Crime, Race and the Perils of Profiling
September 7, 2011
When it comes to rationalizing the mistreatment of people of color, there are few who manage to do it better, or more consistently, than syndicated columnist Mona Charen.
So, for instance, when officers from the NYPD’s Street Crimes Unit were acquitted of any wrongdoing after killing Amadou Diallo — whose wallet they mistook for a gun, leading them to fire 41 rounds his way, 19 of which found their mark — there was Charen ready to leap to their defense. Although she admitted the event had been “sad” and even “tragic” (words she would have no doubt found putridly inadequate had the victim been a nice Jewish boy on his way home from Yeshiva, or a WASP hedge fund manager on his way home from a hard day at the office), to hear Charen tell it, it was almost unavoidable in the case of Diallo, the African immigrant. After all, Diallo was a black man in a dangerous neighborhood, and given generally higher black crime rates, police are understandably afraid of black people. So yes, Charen agreed, Diallo had likely died at least in part because “he looked like so many of the young men in that neighborhood who are seriously dangerous,” but that was just the price he would have to pay for the “high crime rate of American blacks.”
Then there was her Thanksgiving day essay in 2008, in which she brushed aside any lingering regrets for the genocide of indigenous peoples in the Americas, noting that it was merely “the usual course in human affairs,” and simply a matter of “the more technologically advanced civilization” winning. No big deal.
And when Harvard professor Henry Louis Gates Jr. was arrested in his own home in July 2009, for daring to become belligerent with the Cambridge cops who presumed he didn’t belong there — belligerence that, it should be noted, does not constitute disorderly conduct according to Massachusetts law — Charen was at it again. This time, she lambasted President Obama for his rather tame (and factual) suggestion that there has been a “long history” of blacks and Latinos being stopped disproportionately by police, calling it a “left-wing fable.” After all, she noted, “Blacks and Hispanics also commit a disproportionately high percentage of crimes,” so even if the president were correct, we ought not worry about it much. The mistreatment is justified. Oh, and according to Charen, the president’s statement itself amounted to “reverse racism” — a charge so utterly bizarre as to defy explanation, seeing as how it suggests that merely mentioning the history of racism makes one an anti-white bigot.
Now, true to form, dear Ms. Charen has returned to the well of white denial and rationalization, with her most recent column (September 5, 2011), in which she attacks the New York Times for its supposedly biased coverage of a pending lawsuit against the NYPD. The lawsuit, filed by the Center for Constitutional Rights, alleges a pattern of racial profiling in citywide stop-and-frisk policies, and recently a judge ruled that the suit had sufficient merit to go forward, despite attempts by the city to have it dismissed. Apparently by reporting that fact — even though the Times included statements from police officials responding to the allegations of profiling — the paper is somehow guilty of liberal bias. After all, according to Charen, given the much higher rates of criminal violence among blacks and Latinos in New York City, it only makes sense that police would stop and search a disproportionate number and percentage of them, relative to whites. By failing to provide the data on relative racial crime rates in the city, the Times, in Charen’s view, is deliberately misleading their readers, and going on a “racial profiling goose chase.” Charen then proceeds to provide the data herself — or rather she produces data provided to her by conservative policy analyst and writer, Heather MacDonald — and suggests that since blacks and Latinos commit such a massively disproportionate share of violent crime, you would have to be a fool to show the least bit of concern about the racial disparity in stops and frisks.
Although I have previously written several essays about the fundamental illogic of the right-wing position on racial profiling — and although law professor David A. Harris has thoroughly eviscerated every attempt to justify the practice in his meticulously researched book on the subject — the regularity with which such pedantic rationalizations for racial bias bubble up requires that I do it again, if for no other reason than to demonstrate the utter anti-logic of conservative thinking on this subject, and their basic inability to interpret data. And, to be perfectly frank, because making Mona Charen look foolish, though painfully easy to do, is fun.
The Facts About Stop-and-Frisk and Crime in New York City
First, let’s begin with the facts we know, and which are not in dispute.
