"Sen. Obama, this is truly a good day for America. Too often the achievements of our opponents go unnoticed," McCain says to the camera. "So I wanted to stop and say, congratulations. How perfect that your nomination would come on this historic day. Tomorrow, we'll be back at it. But tonight senator, job well done."

That was McCain's classy congratulation to Obama for accepting the Democratic nomination for President on the anniversary of MLK's "I have a dream" speech. This historic event certainly demonstrated just how far America has come towards realizing King's dream. However, we don't yet live in a race-blind society.

Political campaigns are run with the goal of winning an election. As such, they need to fully understand how voters make decisions, and in what ways voters are persuadable. Some campaigns will avoid "dirty politics" -- tactics that would be effective but which it is felt cross an ethical line. Some won't. You as an informed citizen ought to understand how the issue of race can be exploited by campaigns, and then you can decide for yourself whether a particular campaign is playing to racial sentiments.

Let us acknowledge upfront that the vast majority of politicians today are not themselves racist, although this Presidential race has brought out a few instances of highly questionable rhetoric (Rep. Davis, R-Ky, referred to Obama as "boy" and Rep. Westmoreland, R-Ga, described Obama as "uppity" ). In fact, let us be very careful with the term "racism" itself, which is defined in very strong language. Certainly racial messages can express racism, and can be made by racists, but it's not clear to me that we gain anything by insisting upon the necessity that this be so. I'll try to use "racial" throughout, with the simple definition of "relating to race."

Before we dive in, we should also briefly address the so-called Bradley (or Wilder) effect, meaning the alleged tendency of voters to lie to pollsters by saying they will vote for a black candidate in order to not appear racist. There is no evidence that such an effect currently exists, so there is no reason to believe that polling in this election is overstating Obama's support. From the link:

Hopkins looked at all senatorial and gubernatorial races that featured a woman or an African-American candidate from 1989 to 2006 -- a total of 133 races. For each, he found at least one poll released within a month of Election Day, enabling him to measure the gap between a candidate's polling and performance. Hopkins finds some evidence that African-American candidates suffered from something resembling a Wilder effect before 1996, but since then, the effect seems to have disappeared.

See also this excellent analysis of the primary race. Of course events could cause support to shift, and this would then show in the next round of polling, but the take-away point is that the degree to which Obama is losing votes because of his race is already factored into the polls (as is the offsetting degree to which he gains votes because of his race).

Perhaps the defining work on race in politics is The Race Card , by Tali Mendelberg (Google books ). From the Princeton books site:

In the age of equality, politicians cannot prime race with impunity due to a norm of racial equality that prohibits racist speech. Yet incentives to appeal to white voters remain strong. As a result, politicians often resort to more subtle uses of race to win elections. Mendelberg documents the development of this implicit communication across time and measures its impact on society.

[...] Mendelberg's conclusion is that politicians--including many current state governors--continue to play the race card, using terms like "welfare" and "crime" to manipulate white voters' sentiments without overtly violating egalitarian norms. But she offers some good news: implicitly racial messages lose their appeal, even among their target audience, when their content is exposed.

This differentiation between explicit and implicit racial messaging is critical. There is no incentive for a political campaign to express explicitly racist or racial messages; voters will punish them for crossing that bright line. This is to our credit as a society. However, there are voters that harbor prejudices and there is an incentive for a campaign to activate those prejudices. This is possible to do via implicit racial messaging, for example negative advertising that evokes racial stereotypes without explicitly introducing race. Such an approach sidesteps the drawbacks of explicit messaging while still connecting with the target audience. Not only does the campaign maintain plausible deniability, viewers don't even necessarily realize they are being influenced on a racial level.

Since the publication of The Race Card, there has been significant additional work done on this topic, and there have also been many more opportunities to collect data (such as the 2006 Corker-Ford Senate race in Tennessee). One particularly interesting result is Huber & Lapinski's 2006 article The “Race Card” Revisited: Assessing Racial Priming in Policy Contests (PDF). Let me first quote the description of their experiment, to make clear just what "explicit" and "implicit" racial messaging means:

In experiment A, the manipulated section of the advertisement varies in the text spoken by the announcer. The language is either neutral or explicitly racial. In the neutral language version, the middle section of the advertisement opens with a picture of an African American woman at the top of the frame. As the announcer states, “Too many welfare recipients take advantage of our tax dollars,” “Workfare not Welfare” appears below the picture. The announcer then continues, “To end dependency, we must require everyone receiving welfare to work,” as “Help End Dependency” appears below “Workfare not Welfare.” Following the IE model, we label this treatment as being implicitly racial. In the racial language version, the announcer’s first sentence is, “Too many welfare recipients, especially blacks, take advantage of our tax dollars.” [...] The remainder of the advertisement is the same. We label this version, per the IE model’s classification, as being explicitly racial.

These authors "confirm that individuals do tend to reject explicit appeals outright, but find that implicit appeals are no more effective than explicit ones in priming racial resentment in opinion formation." The apparent contradiction with Mendelberg's model is explained by considering an additional factor: the level of education of the viewer.

