Race and class

Mike Macnair examines the critiques of intersectionality developed by Adolph Reed Jr and Walter Benn Michaels

The critiques of ‘intersectionality’ published in Science and Society, which I looked at two weeks ago,1 were from the standpoint of ‘Marxist feminism’. Now I will look at the contributions at Nonsite.org. A third article will review Asad Haider’s short book, Mistaken identity.2

Both the Nonsite articles and Haider are focussed on critiquing classless forms of ‘anti-racism’. These authors are considerably more robustly critical of their targets than the Science and Society contributors. But their positive political alternatives remain weak; and this is a matter of the conceptions of class and of (so far as this idea is present at all) of party.

Nonsite’s issue 23 (February 11 2018) is headlined ‘Naturalizing class relations’. The issue carries six articles - four are directly on the politics of ‘anti-racism’ (two brief and two substantial), while the other two are indirectly relevant.3

The two brief articles are Kenneth Warren’s ‘Tain’t so’, an introduction to the issue which cross-refers to previous Nonsite articles, and Willie Legette’s and Adolph Reed Jr’s ‘The role of race in contemporary US politics: VO Key’s enduring insight’. This is a short review of the authors’ 30-year debate “over whether the post-segregation-era black political class in Louisiana or that in South Carolina is the most politically bankrupt, craven and crudely class-centric in the country”.

The two more substantial articles are Adolph Reed Jr’s ‘Black politics after 2016’, which begins with the politics of the race question within the Democratic Party, to develop both the historical background, and arguments against the pro-Wall Street ‘anti-racists’; and Walter Benn Michaels’ ‘The political economy of anti-racism’, which seeks (as he indicates at the end of the article) “to argue against a liberal egalitarianism that, taking discrimination as its central problem, imagines a justified inequality as its solution”.

Neither Reed nor Michaels is a new antagonist of identity politics. Reed’s opposition goes back at least to his book, Class notes: posing as politics and other thoughts on the American scene (2000), and at least partially to his The Jesse Jackson phenomenon (1986)4 and probably further. Michaels wrote in 2007 The trouble with diversity: how we learned to love identity and ignore inequality and already in 2000 contributed an article on ‘Political science fictions’ to the journal New Literary History, whose issue title was ‘Is there life after identity politics?’ Again the political commitments may be older.5 The point is that the politics of 2016-18 have given considerable added force to the critiques.


Reed’s argument is complex - in annotating the article, I identified eight distinct points in it.

1. The starting point is that Reed documents with clarity the linkage made by Clintonite Democrats between, firstly, attacking the Sanders wing as pandering to backward, white male, working class ideas, and secondly, attacking them as supporting old-fashioned, unrealistic socialist ideas.

2. This Clintonite-identitarian line of argument involves “writing off” large numbers of voters: there may have been seven to nine million 2012 Obama voters who voted for Trump in 2016. Reed argues that these voters - “who were not exclusively white - may have good reasons to feel betrayed by both parties ...”

3. The Clintonista vision of the Democratic Party as based on an alliance of women and minorities suppresses largely, but not completely, reference to the other element of the alliance - Wall Street; speaking in other contexts, identitarian Clintonistas can speak explicitly about “Why Democrats need Wall Street”. This schema places Democrats and Republicans in a common commitment to neoliberalism and the financial sector, differentiated by Democrats identifying their party ‘brand’ by “commitment to diversity”.

4. This identity politics is not merely an alternative to a (working) class politics, but “reflects the perspective of a different class, the professional and managerial strata ...”

5. The contemporary US politics of race is framed by “a notion of ‘race relations’ which presumes that blacks and whites ... are basic, indivisible units of political interest”. There was a real basis for this for much of the 20th century in the form of “legally enforced racial segregation”; in reality, however, “black political differentiation was significant even in the Jim Crow era” and, conversely, black people were also engaged in the 1880s Knights of Labor, the populist movement of the 1890s-1900s, and so on. Down to the early 1960s black political action “focussed on concrete issues”, and it was only after the 1960s legislative victories of the civil rights movement, in the late 1960s-1970s, that “‘racism’ was advanced as the default explanation for inequalities that appear as racial disparities. That view emerged from Black Power politics and its commitment to a race-first communitarian [meaning communalist] ideology ...”

