What is good polic­ing? We have no idea.

We spend $100 bil­lion on pub­lic safety with­out know­ing the ben­e­fits, says Barry Fried­man

The Washington Post Sunday - - OUTLOOK - Twit­ter: @bar­ryfried­man1 Barry Fried­man is a pro­fes­sor at New York Univer­sity School of Law and the di­rec­tor of its polic­ing project. His lat­est book is “Un­war­ranted: Polic­ing With­out Per­mis­sion.”

Watch­ing the de­bate in this coun­try over pub­lic safety, you’d think some peo­ple wish to live se­curely, while oth­ers wel­come Ar­maged­don. Con­ser­va­tive pun­dit Bill O’Reilly re­cently went after “lib­eral politi­cians” in Chicago and San Fran­cisco, not­ing crime in those cities and say­ing, “The sit­u­a­tion is out of con­trol and a disgrace, and that’s what hap­pens when in­com­pe­tent politi­cians de­mand the po­lice stop en­forc­ing laws.”

The truth is, we all want to be safe. The strug­gle isn’t about out­comes, it’s about meth­ods. Should law en­force­ment have ready ac­cess to ev­ery­one’s phone lo­ca­tion-track­ing data? Should po­lice be re­quired to un­dergo deesca­la­tion train­ing be­fore be­ing au­tho­rized to use force?

Th­ese aren’t ques­tions to be re­solved by free-for-alls on ca­ble news chan­nels. They re­quire facts and anal­y­sis. And yet, al­though the United States shells out well over $100 bil­lion each year for pub­lic safety, we have re­mark­ably lit­tle idea whether that money is well spent. It’s pos­si­ble that any given polic­ing tac­tic or tech­nol­ogy — from Tasers to fa­cial-recog­ni­tion sys­tems to body cam­eras — is a fine or poor idea. But we

re­ally don’t have much sense of which tac­tics and tools work, or whether they are worth the cost. We don’t know how much money we may be wast­ing, or whether we are com­pro­mis­ing civil lib­er­ties, or harm­ing peo­ple or prop­erty, with­out good rea­son.

Through­out the rest of gov­ern­ment, we use cost-ben­e­fit anal­y­sis to an­swer th­ese sorts of ques­tions. (Many economists pre­fer to call it ben­e­fit-cost anal­y­sis, or BCA, rightly ask­ing: Why worry about the costs un­til we know if there are any ben­e­fits?) Whether it is en­vi­ron­men­tal reg­u­la­tion, work­place safety, fi­nan­cial rules or the pro­vi­sion of health care, BCA is per­va­sive. But as a 2014 re­port by the Vera In­sti­tute of Jus­tice pointed out, BCA has not been widely taught or used in crim­i­nal jus­tice. That’s a stark un­der­state­ment when it comes to polic­ing.

Take ShotSpot­ter, a tech­nol­ogy that uses sound waves to pin­point where a gun has been fired. The prod­uct is mar­keted as al­low­ing po­lice to know about gun­shots and re­spond quickly, es­pe­cially in neigh­bor­hoods where peo­ple aren’t in­clined to call the cops. ShotSpot­ter leases the tech­nol­ogy to cities at a cost of $65,000 to $95,000 per square mile per year. (The District of Columbia is one of ShotSpot­ter’s ma­jor clients; it in­stalled sen­sors in 2005 with the help of a $2 mil­lion fed­eral grant and spent $3.5 mil­lion main­tain­ing and ex­pand­ing the sys­tem through 2013, ac­cord­ing to a Wash­ing­ton Post in­ves­ti­ga­tion pub­lished that year.) This is se­ri­ous money for cash-strapped cities, so the ques­tion nat­u­rally is: Is it worth it? To an­swer this, we’d want to see good data on whether the tech­nol­ogy is help­ing cops nab shoot­ers, whether there are fewer shots fired when it is in place or whether gun vi­o­lence is down. And we’d then want to know if the tech­nol­ogy is more ef­fec­tive than an al­ter­na­tive, such as hir­ing more of­fi­cers.

Un­for­tu­nately, it is harder than it should be even to get data for BCA around polic­ing. The rea­sons for this are many. An un­nec­es­sary cloak of se­crecy en­velops too much polic­ing. Law en­force­ment’s in­stinct is to give no in­for­ma­tion be­yond what is nec­es­sary, mak­ing demo­cratic en­gage­ment with polic­ing ex­tremely dif­fi­cult. Ad­di­tion­ally, the tech­nolo­gies for po­lice data col­lec­tion are in many ways prim­i­tive, as any­one who has watched a po­lice of­fi­cer fill­ing out in­ci­dent re­ports in du­pli­cate on the back of a cruiser knows. There are about 18,000 de­part­ments, and ag­gre­gated data would be use­ful, but we of­ten don’t have it.

