Google+ Followers

Friday, August 7, 2015

Overstating "White Privilege"

The phrase "white privilege" has become a standard buzzword among social justice warriors in the United States and creeps from time to time into more mainstream fora, such as the (poorly argued) debate between Bill O'Reilly and Jon Stewart on The Daily Show last October. 

In the New York Times a couple weeks ago, I read a Michiko Kakutani review of a new book by Ta-Nehisi Coates, a writer at The Atlantic whose column I used to read occasionally.   The review summarizes Coates' book as "a searing meditation [sic (I have no idea how a meditation can  "sear" anything)] on what it means to be black in America today. It takes the form of a letter from Mr. Coates to his 14-year-old son, Samori, and speaks of the perils of living in a country where unarmed black men and boys ...  are dying at the hands of police officers, an America where just last month nine black worshipers were shot and killed in a Charleston, S.C., church by a young white man with apparent links to white supremacist groups online."

Other MSM participants and, of course, the left wing echo chamber have dwelt on Coates' book at length too.  Like Kakutani's review, David Brooks of the Times calls it a "searing contribution to ... education for white people" before excerpting several hateful passages (example: "‘White America’ is a syndicate arrayed to protect its exclusive power to dominate and control our bodies." Or the passage in which he calls the first responders who died in the 9/11 attacks "menaces of nature")(Parenthetically, I can actually understand someone black feeling that way in the flash of a given moment, but I can't understand anyone retaining that feeling and publishing it after any sort of minimal reflection on the facts of 9/11). 

Kakutani's review excerpts a brief passage in which Coates makes  typical assertions about the benefit of being white in America, which even the review - in the Times! - characterizes as "cliched".  Notwithstanding that judgment, I want to go beyond epithets into a deeper analysis of the illogic and inaccuracies in assertions like Coates' about "white privilege", as I did earlier this year in reaction to a fallacious editorial by Nicholas Kristof about modern bias experiments.  Here is the excerpt from Kakutani's review:

"Mr. Coates contrasts [the] world of the streets with the 'other world' of suburbia, 'organized around pot roasts, blueberry pies, fireworks, ice cream sundaes, immaculate bathrooms, and small toy trucks that were loosed in wooded backyards with streams and glens.' He associates this clich├ęd suburban idyll with ... an exclusionary white dream rooted in a history of subjugation and privilege.  [White people] he contends, 'have forgotten the scale of theft that enriched them in slavery; the terror that allowed them, for a century, to pilfer the vote; the segregationist policy that gave them their suburbs.'"

It's obvious to any reader that white people in suburbs have no monopoly on pot roasts, blueberry pies, fireworks, ice cream sundaes, clean bathrooms or small toy trucks.  So, obviously overstated, but that is not my paramount concern, since any reader can make that judgment for herself or himself.  More significant are the examples of "subjugation and privilege": slavery; vote deprivation; and segregation.  What I want to unpack here is the confusion between "subjugation" and "privilege", which the author, and evidently the reviewer, seem to think of as two sides of the same coin.

I think that is fallacious; that is, probably because discussion of racial bias tends to proceed in terms of  dualities:  white/black; slave/free, etc., which, unquantified, sound equal, reciprocal, zero-sum, the frame obscures a simple quantitative fact, that the percentage of blacks in America at relevant times to the discussion has been consistently 10 - 11%, so small that blacks could indeed be "subjugated" by what Coates identifies, without that subjugation playing a substantial role in causing the status of very many white people today.  Put another way, as I shall illustrate below, there have been so many more white people in America than black at all relevant times, that the current economic status of white people is largely independent of the causes of the current economic status of black people.  Therefore, the concept of "white privilege" is grossly overstated; it would be more accurate to speak about "disparate impact" of policies on black people than to falsely inflate the benefit of "white privilege".  

An easy demonstration of this is the "segregationist policies that gave [white people] the suburbs" according to Coates.  In the period of most rapid suburban expansion, 1950 - 1970, there were three censuses and, in each of them, whites never made up less than 88% of the population:

Year
White
Black
Ratio
1970
177.7
22.6
7.9:1
1960
158.4
18.9
8.4:1
1950
134.9
15
9:1
Net Population Growth
31.8%
50.7%
1.6:1
source:  Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970 (Bicentennial Edition, Part I), Series A73-81; all numbers in millions and rounded to first decimal

Before I delve more deeply into the demographic profile of suburban presence in this period, I will offer a simple example of what these overall demographics mean for "white privilege".
Imagine that every neighborhood in the nation at these points in time had been a microcosm of the national demographics, perfectly mirroring the nation's demographics, with no racial disparity whatsoever.  If, for example, there was a suburban development built in 1950 with 50 homes, the ownership of those homes would have been allocated 45 to whites and 5 to blacks.  By 1970, ownership would have shifted 44:6 (7.9/8.9 * 50). 

