Phillip Atiba Goff: I don’t like talk­ing to peo­ple about what I do for a liv­ing. It’s not that I don’t like my job—I love my job. It’s just it’s not always the most com­fort­able cock­tail par­ty reveal? Right? It’s like, Hey, Steve. Nice to meet you. What is it you do?” 

Oh, I’m in acqui­si­tions. Remind me what you do again, Phil?”

Oh yeah, I’m in the end­ing racism and state-sanctioned lynch­ings business.”

Right, right. Right. Cool cool cool. Cool. I hear that’s like a growth indus­try. Oh hey look, alcohol.”

Cuz the idea is it’s…it’s awk­ward, right. And it makes sense that it’s awk­ward. It’s a heavy top­ic, right. Because par­tial­ly, the worst things that human beings do to oth­er human beings, they often have the specter of racism on top of them. But it’s also awk­ward because frankly, a lot of peo­ple don’t have a lot of hope that we’re fix­ing racism any­time soon, par­tic­u­lar­ly in the con­tem­po­rary sort of polit­i­cal envi­ron­ment that we live in. 

So, today I want to talk to you just a lit­tle bit about how the sci­ence of racism, how racism actu­al­ly func­tions, can bring a lit­tle bit of hope to these dif­fi­cult issues, with­out even need­ing to be espe­cial­ly polit­i­cal. And bet­ter than that, how the sci­ence of racism can lead to some action­able solu­tions to these seem­ing­ly impos­si­ble problems. 

Alright, so what kind of prob­lems exact­ly am I talk­ing about? Well for sure, I mean…racism in edu­ca­tion, and hous­ing and policing—that’s the work that I do every day. But I also mean recruit­ing. And HR prac­tices. And mar­ket­ing, right. How we come to under­stand how vul­ner­a­ble com­mu­ni­ties, those very com­mu­ni­ties that are the dri­vers of cul­ture in every coun­try across the world, how vul­ner­a­ble com­mu­ni­ties are gonna respond to prod­ucts. And to brand identities. 

The 10th sustainable development goal, reduced inequalities.

See, we’re not just talk­ing about the tenth sus­tain­able devel­op­ment goal. If we do racial equi­ty right, we’ve got impli­ca­tions for all sev­en­teen. In oth­er words this isn’t just a moral issue, this is a P&L issue, right. And as mar­kets diver­si­fy, ignor­ing racial equi­ty is going to leave the com­pe­ti­tion behind in the same way that lag­ging on green ener­gy and gen­der equi­ty has. 

So, in my every­day, I deal with racism in polic­ing. But my hope is a lit­tle bit today you all will help me to trans­late from the world I live in to the worlds that are clos­est to you so we can trans­form these awk­ward con­ver­sa­tions about racism, make them just a lit­tle bit less painful, and a lit­tle bit more com­fort­able. Okay, so can I ask you to do that with me a lit­tle bit? Yeah? So it’s okay to speak back, right. Cameras are on me. Thank you.

Alright. So, to get how the sci­ence of racism leaves me opti­mistic, it’s impor­tant we define that term racism.” The most com­mon def­i­n­i­tion of racism is that racist behav­iors stem from defec­tive hearts and minds. And if you think about the ways that we talk about try­ing to solve this prob­lem, you’ll hear it. To fight racism we need to com­bat igno­rance. To fight racism we need to stamp out hatred. Hearts and minds, right. 

Now, a lot of my opti­mism comes from not using this def­i­n­i­tion because, and this is impor­tant, it’s total­ly wrong. Okay. Wrong in every way. But before we talk about the def­i­n­i­tion I do use, I actu­al­ly think it’s use­ful to talk about how this hearts and minds def­i­n­i­tion of racism, the wrong def­i­n­i­tion, how it gets to be so sticky. How it’s so hard to get rid of. So I want to use an illus­tra­tive examples. 

This is a pic­ture of Emmett Till. And for those of you who don’t know where this is going, Emmett Till was a 14-year-old black boy born in Chicago, Illinois to work­ing class par­ents. And as he grew up, he devel­oped a speech imped­i­ment that made it sound like he was whistling when he spoke. 