In ShotSpot­ter’s case, though, the prob­lem is dif­fer­ent. Much of polic­ing tech­nol­ogy is pri­va­tized, and pri­vate ven­dors — claim­ing trade se­cret pro­tec­tion — shield data from the pub­lic and re­searchers. The Post in­ves­ti­ga­tion ob­tained data from the D.C. po­lice depart­ment through a pub­lic-records re­quest and found that ShotSpot­ter sen­sors logged 39,000 in­ci­dents of gun­fire in the District over eight years. Yet while D.C. po­lice say that ShotSpot­ter is “a valu­able tool,” help­ing to es­tab­lish crime trends, the depart­ment told The Post that it didn’t track ar­rests made as a re­sult of ShotSpot­ter alerts. We can’t eval­u­ate polic­ing tech­nolo­gies un­less we can con­nect the use of those tech­nolo­gies with re­sults.

In the mod­ern era of de­ter­rent-based polic­ing, assess­ing the ben­e­fits of law en­force­ment has be­come both more im­por­tant and more dif­fi­cult. Rather than sim­ply chas­ing bad guys, polic­ing has put more fo­cus on dis­cour­ag­ing crime in the first place. Take air­port se­cu­rity: Any­one who car­ries a bomb through a check­point will be ar­rested. But the point of the for­tune we spend on air­port se­cu­rity is to pre­vent peo­ple from even con­tem­plat­ing such an at­tack. The same may be said of closed-cir­cuit TV cam­eras that peer down at us all over our cities and the de­ploy­ment of many other tech­nolo­gies. The goal is to de­ter crime en­tirely.

It is one thing to count crimes that did hap­pen; it is quite an­other to count crimes that didn’t. We can, of course, com­pare crime rates be­fore and after a new tech­nol­ogy or tac­tic is put in place, but it is hard to es­tab­lish the cause of any trends. Did crime go down be­cause of that new tac­tic, or did it de­cline for other rea­sons? For low-in­ci­dent crimes such as ter­ror­ism, we can’t mea­sure a mean­ing­ful be­fore-and-after any­way.

Law en­force­ment agen­cies of­ten are un­clear about the spe­cific crime-fight­ing ben­e­fits they hope to achieve when adopt­ing tech­nolo­gies. Those tech­nolo­gies tend to spread by word of mouth: If Depart­ment A has a new tool, then Agency B wants it, too — of­fi­cials don’t al­ways put much thought into the pre­cise pur­pose of a tool or whether some­thing else (maybe some­thing they al­ready have or some­thing cheaper) could achieve the same thing.

A good ex­am­ple is the cu­ri­ous case of au­to­matic li­cense plate read­ers: fixed cam­eras, or cam­eras in­stalled on po­lice cars, that cap­ture and dig­i­tize images of li­cense plates. Ini­tially, LPR tech­nol­ogy was sold as a way to re­duce car theft. But with auto theft de­clin­ing any­way, and at $10,000 or more for each mo­bile unit and $100,000 per fixed unit — not in­clud­ing main­te­nance, up­grades, and data stor­age and re­trieval sys­tems — it was hard to jus­tify the cost. Then it oc­curred to de­part­ments that they could store the data and call it up in in­ves­ti­ga­tions; if they are look­ing for a van with a cer­tain plate num­ber, stored records may in­di­cate where that van trav­els fre­quently or even where it parks. One hears anec­dotes about LPR help­fully de­ployed this way. And yet, suc­cess sto­ries tend to rely on ran­dom­ness: A po­lice car hap­pened to drive by the van at some point and dumped that data into the sys­tem. The ben­e­fits of what is ba­si­cally ran­dom en­force­ment are likely to be small.

The lat­est emerg­ing jus­ti­fi­ca­tion for LPR is as part of polic­ing for profit. A pri­vate ven­dor, Vig­i­lant So­lu­tions, gives LPR tech­nol­ogy to de­part­ments for “free” and in­cor­po­rates all plate cap­tures into its data­base. Vig­i­lant then uses that data to cre­ate an ex­ten­sive “hot list,” com­posed, in par­tic­u­lar, of ve­hi­cles with out­stand­ing traf­fic fines. When cops get a ping they can pull a car over; po­lice cruis­ers in some ju­ris­dic­tions even are be­ing equipped with credit card ma­chines for in­stant pay­ment. What’s in this for Vig­i­lant? It gets a 25 per­cent “pro­cess­ing fee.”

