Obama's Problem with White Voters

The racial dimension of Barack Obama's electability problem is now apparent, but no prominent Democrat dares discuss it openly. Similarly expect no discussion of the subject in the major media.

The white working class vote 

I am not referring to the ongoing and intense discussion of The Reverend Jeremiah Wright. Wright is a separate problem for Obama. Whether Obama has been, or will be, permanently weakened by his long and close association with Wright, or has soared above it with his Philadelphia speech, is not the subject of these thoughts. Something much simpler than the answer to that question has been starkly apparent for some time, certainly since well before the Wright eruption: Consistently, and by large margins, Obama has lost the white working class vote to Clinton in all states critical to the Democratic ticket this November. The lurking suspicion -- impossible to verify or refute -- is that much of Clinton's handsome portion of this demographic will not go to Obama in the November election.

This has grave implications for a Obama, at least in Ohio, Michigan, Florida, Pennsylvania and New Jersey. Working class whites have voted heavily for Clinton in these states (or, in the case of Pennsylvania, will soon do so). The return of the Reagan Democrats, not the odious fulminations of Reverend Wright and their consequences, is what is now driving Democratic Big Wigs to the bourbon cabinet. Predictably, the media either refuses to acknowledge this now established voting pattern or, in some cases, actually denies its existence.    

The latest example of denial is by Dan Balz, staff writer for the Washington Post, who remarked in his March 17, 2008 article purporting to analyze the white male vote, that Wisconsin (where Obama did relatively well among white males overall) and Ohio (where Clinton crushed him, 66-33%, among white working class males) are "states with striking similarities." It appears Mr. Balz has not looked at the two states closely and thoughtfully. In the crucial details of racial demographics, Ohio and Wisconsin are worlds apart; and it is through these details that Obama's white working class problem can be understood.

A tale of two states

Here are some pertinent facts about Wisconsin and Ohio: Wisconsin has about 5.5 million residents, Ohio about 11.3 million. Wisconsin is about 89% white and 5.7% black, while Ohio is 85% white and about 11.5% black. The small (but statistically significant) difference in percentage of blacks living in the two states was the least part of Obama's problem in Ohio. Obama's real difficulty in Ohio - and it has been a consistent one for him in similar states -- is the widely dispersed and interwoven location of the two racial groups in that state, versus their relative isolation from each other in Wisconsin. Here, I warn the reader, we are entering emotionally rough terrain for those schooled only in the mandatory American racial catechism of the last forty years.

For at least the last two generations America's racial policies have been predicated on a near religious belief that increased contact between the races will produce harmony, good feelings and positive relationships. Our experience during this period has been uniformly the opposite. Urban white liberals have fled the public schools by the hundreds of thousands, self-segregation by blacks on university campuses is widespread, resentment in the workplace (by both races) ubiquitous etc. In his Philadelphia speech Obama himself referred -- perhaps the first such reference by a black politician without open contempt -- to the concerns that many white Americans have about blacks.

The salient fact is this: in settings where the two races deal more directly with each other, and get to know each other better, through shared public schools, workplaces, public conveyances, universities, etc., they seem to like each other less, not more.

This fact is laid bare, at least for anyone willing to see it, by the Democratic primary results thus far.

Consider the following additional facts about Wisconsin and Ohio, those states with "striking similarities." 

In Wisconsin more than 75% of the black population resides in the Milwaukee area, a metropolitan area that accounts for only 32% of Wisconsin's total population. This means that in Wisconsin the white portion of 68% of the state's population (which is more heavily white than the state as a whole because of the concentration of blacks in Milwaukee) rarely if ever encounters blacks. Thus, for a high proportion of Wisconsin whites, blacks are abstractions, approached most closely by turning on Oprah.

Now consider Ohio: to begin with, the black population, in percentage terms, is nearly double that of Wisconsin (11.5% versus 5.7%). But its dispersion within and among the white population is the real difference between the two states' racial demographics. In Ohio 80% of the state's 11.3 million residents reside in the eight largest metropolitan areas (Columbus, Cleveland, Cincinnati, Toledo, Akron, Dayton, Youngstown and Canton). These cities contain, in the order listed, 24%, 51%, 43%, 24%, 28%, 43%, 44% and 21% black residents. Thus, in Ohio a very high percentage of the white population, particularly its working class component, has regular contact with blacks, or, if living in outer suburbs, has direct contact with other whites who do.