Beginning in the 1990s, under the Mayoral Administration of Rudy Giuliani, the NYPD adopted a number of new policing strategies intended to address the problem of crime. Among these were the use of computer-modeling to track crime and then predict where future crime would likely occur, the deployment of additional officers to those “hot spots” identified by the COMPSTAT system, and a “broken windows” philosophy of policing, which holds that if minor infractions and quality-of-life rule violations are enforced aggressively, the result will be a reduction in more serious crimes as well. Although crime certainly dropped in New York City following the adoption of these strategies, it also fell nationwide, including in many large cities that did not deploy such strategies, making it difficult to claim that the new methods per se had been the cause of the crime drops in New York.
Additionally, the NYPD began widespread stop-and-frisk operations at this time, so that whenever officers felt they had reasonable justification to suspect someone of criminal wrongdoing, they would stop and question them. Then, if these interactions led to further suspicion, they would frisk the presumed perp for illegal contraband such as weapons or drugs. In the first few years after the program’s inception, roughly 85,000 or so persons in the city were being searched annually. By 2009, even though the crime rate had dropped substantially in the intervening years, the numbers of stops had exploded to over 580,000. Last year, they topped 600,000. And consistently, since the beginning, about 85 percent of persons stopped by police and searched, have been black or Latino.
On the one hand, conservatives are right to chide those on the left who claim naively that this fact alone, in and of itself, proves racism in the the NYPD, since folks of color comprise only about 65 percent of the city’s population. Obviously, the population percentages of various racial groups are not the only factor that would logically be relevant to a law enforcement practice, since law enforcement is supposed to aim its efforts at lawbreakers, not random people. As such, the right is correct when they note that the relevant demographic information is not the percentage of whites, blacks or Latinos in a given community, but rather, the percentage of crime in a given area being committed by whites, blacks or Latinos. So let’s take a look at that.
As for crime in New York, there is no argument that the rates of violent criminal wrongdoing in the city are much higher among blacks and Latinos than among whites. About that, Charen is correct, and it forms the basis of her defense of the NYPD. According to data from the NYPD for 2010, victims of violent crime who report their victimization to police, overwhelmingly indicate that their attackers were black or Latino. So, for instance, according to victim and witness reports, 91.5 percent of murder suspects in the city are black or Latino, as are 86 percent of rape suspects, 94 percent of robbery suspects, 87.5 percent of aggravated/felonious assault suspects, 85 percent of misdemeanor assault perps, and nearly all shooting suspects. Meanwhile, only about 5 percent or so of violent crime perps, according to that same data, are white (as opposed to about 35 percent of the population of the city).* So perhaps Charen and the conservatives are correct: perhaps the disproportionate rates at which blacks and Latinos are stopped by the NYPD makes sense. Indeed, to the extent such practices might remove dangerous criminals from the streets, it could even be seen as a pro-black and brown folks policy, ultimately helping the very communities hardest hit by crime. Perhaps the NYPD is the new NAACP?
But no. Despite the seeming logic of such a claim, given the raw data, there are a number of problems with the quick jump to this conclusion so readily and happily made by Charen, MacDonald, and other defenders of the NYPD. These flaws have been meticulously demonstrated by scholarly research on the subject, going back over a decade, and most recently by Jeffrey Fagan, a Professor of Law and Public Health at Columbia University, in his expert report, prepared for the court in the current and pending case.
Deceptive Data: How the Right Misuses Crime Statistics to Justify Racist Policing
The claim that the NYPD’s stop-and-frisk policies are a key element of effective crime control seems laughable once you actually look at the numbers produced by the policy. From 2004-2009, out of 2.8 million stops made by the police, fewer than 6 percent resulted in an arrest of any kind — a rate that is actually lower than that which has been produced in other jurisdictions using random checkpoints. When it comes to finding guns, drugs or stolen property — among the principal goals that defenders of the practice cite as justification for it — the stops are even more inadequate, with guns being found in only 0.15 percent of all cases, and drugs or stolen property recovered in only 1.75 percent of all stops. In the heavily-policed Brownsville community of Brooklyn, arrests occur in fewer than 1 percent of all stops made, and out of 50,000 stops in the community just since 2006, only 25 guns have been recovered: that’s a hit rate of about one-twentieth of a percent.