6. The “mundane context” of the shift into communalism was a debate among the Democrats in the early 1960s over how to tackle poverty - with one camp, connected to the labour movement, arguing for social democratic measures, while the other, connected to the Ford foundation and stronger in the Kennedy and Johnson administrations, saw poverty as the result of the family pathologies inherited from the past: the view most explicitly defended in Daniel Moynihan’s 1965 Department of Labor report The negro family. The social democratic trend was defeated, essentially on the basis that its proposals could not be afforded alongside tax cuts as a Keynesian stimulus and Vietnam war spending.

What remained was Moynihan-style initiatives purportedly directed against the ‘pathologies’ of the poor - a policy which has persisted to the present day.6 Talk of ‘institutional racism’ was popularised after Stokeley Carmichael’s 1967 Black power and functioned precisely to abstract questions of race from the broader institutional issues requiring a redistributive politics. The end product is Obama’s Moynihanesque 2008 talk of ‘tough love’ against the imagined black underclass.

7. With this background, the race-baiting of the Sanders movement is

at bottom a class politics ... Only ideological blinders can block out the implication that a fair share of acclaim for Ava DuVernay, Nate Parker or Rihanna is, or should be, more important to black Americans than general access to decent, secure employment and retirement, healthcare and a vibrant public sector.

The arguments offered of a “new Jim Crow” falsify what was actually won in the 1960s; claims centred on “disparity [between races] as the lone, truly actionable injustice” are “not a call to popular political action, but a demand for recognition based on moral priority”.

The material foundation of this ideology is the black professional-managerial strata (mentioned earlier), which have grown exponentially under equalities and diversity policies, and the ‘community’ representatives of localities qua black or whatever. In a particular example, opposition to ‘gentrification’ misses the fundamental point - that the state is expropriating the poor to subsidise improved rent extraction by developers - in favour of a discourse of cultural authenticity, which can be satisfied by token gestures and in which “benefits to individuals can appear to be victories for the generic racial population ...”

8. Finally,

A year into the Trump presidency and unimpeded Republican control of Congress and of most state governments has confirmed what many on the left have known all along: that the right’s agenda is an all-out attack on working people, no matter what their racial and gender classifications and identities or sexual orientations. The alliance of Democratic neoliberalism and an identity-based notion of social justice has contributed to this nightmarish outcome, precisely by diminishing the significance of a policy orientation that abets upward redistribution and intensifying economic inequality and racialising the working class as white losers. Doubling down on that approach, as Clintonites and race- and gender-reductionists exhort, will not effectively counter the right’s strategy.

The alternative Reed offers is a politics based round “how government can best provide for the security and welfare of the vast majority of the population who must work for a living”. And:

This means, among other things, that we must be confident in rejecting claims - no matter how flamboyantly adorned they are with moral posturing and evocations of past insurgencies - that black Americans or other non-whites, immigrants, women, LGBTQ and transgender people somehow will spurn access to quality, affordable housing, healthcare, education, secure employment with decent wages, benefits and rights and protections on the job, and high-quality public goods and services as not pertinent to them.


Walter Benn Michaels’ argument is a good deal simpler. It starts with the point that the family income of students at elite universities is massively above US median family income. Michaels recognises that, as academics, Reed and he are both well-paid, and engaged in trying to talk to “rich kids with short attention spans”. There is a contradiction between these characteristics and their socialism. But, Michaels argues, there is no such contradiction between wealth and ‘identitarianism’: “What’s wrong with the identitarian version of the left is not that its roots are in money, but that its identitarianism is a defence of that money.”

The argument is that discussions of equality address two kinds of inequalities: “individual inequality” and “horizontal inequality” - aka “inequalities between culturally formed groups”. Democrat economist Paul Krugman, for example, demanded during the 2016 election that Democrats should “talk more about horizontal inequality”.

Michaels looks at the distribution of household income in the US in 2015 by quintiles. For example, 6% of US households earn more than $200,000 a year, but only 2% of blacks; 45% earn less than $50,000 a year, but 61% of blacks. If race discrimination was got rid of, since black people are around 13.2% of the US population, they would form 13.2% of the very rich, and so on: and 6% of black people would be in the top quintile, 45% in the bottom quintile. The point is thus made graphically that “the point of eliminating horizontal inequality is to justify individual inequality”. And:

This is why some of us have been arguing that identity politics is not an alternative to class politics, but a form of it: it’s the politics of an upper class that has no problem with seeing people left behind, as long as they haven’t been left behind because of their race or sex.