So his moth­er, Mamie Till, decid­ed it’d be a good idea to send young Emmett away to live with his rel­a­tives in Mississippi dur­ing the sum­mer of 1955. So he would­n’t get bul­lied. What she was­n’t expect­ing is that a white woman would lie about her child. 

See, in August of that year, a white woman told her hus­band that the 14-year-old Till had sex­u­al­ly propo­si­tioned her. Which he had­n’t. But the result was worse than any bul­ly­ing she could’ve imag­ined. Because on the evening of August 28th, 1955, Roy Bryant and JW Milam—these two gen­tle­man in the white—they showed up armed to Emmett Till’s great uncle’s house. They abduct­ed him at gun­point, tied him up, sav­age­ly beat and muti­lat­ed his face and his tor­so, and when when they real­ized that they’d killed him they tied a cot­ton gin fan to his body to sink it to the bot­tom of the Tallahatchie River. 

Pictures of Till’s open cas­ket—and for the squea­mish you might want to look away for a cou­ple of slides. Pictures like this, they sparked not just local but inter­na­tion­al out­rage. Because they shone a light on the racist, vicious vio­lence in the United States. But more than that, by this time in 1955 there’d been over 4,000 of these kind of extrale­gal mur­ders of black peo­ple, includ­ing chil­dren, since the US had abol­ished slav­ery. Four thou­sand.

And see­ing just one image like that, one image like the lynch­ing in Marion, Indiana, like this, where peo­ple brought their chil­dren to see this, it’s easy to imag­ine how you can just reduce this all to evil. This is what evil looks like. What decent per­son could do this to a 14-year-old boy? Our eyes con­vince us that some­thing sick in the hearts and minds of peo­ple is what’s required for this to happen. 

But his­to­ry teach­es us what our eyes can’t. Genocide, lynchings—the worst things that human beings can do to oth­er human beings, those things hap­pen at scale. They’re not individual—you can­not reduce them to indi­vid­ual hearts and minds or even a group of peo­ple’s feel­ings. This is not a those peo­ple over there” prob­lem, this is an us prob­lem. And his­to­ri­ans, social sci­en­tists of racism and group con­flict, they’ve known that for some time. For gen­er­a­tions we’ve referred to this as the banal­i­ty of evil—it’s easy for this to hap­pen in large groups. 

So if the world has known this for some time, why is it that inter­na­tion­al­ly the lan­guage for racism still reduces to indi­vid­ual feel­ings? Well as a psy­cho­log­i­cal sci­en­tist, I can tell you there’s a great deal to be gained from try­ing to take some­thing that is incred­i­bly uncom­fort­able and make it com­fort­able. And our con­ver­sa­tions about racism? They’re pret­ty uncom­fort­able, right. 

So, racism is an uncom­fort­able top­ic but, if I tell you that racism is real­ly just about what’s in your char­ac­ter, all of a sud­den you have the capac­i­ty to be in con­trol of that. Things you’re in con­trol of are way more com­fort­able, right. Cuz if the prob­lem real­ly is just my feel­ings, then all I have to do, all any of us can do, is get our feel­ings right—be of good char­ac­ter. Because if I’m of good char­ac­ter I can’t be racist, when I see racism in the world, I’m not impli­cat­ed. That is way more com­fort­able. And like I said, also wrong. 

So there are two ele­ments of that hearts and minds def­i­n­i­tion that are absolute­ly wrong, right. The first is that it’s sci­en­tif­i­cal­ly garbage, and the sec­ond, it makes it hard­er to solve the prob­lem. And here’s what I mean.

