Data & its Discontents vol. 1: Abuse of Data, Abuse of Bodies

Prologue

“Now the reason that I use that particular word is because for me it has a, what is what I call a “basicness” about it, that is to say that I feel that in any case the first niggers were of course the field niggers and upon that is really the basis of the American economic system, without field niggers we wouldn’t really have such a great and grand economy.” — Julius Eastman, from his spoken introduction to the 1980 performance of ‘Evil Nigger’, ‘Gay Guerrilla’ & ‘Crazy Nigger’ at Northwestern University

A few months ago I attended an exceptional free performance of radical avant-garde composer Julius Eastman’s controversially titled Crazy Ni**er. Elsewhere in the city that night, Ta-Nehisi Coates (hereafter ‘TNC’) spoke to a sold-out library crowd as part of his book tour promoting the also radical & exceptional Between The World And Me.

“The richest men in America lived in the Mississippi River Valley, and they made their riches off our stolen bodies. Our bodies were held in bondage by the early presidents. Our bodies were traded from the White House by James K. Polk. Our bodies built the Capitol and the National Mall.” — Ta-Nehisi Coates, from Between the World and Me

During the requisite post-concert loiter & bullshit, conversation turned to Coates & his book — which i had almost finished & of which my interlocutor had read an extended excerpt. (shoutout to my sole acquaintance w/ equal interests in experimental composition & supra-tumblr radical left public discourse)

We discussed the salient contrast between the incisive , historically-grounded , research-heavy ‘objective’ analysis & argumentation (which first put us each in TNC’s thrall) & the more personal, memoir-istic & emotive nature of the new book. Not that his prose ever lacked pathos or rich, illustrative anecdotes , but this emphatic ‘subjective’ turn meaningfully differed from the journalistic tone of his famed prior magnum opus (The Case for Reparations), & the economical argumentation of his prior rhetorical duals.

Here is what I would like for you to know: In America, it is traditional to destroy the black body — it is heritage.

— TNC, BTWAM

Taking the form of a letter to his son — recounting his most intimate formative experiences of race , violence , love & atrocity — his new book couldn’t be more inward-facing. Same time, being published to the world by a premier international publisher & accompanied by an extensive marketing campaign (& book tour), it couldn’t be more outwardly oriented. (this dynamic is itself a sort of nested, fractal recurrence of the subjective / objective contrast i’m drawing, & accentuates the tension between the two)

ABUSE OF BODIES

My understanding of the universe was physical, and its moral arc bent toward chaos then concluded in a box.

— TNC, BTWAM

The emphatic subjective framing of this very public argument is paralleled & reinforced by TNC’s relentless focus on & explicit reference to bodies — specifically black bodies, & particularly the plunder and destruction of those bodies. The focus is so jarring & unremitting , there’s no mistaking its deliberate strategic centrality in the work. This is bringing the abstract & discursive forcefully back to the concrete & palpable.

It struck me as Ta-Nehisi Coates — the prodigious rhetorician & exhaustive researcher — in the throes of rhetoric-&-research-overload, pressed by the ever-growing weight of actual, physical black bodies piling up without pause, executed by actual physical police using actual physical guns, bullets, choke-holds, ‘rough rides’ and denial of critical medical attention. (this weight further intensified not only by his own black ‘skin-in-the-game’, but also — crucially — his son’s)

But all our phrasing — race relations, racial chasm, racial justice, racial profiling, white privilege, even white supremacy — serves to obscure that racism is a visceral experience, that it dislodges brains, blocks airways, rips muscle, extracts organs, cracks bones, breaks teeth.

— TNC, BTWAM

In the wars of words, the semantics, the ‘logical argumentation’, the second-guessing of data, the policy debates, the demagoguery & platitudes of pundits, politicians & the public—one thing is undeniable: bodies continue to pile up. And his or his son’s could be next; there’s no plausible cause to imagine otherwise.

Rational discourse & quantitative analysis shrivel in the face of real people, real lives, real bodies being destroyed & decomposed (particularly when those bodies are indistinguishable from your own, your family’s, & your community’s). These bodies are destroyed in spite of — possibly obscured, elided, euphemized & thereby abetted by — the words & data.

And so TNC refuses to allow the shroud to stand. He won’t let us look away, won’t soften the blow of a crushed larynx , snapped spine , a bullet-riddled body of a teenager , or a bullet-riddled body of a 12 year old playing alone in a park .

