Why Stop-and-Frisk Does Not Work


The Inconsistency of Our Outrage

The stop-and-frisk debate is instructive in that it spotlights the inconsistency in outrage that both conservatives and liberals display over the violation of constitutional rights. Many conservatives appear completely insensitive to the empirical reality that stop-and-frisk policies have led to fourth amendment violations against minorities time and time again. They argue that the benefits derived from such policies, in terms of crime and homicide reduction, is valuable enough to outweigh fourth amendment concerns.

In the same breath, however, many of these same conservatives appear scandalized by even the most basic and reasonable gun regulation, arguing that any second amendment violation goes too far, irrespective of the potential public health benefits. Curiously, they appear quite willing to empower the police state with stop-and-frisk policies, insisting that “training” and “discipline” is more than sufficient to inoculate the entire police force against bad judgment and racism, and then, with no apparent irony, turn around and argue that they need assault rifles at home because the police can’t be trusted. The second amendment has become sacred, the fourth amendment optional.

Liberals often display a similar inconsistency. They are offended that conservatives can so callously dismiss the constitutional protections of minorities, ignoring data that might point to the empirical benefits of targeted policing, but are quite willing to dismiss second amendment violations if a public health benefit can be shown.

In both cases, the constitution functions as an ideological Rorschach test, you see what you want, and feign outrage when others don’t see the same thing as you. If we truly care about human well being, we need to stop shouting “Constitution!” whenever it’s convenient, and root our discussion in actual reasons other than “because the Founding Fathers said so.”

Stop-and-Frisk Doesn’t Work

Asking whether stop-and-frisk is effective is a little bit like asking whether or not kicking a door in is an effective way to get into a house. You may get through the door, but you could have just knocked. Similarly, framing the debate as a yes/no question of whether or not “stop-and-frisk is effective” obscures meaningful cost-benefit analysis by overemphasizing any potential benefit, while ignoring guaranteed costs experienced by minorities. Maybe there are ways to get crime down without kicking the doors in — this is the question we must address.

We should start by clarifying that nobody really knows how effective stop-and-frisk policies are. It is essentially impossible to parse out the specific contributory effect of stop-and-frisk on crime rates, relative to other police procedures going on at the same time such as computational analysis of crime patterns, “hot spot” policing, and community engagement with citizens. On top of that, controlling for the large number of exogenous variables that have contributed to a significant decline in crimein other cities without the use of stop-and-frisk further makes an effective evaluation of stop-and-frisk essentially impossible.

As Jeffrey Fagan, Professor of Law at Columbia Law School, stated , “Anyone who says we know [what] is bringing the crime rate down is really making it up.” A paperby Hebrew University’s David Weisburd and George Mason’s Cody Telep and Brian Lawton takes a similar position, “Unfortunately, we can only speculate on issues of police innovation and the crime decline in New York. New York City did not put an emphasis on evaluation of its policing strategies… the ‘bad news’ is that we will never know whether what happened in New York was a ‘policing miracle.'”

Consider that, in the first quarter of 2013, there was a 50 percent drop in the number of stops in New York City, while remarkable declines in violent crime rates (almost 25 percent) were observed compared to the year before. Furthermore, two years prior to the implementation of stop-and-frisk there was already a 10 percent decline in violent crime. Though this doesn’t say anything about the causal effect of stop-and-frisk, it does emphasize the reality of how many different variables are at play, and how difficult it is to extricate the effect of any one of these on the crime rates.

That being said, the most sophisticated analysis on NYPD’s stop-and-frisk procedures finds that there is “no evidence that misdemeanor arrests reduced levels of homicide, robbery, or aggravated assaults.” An unpublished study by Dennis Smith and Robert Purtell, found “statistically significant and negative effects of the lagged stop rates on rates of robbery, burglary, motor vehicle theft, and homicide and no significant effects on rates of assault, rape, or grand larceny.” Further research by Professors Richard Rosenfeld and Robert Fornango found that there are “few significant effects of several SQF [stop, question,and frisk] measures on precinct robbery and burglary rates.”

A new study investigated the question of what happens after an arrest resulting from stop-and-frisk occurs. There is a tendency to view arrest as the goal,but new research challenges this idea, pointing out that half of arrests from stop-and-frisk do not lead to a conviction. For those that do, very few are related to violent crime or yield prison sentences longer than thirty days. So not only are we potentially harassing innocent bystanders, when we do make an arrest, it turns out that half of them really are innocent, and the other half aren’t the main drivers of crime.

