces a total of 16.5 million white users and 3.4 million black drug users in all. Whites are therefore two-thirds of all drug users in the country while blacks are fourteen percent of users. Latinos, it should be noted, though constituting about eighteen percent of the population, comprise only fifteen percent of drug users, and use drugs at a rate lower than whites—about 8.8 percent.
But remember, whites are only 51 percent of persons arrested for drug possession, while blacks are thirty percent of persons arrested for drugs. If arrest rates mirrored the rates at which persons in different racial groups actually violated drug laws, rather than 500,000 whites arrested for drug possession in 2013, there would have been more than 660,000 such arrests—in other words, more than 160,000 additional white people arrested for drugs that year. As for blacks, if arrest rates mirrored violation rates there would have been about 140,000 African Americans arrested for drug possession rather than 300,000—in other words, 160,000 fewer blacks arrested for drugs in 2013.
Imagine how different life would be in America if 160,000 more whites and 160,000 fewer blacks were being arrested each year for drugs: a shift of 320,000 persons in all in terms of who would be brought under the control of the justice system and who wouldn’t. Imagine what it would mean over the course of a decade: at least 1.5 million more whites with drug arrest records than the numbers of whites who actually have such records now; and likewise, 1.5 million fewer blacks with drug arrest records when compared to those who have such records now or will in the next ten years. First, had this happened, it would likely mean that the war on drugs would have ended long ago. It is hard to imagine white Americans and lawmakers sitting back and allowing 1.5 million more whites to be arrested each decade for drugs. So it is no exaggeration to say that but for white privilege in the enforcement of drug laws, the war on drugs and all of its destructive impacts on neighborhoods, families and the country as a whole would have been avoided.
Then imagine what a more equitable and representative enforcement of drug laws would mean (or would have meant over the years) for our jail and prison populations; what it would mean for the employability of whites compared to blacks (since people with arrest records have a harder time finding jobs); for the formation of stable families (since arrests can disrupt normal family life and the ability to provide financially for one’s children); and for the right to vote (since incarcerated persons generally are disallowed from voting and in some states, even after being freed from prison, one loses the right to vote). Imagine what such a shift in arrests would mean for the public perception of criminality and deviance—how it might effect policing, media portrayals of crime, and the stereotypes that are held by millions of people currently subjected to the lopsided arrest and prosecution of African Americans.
In short, imagine what the nation would look like in the absence of the obvious, blatant and massive racial disparity in just this one area of the justice system.
And then please, by all means, tell me the one again about how the system is fair, and racial bias a myth.
* These percentage break downs were determined by looking at Census Bureau population tables under the American FactFinder section of census.gov for July 2013. The FactFinder tables provide data for Hispanics as members of each racial group. By comparing those numbers to the overall number of Hispanics in the population, one can quickly determine the share of the Hispanic population that is white, black, etc.