So for the past cen­tu­ry or so, social psy­chol­o­gists have known that atti­tudes like prej­u­dice are real­ly weak pre­dic­tors of behav­iors like dis­crim­i­na­tion. In fact atti­tudes pre­dict about 10% of behav­iors, at best, through meta-analysis. But even more impor­tant­ly than that, even if atti­tudes were great pre­dic­tors of behav­ior, a hearts and minds def­i­n­i­tion of racism says the solu­tion to racism is what? Salvation? I don’t know about you but sal­va­tion for me is real­ly dif­fi­cult just one on one. And it does­n’t real­ly scale. And that’s the rea­son why this hearts and minds def­i­n­i­tion is a bad diag­no­sis. It’s not just that it’s wrong. It’s that it makes it hard­er to solve the prob­lem. And worse, when we fix­ate on try­ing to change behav­iors, we end up dis­tract­ed by try­ing to reha­bil­i­tate poten­tial­ly racist actors and ignor­ing the accu­mu­la­tion of harms that are hap­pen­ing to vul­ner­a­ble com­mu­ni­ties right in front of us. By defin­ing the prob­lem as hearts and minds, we end up being dis­tract­ed try­ing to fix racist peo­ple instead of fix­ing racist outcomes. 

And I was encour­aged to make all of this very dif­fi­cult mate­r­i­al acces­si­ble, so I have a meme that I think explains all of this, very simply. 

A variation of the Distracted Boyfriend meme: the guy (labeled "aspiring anti-racist") looks at a woman passing by (labeled "saving souls of racists") while his girlfriend (labeled "changing behaviors") stares at him in irritation

Nobody wants to be this guy. Aspiring anti-racist…not chang­ing behavior…this guy’s not a role model. 

But I do have some good news. Which is that in my field, we have a dif­fer­ent def­i­n­i­tion which is both more accu­rate and more action­able. As a psy­cho­log­i­cal sci­en­tist we actu­al­ly define racism as the accu­mu­lat­ed pat­tern of behav­iors that dis­ad­van­tage one racial group and advan­tage anoth­er racial group, as well as the sys­tems to facil­i­tate that. 

And this won­der­ful thing hap­pens when you define the prob­lem this way. You can start mak­ing it solv­able, right. And when you can do that, this hope­less thing can become inspir­ing. And we have this great example…but most­ly with sex­ism and cor­po­rate com­pen­sa­tion, that comes to us from our friends at Salesforce. 

See in 2015, the folks at Salesforce had a concern…right, some of you are nod­ding your heads—you know this sto­ry. They had a con­cern that they weren’t pay­ing men and women the same amount for doing the same job. So what did they do? Well they mea­sured whether not they were pay­ing men and women the same amount for the same job. 

What did they find? They weren’t pay­ing men and women the same amount for the same job. So what did they do? They start­ed pay­ing men and women the same amount for the same job, right. 

Now the sur­pris­ing thing for most peo­ple is, the year lat­er when they audit­ed again, they found the same gen­der gap. Because they had acquired dur­ing that year a num­ber of addi­tion­al com­pa­nies that came in with the gen­der gap they’d had before. So the mirac­u­lous thing that should be a sur­prise for exact­ly nobody in the entire town of Davos, the take-home mes­sage is you have to mea­sure the things you care about, ana­lyze them, and then opti­mize for them reg­u­lar­ly. It’s called run­ning a busi­ness with goals. It’s how every busi­ness has ever worked through­out his­to­ry. We just don’t think about all of our val­ues the same way. 

So in my day job, I do that kind of thing, but for racism in polic­ing. My cen­ter, the Center for Policing Equity, hosts the largest-known col­lec­tion of police behav­ioral data in the world. It’s a super hum­ble­brag but we can talk about that lat­er, right. And we use that data to diag­nose and ana­lyze racial dis­par­i­ties and racial bias in police behav­ior, and give that back to com­mu­ni­ties and law enforce­ment so they can hold them­selves account­able to those goals. 

Now, one of the first things that peo­ple will ask me when they hear about what I do is, Alright be hon­est with me, do police real­ly want to be held account­able to data?” 