If we contrast that with the worst-case segregation scenario, with all the suburbs being populated in the year of highest black population and yet complete exclusion of the black minority, we see that 6 blacks would have been excluded, and 6 whites would have benefited disproportionately.  But, 88% of whites would have been there anyway

This is the simple point that needs to be understood about "white privilege".  The vast majority of whites would have been in suburbia under any scenario from the worst case to the most utopian, enjoying all the benefits of the "cliched suburban idyll" Coates depicts.  They would have seen the same home appreciation; paid the same mortgage and property taxes; the vast majority of their kids would have gone to the same public school funded by those taxes, gotten the same grades, and gone to the same colleges; they would have shopped in the same places, joined the same country club, had much the same social interactions, etc. Whatever specific example of the "white privilege" argument one wants to offer, the odds are very high that, as to any given white person, the identified benefit has not been received at the expense of a black person.  Some white people will have benefited disproportionately but, by definition, even in the worst case I've just posited, they cannot amount to more that the black minority's on-average 11% share of the total population -- and  when one  delves into the demographics of residence during the period of suburban expansion more closely, it's really even smaller than that, more like 3% of white people in the relevant period, 

When you get into the demographics of suburban expansion, you may be surprised to find that  the black population in every category classifiable as suburban grew slightly faster than the similarly classifiable white population in this time;

Year
White Urban Fringe
Black Urban Fringe
White Rural Nonfarm
Black Rural Nonfarm
White Other Urban
Black Other Urban
1970
51.4
2.54
41.3
3.8
27.8
 2.7
1950
19.9
0.95
28.5
2.5
24.8
 2.3
Growth
158%
167%
44%
52%
12%
17%
source:   same document, Series A73-90; all numbers in millions

However, given that the total black population grew 60% faster than the total white population in those decades, as the first table showed, these slight differences cover less than half of the black population growth (3.3 million out of 7.6 million) in those decades.  The real change in black population location was that black rural families moved disproportionately into the central cities, as documented in, among other things, "The Warmth of Other Suns", a 2010 history of that migration by Isabel Wilkerson that won the Pulitzer Prize (such a rural - > urban migration is, by the way, not an uncommon demographic movement in nations worldwide, regardless of race; it happened in the Industrial Revolution in England; it has been the source of urbanization in China and India, for example; it goes on today in Brazil and it is a substantial part of the Hispanic migration into the US as well).

Year
White
Rural Farm
Black
Rural Farm
White Central Cities
Black Central Cities
1970
  7.8
0.45
49.5
13.1
1950
19.7
3.16
42
  6.1
Growth
- 60%
 - 86%
18%
115%
source: same as above

If I recompute the suburban population of 1970 to mirror the total national population of that year, the racial distribution would look like this (all numbers in millions):

White non-farm, non-central-city (unadjusted)
White non-farm, non-central-city (adjusted)
Change
Black non-farm, non-central-city (unadjusted)
Black non-farm, non-central-city (adjusted)
Change
120.1
114.6
  -5.5
9.0
14.5
5.5

So, roughly 5.5 million whites would have been replaced in the suburbs (loosely defined  --I know that this national level of analysis is an over-generalization but it is the level Coates argues at, so I think it is where the response should be couched) by blacks, out of a total white suburban population of 120.1 million, or roughly 5%.   That same small shift in the white population would, however, have increased the black suburban population by roughly 60%.  Measured against the total population of each race in 1970, if you want to measure the total impact on each race, the numbers are 3.1% and 24%.  So that was -- roughly speaking, given the 50,000 foot level of national census data -- the racial disparity in housing:  96.9% of whites and 76% of blacks would have landed in the same categories in a racially unaltered allocation of housing. The point, I hope, is well taken: there is virtually no "white privilege" in housing (or much of anything else) because whites are such a large portion of the population, and thus they are inherently a large portion of all of the outcomes of the population.  It's in fact a ludicrously inaccurate form of stereotyping, as much a stereotype as claiming that all black men are criminals.  Yet, even if white privilege is in fact negligible, the black population can experience a disparate negative impact from exclusion.  For this reason, it would be more intelligent to speak in terms of "disparate impact" on blacks than in terms of the insubstantial "white privilege".