Among the reasons for such pathetic “hit rates,” Fagan notes that a disproportionate share of all stops are justified on police incident forms on the basis of vague and subjective reasons, such as an individual making “furtive movements,” being in a “high crime area,” or for other unspecified reasons. In those cases, which represent roughly half of all stops made, the hit rates are even worse than in the larger sample. In other words, as a crime-control tactic, stop-and-frisk is inefficient at best, downright irrational at worst. Officers are apparently suspecting people of criminal activity on the basis of clues and signals that are proving to be ill-informed. Yet rather than rethink their assumptions, they continue to use the same reasons for their racially-disparate stops year after year.
Interestingly, the racial disparities are even harder to explain when you consider what Fagan and a colleague discovered even as far back as the 1990s; namely, that as bad as the hit rates were overall for stops-and-frisks, they were actually far lower for persons of color. When searched, blacks and Latinos historically have been about a third less likely than their white counterparts to actually be found with illegal contraband or other evidence of criminal activity. Although the disparities in hit rates have been reduced since the 1990s, blacks stopped and searched are still nearly 10 percent less likely than their white counterparts to receive some kind of sanction (either arrest or a court summons) after being stopped by the NYPD. Far from suggesting that the cops are bending over backwards to be kind to African Americans, this fact suggests that still today, the police are quicker to suspect blacks for less legitimate reasons than they are whites, and thus, after searching them, less likely to actually find evidence of actual wrongdoing.
Furthermore, and as Fagan amply demonstrates, in what may well be the most rigorous statistical analysis ever performed on the subject of racial profiling, the correlation between police stops in a given precinct and reports of crimes in those precincts is generally pathetic. For violent crime, there is no significant correlation between reports of crime and the number or racial distribution of stops made, and the racial composition of a precinct alone actually predicts stops three times better than reported crimes. In other words, the fact that people of color commit the lion’s share of violent crime in New York cannot possibly justify the level of racial disproportionality in stops-and-frisks.
Of course, this makes sense when you consider that stops of this nature are a pretty inefficient tool for catching violent criminals. In those kinds of cases, police have more precise information to go on, and utilize more sophisticated methods of investigation than simply stopping people on the streets because of “furtive movements,” in the hopes of, let’s say, turning up last night’s liquor store holdup man. This is likely why only about 15 percent of stops by police since 2004 have been for the purpose of investigating violent crime, according to the NYPD’s own records: yet another reason why the Charen/MacDonald evidence on people of color and violent crime is irrelevant when it comes to understanding the disproportionality of police stops.
Rather than violent crime, large numbers of stops are written up as being related to a search for weapons or drugs (about 800,000 of the 2.8 million incidents), trespassing (another 325,000 or so stops), or for “unknown or unclassified” offenses (another half-million of the stops from 2004-2009). That blacks and Latinos commit violent crime at much higher rates than whites in New York, cannot possibly explain such wildly disparate rates of racial stops by police, as those stops (and indeed 85 percent of all stops) had nothing to do with the person stopped being suspected of a violent offense.
As for trespassing, the correlation between stops and reports of this offense, is only one-tenth as strong as the correlation between stops and the racial composition of the community alone. And as for drugs and weapons, stop rates are significantly but negatively correlated with reported drug or gun possession offenses in a given precinct. In other words, the rationale being offered by Charen and MacDonald for the stops (that people of color are committing more crimes and thus, racial disparity should naturally result in the stop rates) is exactly the opposite of reality when it comes to drug and gun possession offenses.
Although there is no clear data on drug possession or weapons possession rates for New Yorkers as a whole, there is data on Borough-level drug use and weapons possession for high school students in New York. So we can at least examine data on youth to get an idea of whether the disproportionate stopping and frisking of people of color by the NYPD is justifiable, based on the relative rates of offending in those categories. And once you take a look at that data, it becomes clear: drug and weapons possession rates are far too similar for young whites and young people of color to justify such racially disparate rates of stops.
According to 2009 data from the Centers for Disease Control, in Manhattan, Queens, and Staten Island — the three New York City boroughs that consistently have a sufficient number of white public school students so as to allow for statistical comparisons — there are no significant differences between the rates at which white and black youth carry weapons in general, or guns, specifically; likewise, rates of drug use between the two groups either indicate no significant differences, or differences that suggest higher rates of use by whites.