Even if the ben­e­fits of cer­tain polic­ing tac­tics can be as­sessed, fig­ur­ing out the full costs of any given tac­tic or tech­nol­ogy is es­sen­tial, al­beit dif­fi­cult. We can cal­cu­late the price tag of the tech­nol­ogy it­self, or per­haps the cost of more of­fi­cers or new train­ing. But what about the cur­tail­ing of civil lib­er­ties? What about the so­cial costs? How do you put a price on pub­lic dis­sat­is­fac­tion with the way the po­lice are do­ing their jobs? What about a po­ten­tial loss of com­mu­nity trust?

Use of stop-and-frisk has been on the de­cline in some ma­jor cities, in part be­cause of lit­i­ga­tion (a cost in and of it­self), but Don­ald Trump re­vived the de­bate with a shout-out dur­ing his pres­i­den­tial cam­paign. The ev­i­dence on whether stop-and-frisk works is thin. A 2008 study found that more po­lice stops led to fewer rob­beries, bur­glar­ies, ve­hi­cle thefts and homi­cides — but did not re­duce as­saults, rapes or grand lar­ceny. A 2014 study took the ear­lier study’s method­ol­ogy to task and called into ques­tion the rob­bery and bur­glary re­sults. One re­view of the em­pir­i­cal ev­i­dence sug­gested that dur­ing the height of stop-and-frisk in New York, fewer youth were car­ry­ing guns. Still, that study’s au­thor con­ceded that the “strong­est ar­gu­ment that New York City’s ag­gres­sive polic­ing strate­gies . . . con­trib­uted to its plum­met­ing crime rate” was “the ab­sence of al­ter­na­tive ex­pla­na­tions.” That’s not so­cial sci­ence, it’s con­jec­ture. And, of course, New York has since dra­mat­i­cally re­duced stopand-frisk, and crime con­tin­ues to fall.

Even if stop-and-frisk were ef­fec­tive in cur­tail­ing gun vi­o­lence( putting con­sti­tu­tion­be­ing

If Depart­ment A has a new tool, then Agency B wants it, too — of­fi­cials don’t al­ways put much thought into the pre­cise pur­pose of a tool or whether some­thing else (maybe some­thing they al­ready have or some­thing cheaper) could achieve the same thing.

al ar­gu­ments to one side), we would still have to look at the so­cial costs of the prac­tice. We’d want to put a value on a loss of trust in the po­lice, con­tribut­ing to an en­vi­ron­ment in which some peo­ple in some neigh­bor­hoods don’t call the cops when shots are fired. That’s why ShotSpot­ter is ap­peal­ing in the first place. We should also con­sider the phys­i­cal vi­o­la­tion, the loss of time and lib­erty, the psy­cho­log­i­cal costs of stop-and-frisk. Sup­pose the cost of be­ing stopped out of the blue, put against a wall and hav­ing hands run over your body is equiv­a­lent to $100, which does not seem en­tirely un­rea­son­able. In New York in 2011 there were more than 300,000 stop-and-frisks, at a cost of more than $30,000,000, us­ing this es­ti­mate. That gives a sense of what the ben­e­fits would have to be in or­der to jus­tify those costs.

The chal­lenges of cal­cu­lat­ing the costs and ben­e­fits of polic­ing should not dis­suade us. If BCA can be ap­plied to com­pli­cated and con­tro­ver­sial sub­jects such as en­vi­ron­men­tal reg­u­la­tion, such cal­cu­la­tions can be ap­plied to pub­lic safety.

It would help if pub­lic safety agen­cies and con­trac­tors were more forth­com­ing with their data. When trade se­crets are a con­cern, pri­vate com­pa­nies such as ShotSpot­ter and Vig­i­lant could make their data avail­able for re­searchers but re­quire nondis­clo­sure agree­ments. A ShotSpot­ter board mem­ber was quoted in Forbes as say­ing it would be ben­e­fi­cial to have an in­de­pen­dent study. Com­pa­nies that be­lieve in their prod­ucts should be will­ing to un­dergo out­side test­ing. Tech com­pa­nies could also lend their ex­per­tise to help po­lice de­part­ments mod­ern­ize their data col­lec­tion and anal­y­sis ef­forts.

Gov­ern­ment could play a big role here. In­deed, al­though there is de­bate and con­fu­sion over the role of the fed­eral gov­ern­ment in lo­cal law en­force­ment, dol­lars spent by the fed­eral gov­ern­ment in assess­ing what keeps us safe, and what is plain waste­ful, are likely to be dol­lars well spent.

Not-for-prof­its have a part, too. They can team up with polic­ing agen­cies to run BCA eval­u­a­tions on tac­tics and tools, such as ve­hi­cle pur­suits and deesca­la­tion train­ing.

Pub­lic safety is foun­da­tional in so­ci­ety. We spend a for­tune on it. It seems that it’s not too much to ask that we de­vote re­sources to fig­ur­ing out what works and what does not, and whether we could do bet­ter.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.