So, it should be clear, at best there is a mild to insignificant effect of stop-and-frisk procedures on crime rates, and even this evaluation is dubious given the impossible task of identifying the specific effect of stop-and-frisk beyond other practices and general trends.

The Cost of Stop-And-Frisk

There is, however, considerable cost to stop-and-frisk procedures. A study conducted by researchers at Columbia University evaluating NYPD’S stop-and-frisk program found that, even after controlling for precinct variability and race-specific estimates of crime participation, persons of African and Hispanic descent were stopped many times more frequently than whites. Even though Blacks and Hispanics make up only 26 and 24 percent respectively of the New York City Population, they represent 51 and 33 percent of all stops. Even after controlling for race-specific crime estimates, Blacks were stopped 23 percent and Hispanics 39 percent more than whites.

The study also found evidence that stops of Blacks and Hispanics were less likely to lead to arrests than stops of White individuals, which suggests that the “reasonable suspicion” standard was relaxed when investigating Blacks and Hispanics. Consider that, between 2004 and 2012, in only 1.5 percent of stops in New York City and 1.1 percent of stops in Philadelphia was a weapon actually found following a frisk. Given that the Supreme Court requires that a suspect be “armed and dangerous” prior to a frisk occurring, one wonders why police are wrong almost all the time. It is clear, then, that whatever arbitrary standard is being used by police departments to determine reasonable suspicion is entirely too permissive in that it is not predictive in any way of actual criminal activity. That reasonable suspicion is informed by social cues and stereotyping, rather than actual suspicious activity, should be more than enough to cast doubt on the efficacy of such policies.

Ironically, police defend the low arrest rate of stop-and-frisk as proof of the deterrence value of such policies. Ostensibly, criminals are now weary about carrying contraband and weapons on their persons for fear of being stopped by the police. One wonders, then, how police establish “reasonable suspicion” so often with Blacks and Hispanics who are apparently no longer acting suspiciously.

The perceived fairness of police procedures matters greatly for any sober cost-benefit analysis of stop-and-frisk policies. Several findings have shown that citizens who believe that police act legitimately are less likely to act disrespectfully, engage in resistance-behaviors, or have a violent reaction to arrest. A recent study found that police practices such as New York’s stop-and-frisk policy inadvertently contribute to higher rates of delinquency by decreasing the legitimacy of police in the eyes of those targeted. For racial and ethnic minorities who experience disproportionate police contact, there is greater reported commitment to deviant peers, less anticipated guilt in future crimes, and higher delinquency frequency.

Another study cautions against a myopic focus on the benefits (instead of the costs) of stop-and-frisk policies: “Given the possible negative impacts of SQF policing, both on citizens who live in such areas…we suspect especially in the long run that this approach will lead to unintended negative consequences…any gains from aggressive police efforts in the short term could lead to long-term increases in offending.”

Despite this fact, Paul Larkin in a recent Atlantic piece writes that “reasonable suspicion does not require an officer to be right, or even to be more likely right than wrong, so we should not be surprised if the police often err.” However, reasonable suspicion does require, I don’t know, reasonable suspicion, prior to a stop. The Supreme Court ruled that a description of race and gender alone “will rarely provide reasonable suspicion justifying a police search and seizure.”

So, under what definition is it “reasonable” to stop someone when doing so almost never leads to a seizure or arrest? Would it be reasonable to stop someone carrying a 7-11 Big Gulp if doing so leads to an arrest one percent of the time? The point is this: Skin color is not a crime, yet if Larkin and other conservatives are to be sanguine about making it one, we should be unreasonably confident that doing so is making everyone safer. There is simply no evidence to suggest this is the case.

The Alternatives to Stop-And-Frisk

Are there ways to get into a house without kicking down the door? As it turns out, yes. We should start with the door knob.