Equality-of-opportunity anti-discrimination discourse blames the victims of economic inequalities, unless they can claim that their victimhood is the result of race or other discrimination. It is a natural consequence that there should be a growth of “white identity politics” among poorer white Americans.

Moreover, introducing class as another element of the ‘intersection’ will not help:

the demand for a working class politics cannot be met by adding class to race and gender. This is in part because most actually existing intersectionalists are not very interested in class, but, even if they were, the great attraction of intersectionality is precisely that it can absorb class too into the logic of discrimination - which is to say, into the neoliberal theory of inequality. On this theory, the fact of being born poor is a problem in the way that the fact of being born black or Latino or a woman can be - if it keeps you from having an equal opportunity to succeed. In other words, we should treat class difference like race and gender by understanding poverty: not as something we should get rid of, but as something we should help people overcome.

But universities which had proportional representation of the poor would mean that “almost all the students they now have would be sent home and almost none of the students they would have instead would be able to pay the tuition”.

Even supposing universal free higher education, with resources magically raising all universities to the standard of Harvard, the US job market is moving in the opposite direction: the fastest growing jobs are for personal care aides, fast food service and related jobs, and so on:

right now the majority of personal care aides (number 1 on our list) are women of colour with no more than a high school education. In the neoliberal, intersectional, everybody-goes-to-Harvard utopia we’ve been imagining, a lot of them would instead be white men with political science or marketing degrees. But they’d still be making $21,920 a year.

“Insofar as economic inequality is the problem”, Michaels argues, “redistribution, not proportional representation, is the solution.” He concludes that “without a politics organised around class - fighting not for the interests of women of colour, but for the interests of personal care aides - the left, when it comes to economic inequality, is just the good conscience of the right”.

These two essays are mostly negative critique, rather than positive alternatives. However, as negative critique, they certainly hit home. Let us imagine for the sake of argument that in this country the Eurocommunists and Blairites had accomplished their mission to liquidate the Labour Party and turn it into a left-of-centre party unified with the Liberal Democrats (perhaps a ‘Democratic Party’, as the Italian Communist Party became ...?). The capitalist class interests represented by ‘intersectionality’ (as it is even now defended and exploited by the Labour right) would then be transparent, as they are in the politics of the US Democratic Party. These class interests are, indeed, as Reed and Michaels argue, those of the managerial and bureaucratic agents for the capitalist class - and lying behind them is the interest of finance capital in covering for upwards redistribution and putting any sort of downwards redistribution ‘off the table’. The fact that the historicalbackground of ‘intersectionality’ and identity-political forms of ‘anti-racism’ can be found in the 1960s-70s left is quite immaterial to their present political meaning.


The positive alternatives are less helpful. In the first place, ‘class’ means for these authors merely income inequality. And the programmatic alternative proposed is, in essence, purely addressed to ameliorating the consequences of income inequality: in Reed’s slightly more concrete version, “quality, affordable housing, healthcare, education, secure employment with decent wages, benefits and rights and protections on the job, and high-quality public goods and services”. This is, in substance, a European social democratic programme.

Having regard to where we are, it would be none the worse for that - if it were not for the fact that the capitalist class - globally and not just in the US - has been insisting on the rejection of such programmes. If capitalists went on the streets to demonstrate, their slogan would be: ‘No return to the 70s! Lefts out now!’

The Clintonista opponents of the Sanders movement, and the Blairite opponents of the Corbynites, are precisely channelling such slogans: thus, for example, “Chris Leslie, Labour’s former shadow chancellor, [on June 18 published] Centre Ground with the Social Market Foundation think tank, aimed at appealing to those who ‘work hard and contribute’, while accepting that ‘there are not limitless public resources’.”7

Why should this be the case? The fundamental point that Reed and Michaels make is that the elites are economic exploiters, who become better-off through the large working class being made worse-off. But what is missing in this story is that, in order to be successful exploiters, the elites have to hold the working class effectively in subordination.