The sur­pris­ing answer for lots of peo­ple, is that they already do. Not just in North America, but in Latin America, Europe, Oceania, Africa, South and Southeast Asia, all over the globe increas­ing­ly law enforce­ment is using a tool called CompStat. Now CompStat when used prop­er­ly, used appro­pri­ate­ly, is a tool that allows you to track crime data, iden­ti­fy pat­terns, and then hold your­self account­able to pub­lic safe­ty goals. And it usu­al­ly works either by direct­ing resources or chang­ing behav­ior once folks show up. So if I’m using CompStat and I see mug­gings in this neigh­bor­hood, I deploy more patrols in this neigh­bor­hood. If I’m using CompStat and I find out that this com­mu­ni­ty is not giv­ing me tips any­more, they’re not talk­ing to us, I want to change how I’m talk­ing to that com­mu­ni­ty. And when you define racism in terms of mea­sur­able behav­iors, you can do the same thing. You can cre­ate a CompStat for justice. 

Now, to be fair jus­tice is kind of tricky to mea­sure. I can’t just count the num­ber of Asian folks who get pulled over and then com­pare that to the num­ber of Asian folks who live in a neigh­bor­hood. That’s called a cor­re­la­tion and I’m a sci­en­tist so…I’m kind of aller­gic to that. 

But what we can do, we can col­lect let’s say police use of force data. And then I can inte­grate that with local data on pover­ty, and crime, and hous­ing inequal­i­ty, and employ­ment inequal­i­ty, and health out­comes, and edu­ca­tion­al out­comes. And then when we ana­lyze those data, we can cal­cu­late rough­ly the por­tion of racial dis­par­i­ties that might be able to be attrib­uted to things that police can’t be account­able for, they can’t con­trol. Like crime and pover­ty. And the por­tion of the racial dis­par­i­ties that police might be account­able for. Like their poli­cies and behav­iors. So if you want­ed to do bet­ter, at least you’d have a place to start. 

Let me give you an exam­ple of how this might work. In fact how it did work. In Las Vegas, we were able to look at their data and help them to see that a dis­pro­por­tion­ate num­ber of their use of force inci­dents, they actu­al­ly fol­lowed from foot pur­suits. Now why would that be? Why would a foot pur­suit lead to a dis­pro­por­tion­ate num­ber of use of force incidents? 

Well if I’m an offi­cer, if I’m chas­ing after some­body, I am con­vinced that they are a bad guy, right. Nobody runs from the law but the bad guys; at least that’s what offi­cers think. But I’m also run­ning in heavy gear and poly­ester (not a breath­able fab­ric). That means my heart rate is up. And I’m sweatin’ and my adren­a­line’s high. That means even if the sus­pect sur­ren­ders at the end and say, Please don’t hurt me,” they’re get­ting a shot to the kid­neys for the price of mak­ing me run. 

And as soon as we gave that infor­ma­tion, just those analy­ses, back to the depart­ment and to the com­mu­ni­ty they said, We can train offi­cers bet­ter than this.” Like any mar­riage coun­selor, we can teach you to count to ten. Or don’t touch them until your back­up has shown up. 

So the fol­low­ing year, across the board—not just in foot pur­suit, across the board, Las Vegas Metro reduced their use of force by 23%. And in the dozens of depart­ments where we’ve worked and we’ve employed these sim­i­lar sorts of strate­gies, we’ve had sim­i­lar­ly pos­i­tive out­comes. So whether it’s foot pur­suit in Las Vegas, or it’s immi­gra­tion enforce­ment in Salt Lake City and Houston, or it’s the home­less prob­lem in Minneapolis, or just low-level enforce­ment of low-level rules vio­la­tions in Baltimore, across our dozens of part­ners we’ve seen an aver­age of 25% few­er arrests, 26% few­er use of force inci­dents, and 13% few­er officer-related injuries. This is healthy for them as well.

Another way to put this is, if you give peo­ple a solv­able prob­lem, they’ll get busy try­ing to solve it. I mean, these are lit­er­al­ly detec­tives, right. They’re gonna try and solve a prob­lem you put in front of them, like any­body in any orga­ni­za­tion ever. 