The same proposition holds true when one looks at income and affluence, instead of housing dispersion.  It is of course the case that, on all statistical fronts, black income is lower than white income.  For example, the Census Bureau calculates the following median incomes for the several main demographic categories in the US in 2012 as follows (h/t Business Insider):

"Among the race groups, Asian households had the highest median income in 2012 ($68,636). The median income for non-Hispanic White households was $57,009, and it was $33,321 for Black households. For Hispanic households the median income was $39,005,"

It seems clear that the same principle holds true, that, owing to the much larger proportion of white earners in the earning population, substantially all white people would be in approximately the same position, if black earners were distributed in perfect proportions throughout the overall income distribution instead of being over-represented in the lower half of that distribution and under-represented in the upper half of it.  For example, I selected this quote from the Wikipedia page "Affluence in the United States" (which, at least as of this date, has a very progressive perspective):

"in 2005, 81.8% of all 114 million households were White (including White Hispanics), 12.2% were African American, 10.9% were Hispanic and 3.7% were Asian American. While White households are always near the national median due to Whites being the by far most prevalent racial demographic, the percentages of minority households with incomes exceeding $100,000 strayed considerably from their percentage of the overall population. Asian Americans, who represent the smallest surveyed racial demographic in the overall population, were found to be the prevalent minority among six figure income households. Among the nearly twenty million households with six figure incomes, 86.9% were White, 5.9% were Asian American, 5.6% were Hispanic; and 5.5% were African American.  Among the general individual population with earnings, 82.1% were White, 12.7% were Hispanic, 11.0% were African American and 4.6% were Asian American."

Note that the percentages add up to > 100% because of double counting of "white hispanics".  Still the order of magnitude of the white/black ratio is such that the double counting doesn't meaningfully alter the main point.  Even if all "white hispanics" are subtracted from the "white" population, the excess percentage of white earners "among the nearly twenty million households with six figure incomes" is 17% ((.869 - .056) / (.821 - .127)), meaning that 5 of every 6 whites in that top earning bracket would be there if its constituents were a perfect mirror of the demographics of all earners.  That would be the most one could say about "white privilege" in the modern income distribution (although one might, I think, fairly question, in an analysis of an asserted birth-based privilege within the polity of the United States, whether new immigrants, who (a) were born outside the US and (b) came here voluntarily, should be weighted; that they dwell in a white majority polity and compete against a white majority workforce and have to sell to majority white consumers are not accidents of birth, but the results of their conscious choices. (In fact, one might fairly call it a "privilege" for a person to be able to live and work in a wealthier polity than the one he or she was born into). Having been born outside the polity, they are not on an apples-to-apples basis with the vast majority of the population set being scrutinized; were they to be subtracted, the white proportion of the population set would skew higher and the percentage of "excess" white membership in the top earning bracket  would shrink materially). 

A subsidiary point I want to draw out circles back to my comment at the outset that the Jon Stewart / Bill O'Reilly "debate" over "white privilege" was poorly argued.   As I think I've shown, in a nation that is mostly white, most of the successful people will likely be white, even in a perfectly racially distributed set of outcomes.  So any white person, if challenged to defend his success against contentions of white privilege, can quite rationally and legitimately respond by saying, "no, I am confident I would have been in this position even in a world of perfectly racially distributed outcomes."  Obviously some small number of white people would be wrong in making that statement, but they would be a small minority.  So that is one relevant point O'Reilly should logically have said in response. 

But there is actually a stronger point that the most successful people like O'Reilly and Stewart can make. When a white person,  like each of those men, succeeds in generating a very high income, he has outcompeted, not just the people of color in his demographic cohort, but the whites in it as well.  If he's in the top 99%, or 99.9%, he's outperformed at least 98% of the other holders of "white privilege".  So mere "whiteness" can't explain very much of his income, or median white income would be up at his level.  A "white privilege" attack on the most successful seems to me to be remarkably sophomoric and truly superficial, even though they are the most visible targets of such attacks.  .