The ‘golden’ 1950s-70s had that character because there was a system of concessions by US capital - both to the US, European and colon-society working classes; and to ‘national capitals’ and nationalists. These concessions were there in order to make ‘containment’ of communism (both in the form of Soviet tanks on the Elbe and the Chinese People’s Liberation Army, and in the form of mass communist parties elsewhere) work politically: to legitimise massive military expenditures, and to persuade European, and so on, workers that ‘social-democratised’ capitalism was preferable to Stalinism.

But they had a major disadvantage to capital: the working class, with increased security, began to demand more than just a bigger slice of the cake (and the bigger slice demanded was itself uncomfortable to capital, as in the 1960s defeat of the US social democratic trend, which Reed’s article records). When the US was defeated in Vietnam and the colonial liberation struggles also triggered the Portuguese revolution of 1974-76, the US turned to a new policy - of which official anti-discrimination and what became ‘intersectionality’ were a part, alongside the positive assertion of the merits of free markets and uselessness of government interventions, ‘human rights’, and US support for Islamism against secularists and for guerrilla movements against nationalist and pro-Soviet regimes.

This policy is today beginning to come to an end, and we are seeing a gradual global turn to rightwing, populist nationalism, which will probably end in a new round of great-power war some time in the next few decades.

The left version of ‘intersectionalism’ maintains the illusion that capital cannot be anti-discriminatory, as if nothing had happened since 1976. Reed’s and Michaels’s polemics are powerful against this view.8 But the converse is that it is not true that capital is inherently anti-discriminatory, contrary to Gary Becker.9 Capitalists need to hold the lower classes in subordination in order to procure their own individual liberty.

For a very simple example, Steve Jobs’ aesthetic design creativity depended on radical subordination of management below him - and beyond them, of the workers at Foxconn and so on. The methods of holding those below in subordination will vary, but will almost always include some form of the old maxim diuide et impera: divide and rule. Capitalists can divide and rule by anti-discrimination, as they have for the last 40 years - or by discrimination, as they did before then.

Capitalists can rule through the party of liberty (currently the Democrats, or the 19th century British Liberals) - or the party of order (currently the Republicans, or the British Tories). Both trends are natural ideologies of capitalism.

The party of liberty ideologises the market sphere, of which Marx said: “This sphere that we are deserting, within whose boundaries the sale and purchase of labour-power goes on, is in fact a very Eden of the innate rights of man. There alone rule Freedom, Equality, Property and Bentham.”10

The party of order ideologises the other side of the coin: the despotism of the factory, and the aspect of production where

The contract by which he sold to the capitalist his labour-power proved, so to say, in black and white that he disposed of himself freely. The bargain concluded, it is discovered that he was no “free agent,” that the time for which he is free to sell his labour-power is the time for which he is forced to sell it, that in fact the vampire will not lose its hold on him, “so long as there is a muscle, a nerve, a drop of blood to be exploited”.11

As long as the capitalist class controls the levers of political power, it can shift ground between the party of order and the party of liberty, as the political ‘purchase’ of one or the other runs out of steam; and it can make economic redistribution concessions, like those Reed’s proposed platform advocates - and withdraw them when the time is right, as it has done gradually over the last 40 years.

The capitalist class controls the levers of political power through the institutional forms of politics - in the US, the constitution, the free market in legal services, the presidency as symbol of one-man-management, ‘first past the post’ and the two-party system of corrupt politicians (Reed’s account presents us with one element of this ‘political class’: the black ‘political entrepreneurs’), and so on.

The consequence is that a purely economic programme, of the sort Reed advocates as a possible basis of unity of the working class, runs up against the idea of ‘No return to the 70s’ shared by major capitalists, Democratic Leadership Council (DLC) types, and ex-Eurocommunists - but, on the other hand, offers no tendency to undermine these groups’ political power.

The choice to offer a purely economic programme is not a stupid one. It reflects an aspiration to build a broad class movement, by focussing on the issues which unite us and avoiding those which divide us. It is in this sense a form of political ‘realism’.

It is the same as the political ‘realism’ of Jeremy Corbyn, John McDonnell and their advisors in this country, who similarly aim to focus on the economic issues and downplay others - and, before them, of the political ‘realism’ of a series of left projects, which claimed Labour had become a purely capitalist party and the remedy was to create a new left, social democratic party: the Socialist Labour Party, Socialist Alliance, Trade Unionist and Socialist Coalition ...12 The idea has a bit more plausibility in America, because the US does not already have a Labour Party.