Just think about how this works on the flip side. How many indus­tries talk a good game about equi­ty, diver­si­ty and inclu­sion, and mea­sure noth­ing. I can’t tell you the num­ber of uni­ver­si­ties that’ve asked me to show up and look at their strate­gic diver­si­ty policy—which mod­el is usu­al­ly sit­ting right next to their cap­i­tal cam­paign, with all of its lead­ing met­rics, right, on fundrais­ing and suc­cess. But there’s their strate­gic diver­si­ty pol­i­cy, and it says We at Blank University are com­mit­ted to diver­si­ty!” And that is the end of the strate­gic pol­i­cy. No met­rics, no sen­ti­ment sur­vey, noth­ing, right. And I got­ta say, if that is how you think about doing diver­si­ty…you’re not help­ing.

And I got­ta say anoth­er way to think about this, I’ll let you in on a lit­tle secret. I have been black my entire life. I took like a week off in col­lege, but oth­er than that straight through. And I have nev­er felt the sweet relief of jus­tice because some­one told me they were com­mit­ted to feel­ing bet­ter about me. 

Let me flip that around a lit­tle bit. That white woman whose lie cost Emmett Till his life, she’s still alive. And she recent­ly admit­ted to the lie, and talked about the change of heart she’d had. How awful she felt. How ter­ri­ble that was. No one could’ve deserved that. 

And I ask you, does that change of heart make it bet­ter? Or is Emmett Till still dead? Because that’s the stark­ness of think­ing about racism as hearts and minds ver­sus behav­iors. So if we’re think­ing about reme­dies, we have to keep that in mind. 

So I don’t know what reme­dies the peo­ple in this room, or the peo­ple who’re watch­ing this else­where care most about. Maybe it’s the stuff I do every day. Great. We’ll have a great con­ver­sa­tion going for­ward. Or maybe it’s the new hir­ing pol­i­cy that has fan­tas­tic prin­ci­ples but no met­rics attached to it. Or it’s the com­mu­ni­ty engage­ment ini­tia­tive that gives you fan­tas­tic sto­ries, but has no clear way of mea­sur­ing the ben­e­fit to the com­mu­ni­ty, or to the com­pa­ny for that mat­ter. Or it’s the mar­ket­ing strat­e­gy for the oppor­tu­ni­ty zone with real­ly thin KPIs. But the point that I think you’re pick­ing up on, if an orga­ni­za­tion, if any busi­ness, has­n’t fig­ured out how to mea­sure the ways it facil­i­tates racially-disparate impacts on the com­mu­ni­ties of touch­es, it is at both moral and finan­cial per­il, at risk. We already know this from envi­ron­men­tal impact. We know this for eco­nom­ic impact. But how many envi­ron­men­tal and eco­nom­ic impact plans actu­al­ly have met­rics built in look­ing at racial dis­par­i­ties? How much of what busi­ness­es are doing right now are designed to mea­sure what hap­pens to the most marginalized? 

Now if the folks in this room know, fan­tas­tic. Please share. And if you don’t, the good news is that means there’s some con­crete things you can do as soon as you leave the room. And believe it or not, that’s why I actu­al­ly love my job. Why I think lots of peo­ple can love doing this work. Because every day, I get to wake up and try and solve a prob­lem that most of the world thinks is impossible. 

It is also why I hate talk­ing about my job when oth­er peo­ple think that my job is about their char­ac­ter. When they imag­ine that the racism I’m work­ing to erad­i­cate every day reduces to what’s in their hearts and their minds and it’s my job to tell them whether or not they’re racists. 

Another lit­tle secret: I don’t care. I don’t care about whether or not some­body’s racist. I don’t want to have that con­ver­sa­tion. That con­ver­sa­tion is awk­ward, but worse it’s frus­trat­ing. And it is use­less. I would much rather live in a world where our con­ver­sa­tions about racism, are con­ver­sa­tions about what we can do. How we can best mea­sure the prob­lems and deliv­er solu­tions to the com­mu­ni­ties that need them. Because solv­ing impos­si­ble prob­lems, for real is a growth indus­try. And it does­n’t need to be uncom­fort­able to talk about at all. Thank you guys for lis­ten­ing. And let’s have a conversation.