Although the 2009 data from Brooklyn indicates higher rates of weapons possession and drug use for blacks than for whites, in almost every category of comparison, the sample sizes for white youth are simply too low for statistically valid comparisons or estimates. And even if we overlook the sample size problem, and accept the data at face value, the fact would remain that less than 4 percent of black youth in the borough are carrying a gun in a given month, hardly sufficient to warrant widespread assumptions about their possession of a firearm. Interestingly, the data from 2007, which had larger samples of white students — and thus allowed for more accurate side-by-side comparisons — showed no statistically significant differences between whites and blacks when it came to weapons or gun possession, and generally higher rates of drug use by whites; likewise, the data from 2003 (which also included a sufficient size white sample to allow for comparison), shows that while black youth were slightly more likely to carry a weapon than white youth, whites were equally or more likely than black youth to use, and thus possess, drugs.
In other words, the data on New York City’s youth provides very little statistical rationale for racially-disparate assumptions when it comes to such categories as weapons or drug possession: two key categories for which stops and frisks could at least theoretically be justified. Unless there is some reason to believe that the data for older New Yorkers differs dramatically from the data for youth — so for instance, unless whites in their twenties and thirties just suddenly stop doing drugs or carrying weapons, while black and brown adults continue to do so — it is hard to then justify the disparities in stop and frisk rates between whites and blacks at any age group. This is all the more true when you consider the data on drug use for adults nationally (which is likely mirrored in New York), all of which points to roughly equivalent rates of drug use (and thus, possession) between whites and blacks. Finally, given the hit rates for uncovering drugs or stolen property (defined in the Fagan research as “contraband”), it is clear that racial disparity cannot be justified in these areas: from 2004-2009, blacks searched were actually 15 percent less likely than their white counterparts to be found with contraband on them, and Latinos were 23 percent less likely than whites to possess such items when searched.
Overall, even when you control for the various non-racial variables that could explain the disproportionate stopping of blacks and Latinos by the NYPD (such as local crime levels, the demographics of the community, or the level of police saturation due to higher crime, all of which could logically explain some portion of the higher levels of contact and stops, even if there were no racial bias operating), Fagan’s analysis shows that police are still far more likely to stop people of color than would be expected. For some categories of suspected crime, simply being black or Latino will make one more than twice as likely as whites to be stopped and frisked, even after these other factors have been taken into consideration. Indeed, even in mostly white neighborhoods, people of color are disproportionately stopped and searched, especially for suspected weapons possession or “trespassing.”
The Cost of Racist Ignorance: How Profiling Hurts Us All
It should be understood, however, that the problems with the NYPD’s stop-and-frisk policies are not merely academic, legal, or even ethical concerns. There is a practical concern as well; namely, the utilization of racially-disparate stops by the police can actually make effective crime control more difficult and detract from more promising methods for addressing the very real problem of criminal victimization.
One thing about which all persons — whether on the right or left — should be able to agree is this: even though crime rates have dropped substantially over the past 20 years, the prospects of criminal victimization are still terrifying, and understandably so. As such, we all have an interest in addressing the problem of crime, none more so than lower-income persons of color who bear the brunt of such disproportionate rates of victimization. And no doubt, such folks would, in most cases, support increased police presence in their communities. However (and it’s a big however), that police presence cannot operate in an adversarial and combative way, stopping hundreds of thousands of people of color for ill-defined reasons simply because they happen to be in a “high crime area,” as is so often the justification offered for stops by the NYPD. Such practices can only erode support for law enforcement and make cooperation with police in solving crimes less likely.
Currently, for instance, when people are stopped by the NYPD, even if there is no evidence of criminal wrongdoing uncovered (which is the case in roughly 94 of 100 stops), the names of the persons stopped are entered into a database, ostensibly for use in solving future crimes. As such, it is no exaggeration to say that in some New York City neighborhoods, virtually every black male is in the police database as a potential perp, even if they have never been found to have committed any crime. Such practices as this, in effect, presume that although the people stopped today may be innocent, eventually they will be guilty of something, so we’d best get their names on file now. That such a practice would never be tolerated in the suburbs (or in any white community) should be obvious. And that such a practice will almost by definition create enmity between the black and brown public and the police should be apparent as well. How such strained relationships can possibly help reduce crime in the long-run remains a mystery about which Charen and her ilk seem unconcerned.