The widely disparate residential patterns of the races is obvious: in Wisconsin, the vast majority of whites live, work, shop, and send their children to school in a world that includes few if any blacks; in Ohio the reverse is true, and the races regularly brush up against each other in all these categories of daily life. Judging from how well Obama did among white voters in these states (satisfactorily in Wisconsin, abysmally in Ohio) increased racial familiarity is not a boon to the Illinois Senator.

The sad truth about racial interaction

Good debaters (and those on the ideological Left) will point out that I have linked two phenomena causally (racial interaction, on the one hand, and disinclination by working class whites to vote for a black candidate, on the other) without actually demonstrating cause and effect. But fortunately it does not take a Ford Foundation grant and a two year study to see what is happening. In this year's Democratic primary results the two phenomena -- extensive racial interaction and poor outcomes for Obama among working class white voters -- have been so universally conjoined that cause and effect can be reasonably presumed.

Without exception, the Wisconsin pattern (little interracial contact) and the Ohio pattern (much more such contact) have correlated with identically opposite results throughout the Clinton/Obama battles: every state outside the South where Obama carried the white vote and won the primary or caucus was one with a small to negligible black population (Wyoming, Vermont, Wisconsin, Maine, Washington, Nebraska, Minnesota, Kansas, Utah, North Dakota, Idaho, Alaska and Iowa);  in every state where a substantial and widely dispersed black population regularly interacts with whites, Obama lost the white vote and lost the primary: Texas, Ohio, Rhode Island, California, New Jersey, and Massachusetts. I have omitted the candidates' home states (New York for Clinton, Illinois and Hawaii for Obama). Pennsylvania, where Clinton has a commanding lead, will follow the Ohio pattern, as will Florida and Michigan in the increasingly unlikely event of do overs.

Simply put, blacks and whites are not doing well together in America, circa 2008. Obama's battle with Clinton, all the pretty rhetoric notwithstanding, is remorselessly exposing that undiscussed fact. Obama is hurt by this -- severely it would appear -- in states where the races interact extensively, particularly at the working class level; while, in states with few blacks, the lamentable state of America's race relations is masked and Obama does reasonably well among white Democrats.  

But the states with extensive racial interaction are precisely those that Democrats regularly carry, or need to carry, to win. Of course in several such states whites in general vote sufficiently Democratic to overcome the now obvious disinclination of working class whites to vote for Obama (e.g., Massachusetts, New York, California). But that is not true of the critical states mentioned above and, possibly, several others.


Michelle Obama attended Princeton and the Harvard Law School. Taking her at her word, interacting with whites in these rarified settings did little to improve her feelings about her country, including, presumably, the whites who made up the majority of her classmates. Given America's current rules of racial engagement -- which allow negative views of whites by blacks to be expressed but forbid the reverse - Mrs. Obama felt free to express herself publicly (though now, no doubt, wishes she had been less candid).  

On the other side of the divide, the only remaining permissible venue for white expression of racial grievance is the voting booth. Where social policy, proximity, and numbers create mandatory interaction by whites and blacks in settings less elegant than Princeton and Harvard, white disenchantment engendered by that interaction finds its outlet in elections.

The theory that greater familiarity is an antidote to mutual antagonism holds only if each party likes what it sees in the other as the familiarity develops. This does not appear to be the case with either principal race in America. The consequences are playing out at the ballot box.

Doubters of this reality should not only consider Mrs. Obama's words, but take a look at the racial demographics of states outside the South where her husband won the white vote (and the state), and compare them with the racial demographics of the states where he lost the white vote (and the state).

Whether this voting pattern will persist is a matter on which no guess is ventured. Whether the Wright fiasco will worsen it for Obama is unknowable. That the pattern does exist is an indisputable fact.   

James Edmund  Pennington is the pen name of an attorney.