This relates, in turn, to the party question. This is unmentioned in these articles - except insofar as it concerns the factional struggles between left and right in the Democratic Party. And Reed’s narrative displays a certain nostalgia for the ‘social democratic’ trend in the 1950s-60s Democratic Party (and for ex-communist and civil rights leader, later social democrat cold warrior, Bayard Rustin). In The Jesse Jackson phenomenon Reed argued explicitly, on grounds of political realism, against any attempt to construct an independent third party (pp131-33).

Reed was, however, a leading participant in the attempt in the 1990s-early 2000s to build a US Labor Party based on the trade unions, led by Tony Mazzochi of the Oil, Chemical and Atomic Workers Union. Reed retains this experience as a reference point, as is shown by his short Nonsite piece, ‘Mazzochi and the moment’ (November 21 2016). The balance sheet Reed, Mark Dudzic (former Labor Party national organiser) and Katherine Isaac (former Labor Party secretary-treasurer) draw of this failure is that its fundamental conceptions were right, but premature.13

Though the 1996-version Labor Party obtained support from unions representing substantial numbers of workers and had over 1,400 delegates at its 1996 founding convention, and between 15,000 and 20,000 members at its peak,14 it lost momentum - especially after the terrorist attacks of September 11 2001 sharply revived the political salience of issues of imperialism, war and hence - albeit in the ‘new’ form of discrimination against people who appear to have ancestral backgrounds in ‘Muslim’ countries - race.

The ‘1996 Labor Party’ strategy was a project for working class political representation, on the basis of a limited economic programme which could unite the class, with the trade unions at its core. This project could, in fact, only make sense as an electoral project - and that on the basis of willingness if necessary to be ‘spoilers’, who caused the Democrats to lose, in order to force them to step back from DLC ‘triangulation’ towards their right. This logic was already known not merely by the ‘Marx party’ of the later 19th century, but also by the non-Marxist element among the founders of British working class electoral representation.

It is thus an apparent paradox of the positive claims made for the strategy that the 1996 Labor Party was unwilling to stand candidates unless it was already certain that they could get ‘serious votes’. Reed, and Dudzic and Isaac, gives two parallel reasons for this course of action. One, which was certainly ‘operative’, is, as Dudzic and Isaac put it, that “it is not realistic to demand that today’s labor movement completely disengage itself from its current ties with the Democratic Party”.15 The other, which seems more ‘ideological’, is a ‘new left’ idea: that the basis for a movement which can in future stand in elections will be built by non-electoral grassroots campaigning and ‘coalition-building’.

In fact, the strategic conception involves a fairly simple misunderstanding of the origins of the British Labour Party. The foundation of this party was not a pure action of the trade union leaderships. Rather, it involved a substantial period of prior electoral campaigning by small groups, the Marxisant Social Democratic Federation (SDF) and the Christian-socialist Independent Labour Party (ILP). Functioning chiefly as ‘spoilers’ in parliamentary elections, by the late 1890s the groups had begun to make serious gains standing locally. They were initially involved in founding the Labour Party as organised affiliated parties, though the SDF walked out. Moreover, these developments took place against the background of the growth to mass influence of the German Social Democratic Party - a party founded in 1875 by the fusion of two smallish groups on a platform for the overthrow of capitalism and its constitution.16

The British trade union leaders were just as much enmeshed with the Liberal Party as today’s US trade unions are enmeshed with the Democrats - and they would have been happy to pursue the US 1996 Labor Party policy of “no unrealistic candidates” and “non-electoral campaigning” if it were not the case that imposing this policy would have resulted merely in the ILP as well as the SDF bolting the Labour Party in order to continue with their moderately successful small-scale electoral work.

In fact, in the USA itself, the most successful workers’ organisations were the Socialist Party of America - founded in 1901 as a fusion of smaller groups and in political solidarity with the Second International17 - and the Communist Party USA, founded in the split in the SP in the aftermath of 1917, which, though it never reached the heights of the SP’s electoral success, was profoundly influential down to and including the 1960s civil rights movement.

How does this problem of party relate to the intersectionality/race and class issue? The question takes us back to what the political platform of any proposed party alternative to intersectionality is.