Meanwhile, although the stops-and-frisks produce very few arrests and even fewer guns, other efforts have paid far greater dividends. For instance, the NYPD operates a program that pays cash rewards for information about persons with illegal weapons, and another that utilizes undercover investigation teams, working with confidential informants to make gun buys from traffickers and other street criminals. Unlike the stop-and-frisk efforts, these have paid off substantially. In one 18-month period from 2002-2003, the cash reward program resulted in the seizure of 455 guns and 757 arrests, from a total of 1234 anonymous tips, while the undercover gun buy initiative, and related investigative strategies utilizing CIs, recovered nearly 500 additional guns in 2002. One operation in a Brooklyn public housing project in early 2003 netted 65 guns and 36 arrests — far more successful than the stop-and-frisk initiative. Research in Chicago and Detroit has also found that focused undercover stings — which are quite a bit different than stop-and-frisk efforts — can reduce the flow of new guns to criminals by nearly 50 percent during the time of the sting operations.**
Sadly, these kinds of programs — the kinds that actually work — are likely to suffer, the more the NYPD relies on widespread stops and frisks. By sowing resentment and mistrust among the city’s lower-income black and brown communities, potential informants may become reluctant to work with police, for fear that they would become targets themselves. In order to produce good CIs, police need positive relationships with the community; but with largely rookie cops being turned loose on many of the city’s highest-crime communities, and treating everyone in them as a potential criminal perp, it’s hard to see how those relationships can remain constructive over time.
Sadly, it is utterly predictable that persons like Mona Charen and Heather MacDonald will continue to rationalize the mistreatment of black and brown folks, evidence notwithstanding. MacDonald has made a name for herself trying miserably and amateurishly to debunk racial profiling, and Charen is, well, Charen: someone who has never once managed to decry an act of racism perpetrated against a person of color without — as in the case of the 1998 dragging death of James Byrd in Texas — turning the discussion around to supposed black hatred of white people. But despite their own denials and prevarications, the rest of us should be clear as to just how dishonest and/or fundamentally ignorant they really are.
* It should be noted here that the actual black and Latino shares of these respective crimes in NYC could be somewhat lower than these estimates suggest, and the white shares could be somewhat higher. This is because the data from the NYPD only represents those crimes where the race of the suspected perp is known or reported to them by victims or witnesses. However, research has long found that blacks and Latinos (especially blacks) are more likely than whites to report crime. Since people of color are mostly victimized by other people of color, and whites are mostly victimized by other whites, this difference in reporting would tend to underestimate the share of all crimes committed by whites and overestimate the percentage committed by blacks and Latinos. However, with that said, it is unlikely that the differences represented by those crimes that are not reported to police, or where the race of the perp is unknown would dramatically alter the bigger picture: people of color do indeed have disproportionate crime rates, relative to whites, due to the economic conditions that have long been found to be highly correlated with crime, and which also tend to be experienced disproportionately by people of color in the U.S.: namely, concentrated poverty, crowded housing, high population density, and disorganized/decaying urban landscapes.
** I want to be clear here: I realize that even these efforts, as effective as they may be, can still prove problematic. If stings or other investigative techniques are carried out by overzealous or corrupt law enforcement officers (as has certainly been documented in more than a few instances), the erosion of Constitutional liberties may end up being the cost of such efforts — certainly an unacceptable trade-off. And I also realize that given the racial and class biases in the justice system, from the point of investigation to prosecution to sentencing, even these otherwise valid techniques can contribute to an overall system that is highly flawed and in need of significant reform. That said, when it comes to the stated goals of conservatives — to get guns and drugs off the street and catch criminal perps — it is still worth noting that the methods they endorse for doing that fail in comparison to less heavy-handed (if yet still sometimes troublesome) approaches.