Abuse of Data (I)

The entire narrative of this country argues against the truth of who you are.

— TNC, BTWAM

Some days after our post-concert convo, my fellow White Experimental Music & Politics Enthusiast (lets call him ‘WEMPE’) posted a NYT op-ed to facebook. The headline read Police Killings of Blacks: Here Is What the Data Say. It contained the following assertions [all bolding added by me for emphasis]:

• I have no special insight into the psychology of police officers or into the complicated forensics involved in [cases of police killings of unarmed black civilians].
• As an economist who has studied racial discrimination, I’ve begun to look at these deaths from a different angle. There is ample statistical evidence of large and persistent racial bias in other areas — from labor markets to online retail markets. So I expected that police prejudice would be a major factor in accounting for the killings of African-Americans. But when I looked at the numbers, that’s not exactly what I found.
• I’m not saying that the police in these specific cases are free of racial bias. I can’t answer that question. But what the data does suggest is that eliminating the biases of all police officers would do little to materially reduce the total number of African-American killings. Police bias may well be a significant problem, but in accounting for why some of these encounters turn into killings, it is swamped by other, bigger problems that plague our society, our economy and our criminal justice system.
• The data is unequivocal. Police killings are a race problem: African-Americans are being killed disproportionately and by a wide margin. And police bias may be responsible. But this data does not prove that biased police officers are more likely to shoot blacks in any given encounter.
• Instead, there is another possibility: It is simply that — for reasons that may well include police bias — African-Americans have a very large number of encounters with police officers. … having more encounters with police officers, even with officers entirely free of racial bias, can create a greater risk of a fatal shooting.
• Arrest data lets us measure this possibility. For the entire country, 28.9 percent of arrestees were African-American. This number is not very different from the 31.8 percent of police-shooting victims who were African-Americans. If police discrimination were a big factor in the actual killings, we would have expected a larger gap between the arrest rate and the police-killing rate.
• This in turn suggests that removing police racial bias will have little effect on the killing rate. [!]
• If the major problem is then that African-Americans have so many more encounters with police, we must ask why.
• But this is too large a problem to pin on individual officers.
• First, the police are at least in part guided by suspect descriptions. And the descriptions provided by victims already show a large racial gap: Nearly 30 percent of reported offenders were black. So if the police simply stopped suspects at a rate matching these descriptions, African-Americans would be encountering police at a rate close to both the arrest and the killing rates. [!!!]
• Second, the choice of where to police is mostly not up to individual officers. And police officers tend to be most active in poor neighborhoods, and African-Americans disproportionately live in poverty.
• In fact, the deeper you look, the more it appears that the race problem revealed by the statistics reflects a larger problem: the structure of our society, our laws and policies. [he goes on to thoughtfully & at times poetically expound on these structural & institutional issues. well & good, not at issue here]
• Individual police officers did not set these economic policies that limited opportunities or create the harsh sentencing policies…
• We should eliminate police prejudice because it is wrong and because it undermines our democracy. It blights — and all too often destroys — lives.
• But there are also structural problems underpinning these killings. We are all responsible for those.

i reacted emotionally , & at length .

i’ll recount some of those reactions below.

The point of this language of “intention” and “personal responsibility” is broad exoneration. Mistakes were made. Bodies were broken. People were enslaved. We meant well. We tried our best. “Good intention” is a hall pass through history, a sleeping pill…

— TNC, BTWAM

Now , it should be said that the op-ed author (as well as good liberal leftist WEMPE) is concerned with the problem of racially biased police violence & offers the standard liberal structural critiques & appeal to shared collective responsibility. I don’t argue with this, except to say that his conclusion (after such tortuous & ultimately insidious ‘data-driven’ argumentation) is as baffling as it is vacuous & obvious.

He meant well.

Abuse of Data (II)

…you know now, if you did not before, that the police departments of your country have been endowed with the authority to destroy your body. It does not matter if the destruction is the result of an unfortunate overreaction. It does not matter if it originates in a misunderstanding. It does not matter if the destruction springs from a foolish policy.

…The destroyers will rarely be held accountable. Mostly they will receive pensions. And destruction is merely the superlative form of a dominion whose prerogatives include friskings, detainings, beatings, and humiliations. All of this is common to black people. And all of this is old for black people. No one is held responsible.