Rather, where the effects of "white privilege" would be highest would be the lower one goes in the income distribution. Think of this in terms of displacement.  In the very top percentile, let's say, for argument's sake, 1/6 of the whites are there by virtue of "white privilege".  If you extract them from the top percentile, they have still outcompeted all the whites below them.  They don't drop to the bottom, they just drop to the next level of measurement, the 98th percentile; meanwhile, the people of color from the 98th percentile and lower would hypothetically move up to the 99th to replace them.  But this process repeats at each percentile, and when it does, the proportion of whites that have to be moved down grows.  When you examine the 98th percentile, because you've just moved whites in from the 99th percentile, while moving people of color out, it has become even more lopsidedly white, and therefore, the "1/6 of all whites" in that cohort is an even larger raw number than in the 99th percentile, so to make it perfectly racially representative, even more whites within it have to move down to the 97th percentile, from which even more people of color get moved up again, and the process snowballs all the way down the income rankings. 

What this confirms is that, at the very top of the income distribution, the whiteness of anyone in that cohort is least meaningful in terms of its contribution to his or her income, compared to other positions in the income distribution.  Which is really why I thought the O'Reilly / Stewart "debate" missed the point and was poorly argued. 

But, conversely, the lower one goes in the income rankings, the more likely it becomes that at least some of the disparity between a white person's income and that of blacks below them in the rankings can be attributed to the status of being white, because that white person has outperformed less white people than those above him in the rankings.  
 
Maybe this explains in part why "limousine liberals" and other white elites have been historically more progressive on race, while opposition has tended to be strongest at the working class level, that intuitively, the former perceive - I would argue largely correctly - race to have been essentially irrelevant to their success and therefore they are not really sacrificing any of their status to support minority progress, while the latter may intuitively sense their status is more at risk.  Certainly the numbers bear this out.

Coates's passage itemizes other wrongs as well - slavery and voting discrimination in particular. I am not going to write a tome on racial issues and civil rights, but I would analyze them with the same theme, that the black population may suffer a disparate impact from any given policy, but the outcomes of the white population may not be meaningfully affected by that policy, and "white privilege" is a terribly inaccurate frame for the debate.  This is pretty obvious regarding voting, where a 10-11% share of the electorate is not going to, in and of itself, win any elections, especially were it to be dispersed proportionally among all neighborhoods (in fact, I wonder whether the Obama Administration's recently announced plan to push for more racially distributed housing patterns might ironically wind up diluting black representation in legislatures over a few decades).  Claims of black vote suppression in the 21st century strike me as a strategic overstatement made for political purposes; the Democratic party has a huge incentive to maximize black voter presence and decades of electoral results indicate they have been by and large successful, so I doubt there remain significant impediments to black votes mattering.  In the 2 most recent Presidential elections, black turnout percentage has surpassed white turnout percentage, which doesn't plausibly reconcile with a claim of black voter suppression; maybe the candidate matters at the margin (Gore, Kerry, ...  Obama).

Slavery has been covered extensively and I have little to add: all the slaves and slave owners and children thereof and 99%+ of the grandchildren are dead; the vast majority of the white population did not own slaves and were so poor they derived negligible benefit from slavery; there were white slaves and indentured servants throughout the 17th century and substantial albeit non-slave-level discrimination against certain white ethnic groups in the 18th and 19th centuries (in my mother's hometown, when her Irish ancestors arrived, the town experienced "German flight"); slavery was endemic in Caribbean nations, yet black immigrants from those nations often outperform African-Americans in the modern US economy; the Union Army that freed the slaves at staggering human cost to themselves and their families was almost entirely white; the vast majority of the current population had no ancestors who profited, even indirectly, from American slavery, and even among those people who do have such a connection, it represents a small fraction of their ancestry and personal heritage, so it's myopic and distorting to focus on that to the exclusion of the rest of their identity.  

Most of the wealth created in the era of slavery, directly or indirectly, has been destroyed and dissipated, not just by the Civil War but by the several financial crises and panics since then, as well as the creative destruction of capitalism at work, not just here but abroad.  Most of the wealth existing today in the US comes from services provided, goods manufactured and inventions invented in the post-Jim-Crow era, sure, maybe you can find in some institution, like a bank or university, that has been around for two hundred plus years, some record of making money from the slave trade hundreds of years ago, but that has to be measured against all the money it made from other sources in its history to gauge the impact of slavery on the status of that institution today. And a truly full accounting of value transfers between whites and blacks would need to add in the billions of dollars of transfers, through progressive taxation and welfare state mechanisms like Medicaid and food stamps, from the predominantly white upper class to the lowest classes, in which blacks are disproportionately represented, and the value of the urban infrastructure that "white flight" left behind when the suburban exodus occurred -- the water pipes and reservoirs, electric power plants and wires, sewers, paved roads, and so on -- that the new black arrivals did not have to create from scratch.