Reed’s wish-list of demands - remember, “affordable housing, healthcare, education, secure employment with decent wages, benefits and rights and protections on the job, and high-quality public goods and services” - is more economistic than the 1996 Labor Party USA platform, which included headings “4. End bigotry: an injury to one is an injury to all” with the usual demands, and “13. End corporate domination of elections”.18 Let alone that it is more economistic, and more reformist, than the 1901 Socialist Party of America platform, the 1875 Gotha Programme, or the 1880 Programme of the Parti Ouvrier.19

Remember that the Clintonite race-baiting of the Sanders movement exploited Black Lives Matter. Reed has argued forcefully that race is not particularly explanatory of police violence.20 Take it for the sake of argument that this is true. It is, however, not politically realistic to expect people to restrict agitation to economic issues for the sake of unity, when cops are being filmed killing unarmed suspects and are then acquitted in the resulting trials. In contrast, in this country in 1984-85 Lesbians and Gays Support the Miners precisely built solidarity round common experiences of police oppression.

The Socialist Party in 1901 could without hesitation demand gender equality in its platform, and could have in that year a serious debate about race. It could also denounce US policy in Puerto Rico, and “call the attention of the working class to the fact that our judiciary is but a servile tool in the hands of the capitalist class and hostile to the interests of labor”. A party with a political programme and character like the old Socialist Party’s could represent an alternative to intersectionality based on class unity. A party which commits itself in advance to a limitation to economic demands for the sake of unity, and to what is acceptable to the trade union bureaucracy, disables itself from offering such an alternative.

It will thus inevitably be derailed - as the 1996 Labor Party project was by 9/11.



1. ‘Intersectionality is a dead end’, June 6.

2. London 2018.

3. At https://nonsite.org/ issue #23. The two indirectly relevant articles are J Sowa, ‘Populism or capitalist demodernisation at the semi-periphery: the case of Poland’ - an extremely illuminating article on the development of Polish right-populism as beginning already in 1990, and about Trump and Brexit as illustrating, in a sense, a ‘Polonisation’ of western politics; and T Cronan and C Palermo, ‘More neoliberal art history’ - a savage critique of the inability of Foucaultian ‘history’ to address the presence of class in pre-neoliberal politics.

4. Yale 1986. Chapter 7 of this book has an interesting discussion of the deployment of ‘anti-Semitism’ allegations against Jackson in the interest of the Democratic right - and Jackson’s willingness to play up for these for a sort of intersectional street cred.

5. In the 2014 interview of Michaels by JJ Williams (‘The political education of Walter Benn Michaels: an interview’ symploke Vol 22, 2014, pp337-59), Michaels presents himself as having broadly naive left political ideas between 1968 and his 2001 job move to the University of Illinois at Chicago.

6. It should be said that these approaches are merely the modern incarnation of the early modern English differentiation between the ‘deserving’ and ‘undeserving’ poor, and the various training schemes which were supposed to fit the poor for employment.

7. The Times June 18 2018.

8. See also Reed’s ‘Splendours and miseries of the anti-racist “left”’, Nonsite.org, November 6 2016.

9. GS Becker The economics of discrimination Chicago 1957.

10. www.marxists.org/archive/marx/works/1867-c1/ch06.htm.

12. www.marxists.org/archive/marx/works/1867-c1/ch10.htm.

12. All these groups were willing to take some formal positions on constitutional issues, but not to actually campaign round them. Respect and Left Unity were for different reasons partial exceptions to this particular catalogue of failures, by being willing at least to address British foreign policy. On the other hand, both in different ways internalised left-Clintonite intersectionalism.

13. Dudzic, interviewed in New Labor Forum Vol 23, pp60-64 (2014); P Reed ibid pp65-67; Dudzic and Isaacs: www.thelaborparty.org/d_lp_time.htm.

14. www.jacobinmag.com/2015/10/tony-mazzochi-mark-dudzic-us-labor-party-wto-nafta-globalization-democrats-union.

15. www.thelaborparty.org/d_lp_time.htm.

16. See the text at https://history.hanover.edu/texts/gotha.html.

17. The 1901 platform, which has the same general character as other Second International platforms, is at www.marxists.org/history/usa/parties/spusa/1901/0817-herald-socialistconv.pdf.

18. www.thelaborparty.org/d_program.htm.

19. www.marxists.org/archive/marx/works/1880/05/parti-ouvrier.htm.

20. https://nonsite.org/editorial/how-racial-disparity-does-not-help-make-sense-of-patterns-of-police-violence.