— TNC, BTWAM

Theres so much going in the above op-ed , i’ll try to take it piece-by-piece.

but tl;dr: It’s always worthwhile to point out structural and institutional factors responsible for horrific injustice. But this needless, reaching, statistically dubious framing — dismissing personal racial bias in police just to reach an over-obvious conclusion — is emblematic of whats wrong with so much journalism in general, & misuse & abuse of data in particular.

It also begins to make a dangerous argument for exonerating individual culpability in racially biased police behavior. And conceivably sketches a ‘data-driven,’ ‘scientific’ justification for the ongoing rampant prosecutorial immunity of individual police in these cases.

The struggle to understand is our only advantage over this madness.

— TNC, BTWAM

ok,

so starting with the headline

…Here Is What the Data Say

(saying nothing of the archaic nerd faux pas of ‘data’ as a plural count noun) , data doesn’t SAY anything, data is numbers; people say things about data. Mostly dumb ass things. The phrase “the data say” almost invariably precedes some sort of bullshit misrepresentation or overstatement of fact.

I realize that columnists are often not responsible for headlines (chosen by editors), so lets take some statements from the body of the text:

There is ample statistical evidence of large and persistent racial bias in other areas … So I expected that police prejudice would be a major factor in accounting for the killings of African-Americans. But when I looked at the numbers, that’s not exactly what I found.

what the data does suggest is that eliminating the biases of all police officers would do little to materially reduce the total number of African-American killings.

The data is unequivocal, Police killings are a race problem: African-Americans are being killed disproportionately and by a wide margin. And police bias may be responsible.

But this data does not prove that biased police officers are more likely to shoot blacks in any given encounter.

Arrest data lets us measure this possibility.

If police discrimination were a big factor in the actual killings, we would have expected a larger gap between the arrest rate and the police-killing rate.

This in turn suggests that removing police racial bias will have little effect on the killing rate.

Sometimes things are ‘found’ in the numbers, sometimes data ‘suggests’ things, sometimes ‘the data is unequivocal’, sometimes the ‘data does not prove’ something (yet disproves it?), sometimes the ‘data lets us measure’ something.

Such haphazard phrasings strike me either as utterances of someone deeply confused about the nature of data & what statistical methods do , a charlatan trying to sell a compelling narrative , or at best an accomplished academic being a little too loose with the numbers & his characterizations of them on deadline for a minor op-ed.

But this is deeper & more troubling than semantics. The claims are not even mutually consistent.

The data is unequivocal, Police killings are a race problem

But this data does not prove that biased police officers are more likely to shoot blacks in any given encounter.

This in turn suggests that removing police racial bias will have little effect on the killing rate.

So the data is UNEQUIVOCAL on the issue of police killings as a race problem, but they don’t prove that individual police are biased, yet they somehow suggest that racial bias has no effect on police killing of black people. (and of course the headline promises to tell us What The Data Say, and he himself tells us this is what he found in the numbers).

So we went from unequivocal evidence of systemic racial bias, to lack of proof of individual bias, to that somehow equating to DISproof of individual bias effecting rates of police killing.

There are glaring technical & framing problems here (e.g. use of aggregate national statistics to assess very localized [& non-normally distributed] phenomena; his model of determining individual police bias; etc) as well as logical fallacies and gross overstatement of conclusions from cursory data, but i’m gonna move on because :

1. these issues are a bit esoteric,
2. i can’t say too much with certainty because no data sources or statistical methods/results are cited &
3. some of this will be covered below

…what one “means” is neither important nor relevant. It is not necessary that you believe that the officer who choked Eric Garner set out that day to destroy a body. All you need to understand is that the officer carries with him the power of the American state and the weight of an American legacy, and they necessitate that of the bodies destroyed every year, some wild and disproportionate number of them will be black.

— TNC, BTWAM

So , we’ve seen that a central claim of the NYT piece is: there is no evidence of individual racial bias in police killings, and somehow this proves that racial bias of individuals plays no part (again, Argument from ignorance fallacy), despite indisputable evidence of racial bias in the aggregate.