After being accused by the Department of Justice (DOJ) of engaging in numerous false arrests, excessive force, and unreasonable search and seizures, the Los Angeles Police Department (LAPD) agreed to enter into a consent decree that focused on nine major areas of improvement including supervisory measures to promote Civil Rights Integrity. Since the implementation of this system, recorded crime is down in every police division in the city. 83 percent of residents now say that the LAPD is doing a good or excellent job, and the frequency of serious force has fallen each year since 2004. There is also substantial evidence that the quality and quantity of enforcement activity has improved, evidenced in the doubling of pedestrian and motor vehicle stops since 2002, and a rise in arrests occurring over the same period. Furthermore, more than two-thirds of Black and Hispanic residents also report that the LAPD is doing a good or excellent job.

Another popular approach is called “focused deterrence,” which is more-or-less the opposite of stop-and-frisk. This approach operates from the standpoint that there is a small proportion of young men responsible for the majority of violent crime, and it seeks to target these individuals for criminal prosecution. This approach was responsible for Boston’s decline in its homicide rate in the 1990s, and was successfully adapted to challenge violent crime in High Point, North Carolina. This policy has the added benefit of bolstering community-police relations because it avoids harassing innocent bystanders, and places emphasis on actual criminals.

An advocate of stop-and-frisk, Paul Larkin writes that, “If the stop-question-and-frisk technique is effective, then the surest way to discriminate against African Americans is to abandon the use of that approach in drug-infested black-majority neighborhoods. Why? Because 99-plus percent of the residents of those neighborhoods are not involved in drug trafficking. Denying them the benefit of a worthwhile law-enforcement technique would only worsen their plight.”

There could not be a more clear demonstration of Larkin’s preoccupation with the benefits of stop-and-frisk, with zero consideration of the cost. We can see from the recent mayoral election and widespread criticism of stop-and-frisk policies that many African Americans believe that these procedures are racist and humiliating. Instead of pretending that humiliation is the cost of security, we owe it to vulnerable populations among high-crime communities to find law enforcement practices that preserve dignity and security. Thankfully, these alternatives exist. There doesn’t have to be a trade-off between effectiveness and justice, nor are citizens asking for there to be one. We should have both, and we can have both.


Pre-Empts to Comments

—“Skin color is not a crime, but crime is undeniably correlated with skin color. They perform stop-and-frisks in areas of high crime rates, i.e. the slums where minorities are the majority.”

Yet,  as the article explicitly states: “Even after controlling for race-specific crime estimates, Blacks were stopped 23 percent and Hispanics 39 percent more than whites.” So even after controlling for different crime rates among races, there is still a significant effect of discriminatory profiling

–“I dunno, robbery, burglary, motor theft and homicide are worth something. ‘Fails’ by some criterion, but not ‘fails to work’ as claimed.”

You have picked the one non-published study out of the batch, which means it hasn’t been through peer review. The two studies on either side, both published, find that there was no/very small effect. Unless the article is holding back on a whole host of research, it is abundantly clear by any objective measure that the costs far outweigh the benefits. And that’s not even including the opportunity cost of pursuing better policing strategies (the article lists several). Its not even close. By any rational measure of “work”, this policy doesn’t.

–“How is it possible to claim that the policy doesn’t work?”

We have a very small but hard to measure potentially positive effect paired with a larger and not as difficult to measure negative effect combined with the opportunity cost of not using superior and less costly policing measures. How is it possible to claim that holistically the policy does work? Do certain parts work? Yes, just as nuking your own house (a tad more extreme than the door kicking example) to stop a burglary-in-process “works” to some degree (i.e. it stops the burglar). However, no sane person would posit that, holistically, nuking your house to stop a burglary works. To determine whether something works or not you have to consider the costs and benefits (which the article does), not just the benefits.

–“How can we know that there will be backlash to Stop-and-Frisk?”

The question — “Are we really relying on surveys that ask ‘what are your feelings about stop and frisk?'” Would be a semi-valid point, except in literally the next sentence after the “survey of feelings” in the article: “A recent study found that police practices such as New York’s stop-and-frisk policy inadvertently contribute to higher rates of delinquency by decreasing the legitimacy of police in the eyes of those targeted.” So yes, “feelings” do matter when they result in even more criminal activity. And to even further refute this point, the paragraph directly after states: “Another study cautions against a myopic focus on the benefits (instead of the costs) of stop-and-frisk policies: “Given the possible negative impacts of SQF policing, both on citizens who live in such areas…we suspect especially in the long run that this approach will lead to unintended negative consequences…any gains from aggressive police efforts in the short term could lead to long-term increases in offending.”

Related posts