I can imagine a number of rejoinders to this argument.  Most of them would, I think, from having read hundreds of arguments along these lines, be totally ad hominem and therefore of no intellectual merit; the rejoinders to Brooks's column that popped up when I googled to find the column, for example, all fall into this camp.  A number would consist of rage and epithets and personal insults and claims to possess a higher truth by dint of birth that are again of no intellectual merit. 

Some would sophomorically make the "if Bill Gates were black, he wouldn't have become the  richest person in America" argument, which is (a) fallacious (ecological fallacy), because a statistical analysis, by definition, does not purport to be true of each and every member of the population studied (if it were, you wouldn't need a statistical analysis; and in fact you couldn't do one, as there would be only one statistic); (b) intellectually dishonest, because it materially changes the data set being studied, and (c) even if the sentence in quotes were true, the odds are very high that the person who replaces him as "the richest person in America" would be as white as he is, given the proportion of whites in the population. 

Some would nitpick (e.g.,, blacks are systematically undercounted in censuses - even so, but not to the extent that the white / black proportions would change materially, and I know from ancestral research that whites have been  repeatedly undercounted as well -- among my 8 great-grandparents, for instance, I can come up with 4 omissions off the top of my head from the censuses of 1910, 1920 and 1930, which is a 16.6% undercounting). 

Some would point out, intelligently, that the census data is insufficiently granular to capture the full extent of housing segregation -- there are suburbs and then there are suburbs, and even within suburbs. there was educational segregation (which in my opinion was the largest factor), and everywhere there were job opportunity barriers, often erected by labor unions and politicians responding to their demands (Google "Davis Bacon Act racist" and learn about its racist origins, for instance; why aren't there reparations lawsuits against unions for damages?)  True, I agree with all those points.  In fact, I would say the root of black economic underperformance lies in the fact that they arrived in urban economies, not only with just rural skills (technical and social), but also, unfortunately, at just the time when the economy was changing over to one that required much more sophisticated skills than even the white working class possessed so their gap was larger and has never closed.   But I must also point out that Coates has couched his argument in a simplistic black-urban / white-suburban duality and I am merely responding apples-to-apples.  Were he to write a more nuanced essay, it would be of greater intellectual merit, but likely not garner him the same publicity, so he sort of picked his poison and I decline to be judged by a higher standard than he set for himself.  I also note that Coates sharply criticizes education as an answer to black economic status, and is himself a dropout, so again, I am just responding to the man's particular argument, and I think in respect of education he is seriously wrong.  As well, education funding has been mainly locally raised and therefore, if a white community has higher education spending, they're not taking the excess out of black people's wallets, in some kind of "theft", to use Coates's word; they're funding it themselves by taxing their own white selves.  The progressive argument in this sector is really the opposite of what Coates is arguing, not that whites are taking from blacks, but that white income should be taken and spent in majority-minority school districts by replacing the local property tax method of funding with statewide tax and transfer mechanisms.   

Some would miss the point of this post and assert that this post whitewashes the harm done to black people from various laws, which in fact this post recognizes at multiple points; the point of the post is that contemporary white outcomes are almost entirely independent of contemporary black status and of policies and actions that harmed black people in the past.  Yes, there was segregation, discrimination and racism, and all of that was harmful and wrong, but the demographics are such that the vast majority of white outcomes would have been the same or very close to the same, even if there had been none.  That is just a quantitative fact when one population constitutes 88-90% of the total.

The "white privilege" rubric is just a tactical attempt to advance "social justice"  by de-legitimizing the success any given white person may have had, and thereby legitimizing the transfer from that person of some of his or her economic gains to minority populations, because, even though the most successful white people are the ones whose success is least attributable to race, they, as Willie Sutton once explained about robbing banks, are where the money is.  But it is an intellectually false dogma and deserves to be rebutted by anyone in the "reality-based community" where I've always prided myself on dwelling.