Indeed, he begins the piece:

I have no special insight into the psychology of police officers …involved in [cases of police killings of unarmed black civilians].

well let me google that for u:

• BBC: Ferguson police racially biased says US Justice Department
• FBI report on White Supremacist Infiltration of Law Enforcement
• Miami New Times: MIAMI BEACH POLICE CAUGHT SENDING HUNDREDS OF RACIST AND SEXIST E-MAILS
• USA Today: 7 San Francisco officers suspended over racist texts
• DailyKos: Texas sheriff involved in the death of Sandra Bland fired from previous post for racism
• Independent: Cleveland police officer under investigation after promoting ‘racist and anti-Muslim’ messages on Twitter
• St. Louis Post-Dispatch: St. Louis police chief orders probes after officer gets racist letter
• Washington Post: Being a cop showed me just how racist and violent the police are. There’s only one fix.
• Wikipedia: Kerner Commission

now that that’s settled…

You must struggle to truly remember this past in all its nuance, error, and humanity. You must resist the common urge toward the comforting narrative of divine law, toward fairy tales that imply some irrepressible justice. The enslaved were not bricks in your road, and their lives were not chapters in your redemptive history. They were people turned to fuel for the American machine.

— TNC, BTWAM

let’s address the issue of the stats themselves.

There are no citations of sources for the article’s data, or any mention of what statistical methods were used to arrive at his conclusions. The only source mentioned in the body of the article is the F.B.I.’s Supplementary Homicide Report, but presumably a report on homicides does not contain ‘arrest data’ or stats on ‘victim’s descriptions of offenders’, which the author relies on for his analysis.

This lack of sourcing & data transparency is always egregious, but particularly so when dealing with notoriously poorly kept data sets & such grave matters:

FBI director calls lack of data on police shootings 'ridiculous,' 'embarrassing'

Here we have the director of the agency responsible for the SOLE DATA SOURCE REFERENCED by the author admitting that their data is embarrassingly inadequate. And yet the author has no reservations in enlightening us about WHAT THE UNEQUIVOCAL DATA SAY about MATTERS OF LIFE & DEATH, MURDER & CORRUPTION, PLUNDER & IMPUNITY.

FUCK YOU.

…the changes [brought about by struggle] have taught me how to best exploit that singular gift of study, to question what I see, then to question what I see after that, because the questions matter as much, perhaps more than, the answers.

— TNC, BTWAM

ahem,

he proceeds:

As an economist…

yea imma stop u right there.

Only an economist (perhaps only a behavioral economist) could suggest that racial bias among individual police officers is somehow distinct & disentangle-able from the systemic & institutional bias in which they operate — much less that individual bias could be ‘eliminated’ without eliminating (or impacting AT ALL) the institutional bias. This is borderline incoherent to me (and probably to most social scientists with a remotely competent grasp of social cognition & structural bias etc).

A weird wrinkle here: this author appears to BE such a social scientist himself in his more notable work (pointed out to me by WEMPE). In that more celebrated (and seemingly sound) work, his account of prejudice describes something deeply personally ingrained , often subconscious, and likely interwoven with the systems , institutions & cultures in which we operate. By this account, it would be absurd to suggest eliminating one without consequence for the other. I simply can’t reconcile these two contrary conceptions of prejudice / bias , & I doubt he can either.

I would not have you descend into your own dream. I would have you be a conscious citizen of this terrible and beautiful world.

— TNC, BTWAM

I’d like to go deeper into some of the technical issues here but i’m already on at some length , constrained by opaqueness of sourcing & method , & not entirely sure i’m equipped for the sort of critique i’d like to do anyway.

instead i’ll highlight a few last (tho not exhaustive) problematic aspects of the piece & why i find it such a reckless & egregious case of abuse of ‘data’.

• the author’s model of proof of individual police bias is essentially: rates of killing > rates of interaction. This is perhaps a decent rough baseline, but hardly a definitive basis by which to RULE OUT such bias — PARTICULARLY when dealing only with (dubious) aggregate national data which could be expected to drown out & obscure more flagrant local realities. Further, his dismissal of the role of individual bias in biased aggregate interaction rates is as facile, speculative & overstated as the rest of the article in general — hardly conclusive or even cogent — and represents another shaky cornerstone in this house of cards that collapses under mild scrutiny.
• there is dizzying incoherence & internal inconsistency of the most crucial claims:

I’m not saying that the police in these specific cases are free of racial bias. I can’t answer that question.

…Police bias may well be a significant problem…

so far so reasonable

…but in accounting for why some of these encounters turn into killings, it is swamped by other, bigger problems that plague our society, our economy and our criminal justice system.

…what the data does suggest is that eliminating the biases of all police officers would do little to materially reduce the total number of African-American killings.

This in turn suggests that removing police racial bias will have little effect on the killing rate.

…this is too large a problem to pin on individual officers.

Individual police officers did not set these economic policies…

sooo, individual ‘police bias may well be a significant problem’ ….except that the numbers show that it’s not, at least not in regards to killing black people.

predictably, his conclusion is comparably non-sequitur & incoherent:

We should eliminate police prejudice because it is wrong and because it undermines our democracy. It blights — and all too often destroys — lives.

But there are also structural problems underpinning these killings. We are all responsible for those.

He just finished explaining how UNEQUIVOCAL data SUGGESTS that racial bias in individual police plays no part in the rate of killings of black people, yet it blights— AND ALL TOO OFTEN DESTROYS — LIVES ???

According to his whole thesis, ‘eliminating the biases of all police officers would do little to materially reduce the total number of African-American killings’, so what lives are ‘too often destroyed’ by it ?? Non-material total numbers of them ??

I give the fuck up.

• finally, if it’s not already obvious, my biggest problem with this hapless ‘data-driven’ doggerel is how rampant police impunity is bolstered by an argument where only systems, not the agents vested with the supreme power & authority of those systems, are deemed culpable. (a familiar dynamic). This argument is unmistakable in assertions such as:

...eliminating the biases of all police officers would do little to materially reduce the total number of African-American killings.

…removing police racial bias will have little effect on the killing rate.

…this is too large a problem to pin on individual officers.

Systemic racial bias in the criminal justice system has been widely & repeatedly demonstrated & the point clearly needs to be further driven home to effect change (not that it seems to matter to white people). But i worry that weaving this argument in with another essentially exonerating any individual officer of such bias can only be counterproductive to efforts to reduce the actual abuse of bodies.

EPILOGUE

…a society that protects some people through a safety net of schools, government-backed home loans, and ancestral wealth but can only protect you with the club of criminal justice has either failed at enforcing its good intentions or has succeeded at something much darker.

— TNC, BTWAM

During the course of our discussion of the book & our somewhat discomfited expectations, i acknowledged that “We’re not exactly the target audience of the book.” To which WEMPE replied “who is?” My initial answer was “Well, ostensibly his son…” and conversation was tangentially interrupted by mingling before i could continue “…and presumably by extension, black youth of America, his past self, etc.”

This wasn’t a shot in the dark; i had seen TNC speak a few months prior at UPenn, during which he responded to a question about how to discuss these issues with people resistant to hearing them — namely white people. His response was essentially: ‘I don’t know. I don’t do that, that’s not my job. I try to write the truth as strongly & precisely as I can, and to the extent that I have an audience in mind, it’s other black people.’

And this was further confirmed in the last pages of BTWAM, where he writes to his son:

I do not believe that we can stop them, Samori, because they must ultimately stop themselves.

…do not struggle for the Dreamers. Hope for them. Pray for them, if you are so moved. But do not pin your struggle on their conversion. The Dreamers will have to learn to struggle themselves, to understand that the field for their Dream, the stage where they have painted themselves white, is the deathbed of us all.

(here and thruout the book, the term ‘Dreamers’ refers to those of us lost in the mutual dream of mythical American innocence & our own whiteness)

And so i am trying to struggle myself, in my own personal & specialized domains (of whiteness , of data analysis), in the hope that one day we might stop ourselves.

If the streets shackled my right leg, the schools shackled my left.

— TNC, BTWAM

Ta elaborates on the parallel dangers of schools & streets:

I came to see the streets and the schools as arms of the same beast. One enjoyed the official power of the state while the other enjoyed its implicit sanction. But fear and violence were the weaponry of both. Fail in the streets and the crews would catch you slipping and take your body. Fail in the schools and you would be suspended and sent back to those same streets, where they would take your body.

It often seems to me that academics & ‘The Data Say’ types are a bit too enamored with their own disciplines & skill-sets. Having a fancy complicated hammer makes us want to pull it out and test it on every would-be nail. We’re overly confident of its powers & insufficiently skeptical of its intrinsic biases & limitations (much less our own). We say dumb things & peddle bullshit. Sometimes the stakes are graver than spreadsheets & probability distributions can tell. The abuse of data vanishes in comparison to the abuse of bodies. But we should not let academic methods add to any of society’s shackles. We should especially not let the abuse of data defend & contribute to the abuse of bodies.

Some of us make it out. But the game is played with loaded dice.

— TNC, BTWAM