Among less-educated individuals, either implicit or explicit appeals appear somewhat effective in activating racial predispositions in opinion formation on non race-related policy. For candidates targeting this set of individuals, however, there is no clear advantage to offering an explicit message over an implicit one. Moreover, more-educated individuals do react more negatively to explicit appeals than implicit ones. These individuals appear relatively impervious to priming in any case, but candidates may still face real costs if they anger voters by violating perceived racial norms. Consequently, given that any effort to target racial messages to less-educated individuals might still reach some of these potential voters, there are large potential costs and no clear benefits to using explicitly racial language in campaign ads. Doing so risks public censure without offering clear priming benefits.

Overall, it is my impression that race is losing power as a lever with which to manipulate public opinion, even implicitly. (Note that this doesn't mean there aren't politicians still seeking to exploit it, or that such efforts don't still enjoy some success -- there are, and they do.) There are a couple reasons for this reduced relevance of race-based messages: Americans are becoming more tolerant, issues that offered a hook in the past (e.g., welfare) aren't politically relevant any longer, and given the current state of the economy/world prejudices are a luxury we can't afford. Consider this anecdote :

So a canvasser goes to a woman's door in Washington, Pennsylvania. Knocks. Woman answers. Knocker asks who she's planning to vote for. She isn't sure, has to ask her husband who she's voting for. Husband is off in another room watching some game. Canvasser hears him yell back, "We're votin' for the n***er!"

Woman turns back to canvasser, and says brightly and matter of factly: "We're voting for the n***er."

I guess that's progress of a sort, right? But there might be another reason that race-based appeals to voters aren't as prominent: the nature of leading black politicians. Consider this article from the American Prospect, titled "Young, Black, and Post-Civil Rights":

So how much has the country changed? This is the question of the moment as we watch the mutations in our national racial DNA triggered by Barack Obama's presidential campaign. Obama, we are reminded constantly, is a singular political talent. But he is in many ways the full flowering of a strain of up-tempo, non-grievance, American-Dream-In-Color politics. His counterparts are young, Ivy League professionals, heirs to the civil-rights movement who are determined to move beyond both the mood and the methods of their forebears.

[...] These new leaders are not what used to be called race men. They argue, somewhat convincingly, that they don't need to concern themselves primarily with the uplift of their race. They appeal to black voters, to be sure, but to white ones as well. They talk about income inequality, not black unemployment. They rail against inadequate educational opportunities, not the endemic poverty in black neighborhoods that results. They attack globalization and outsourcing, not necessarily the loss of high-wage, low-skill manufacturing jobs that built and sustained large working- and middle-class black communities after World War II. 

Candidates who don't fit the stereotype of "typical black politician" are probably less vulnerable to attack ads based on activating innate prejudices, since viewers might be less likely to link negative views of blacks in general with those candidates.

The final reason why I see race-based messaging decreasing in use and relevance is the increasing sophistication of viewers and analysts. It's difficult to sneak a coded message past the 24/7 cable news chatterers and the hordes of excitable bloggers -- heck, sometimes they find "messages" that few people would have noticed! However, the debate over the RNC "Call Me" ad in the TN 2006 Senate election illustrates that it is still difficult to establish a bipartisan consensus on what is or is not implicit racial messaging -- conservatives in general did not view that ad as playing to stereotypes about black men and white women, while many liberals did interpret the ad in such a manner. Unfortunately any interest in objectively evaluating the content of the ad was lost in a loud chorus of accusations of racism and race-baiting (not from the candidates, though: Ford said "I don't think race had anything to do with that ad" and Corker called it tacky and asked for it to be pulled). As to the effect of this ad, it seems clear that Ford lost ground in the polls shortly after it aired, but there were other events around that time (including Ford attempting to crash a Corker press conference, and of course the meta-level discussion of the ad) that could also explain the slide. Notice that his support in the polls actually slightly understated his performance on election day; there was no Bradley effect.

When we as a society can consistently and in good faith discuss whether an ad uses implicit racial messaging then the usefulness of this technique will be compromised. And I do mean we as a society; a black candidate who complains about implicit racial messaging is likely to face a large backlash, just given simple demographic realities. Academic work, including that cited above, establishes that implicit racial messaging is often not identified as objectionable, and a natural defensive reaction to the suggestion that a viewer missed implicit racial messaging is to hear an accusation of racism, strongly deny it, and accuse the critic of "playing the race card" or some such. Bizarrely, a campaign might benefit from using race implicitly in their advertising beyond any immediate influence on voters if they were able to provoke criticism from the opponent and convince voters that the black candidate was levying unfair accusations regarding their use of race. Hence, effective identification of such techniques would have to come from outside observers. Ideally viewers themselves would subject ads to analysis, attempting to compensate for their political biases, based on an understanding guided by work such as that described above. Something to keep in mind when you do this: political campaigns are perfectly free to employ "counter-stereotyping" in their advertisements to nullify the possibility of priming racial resentment through implicit messaging; for example, they could use an image of a white man when discussing urban crime, or avoid hot-button words like "welfare" when describing non-welfare related policies like taxes. If a campaign chooses not to do so, that is itself a data point worth considering, although obviously not conclusive.

So, this hopefully non-controversial introduction gives us some tools to decide for ourselves whether a campaign or ad is attempting to use race to influence perception. It's worth reiterating that this isn't the same thing as characterizing an ad, a campaign, or a candidate as racist, and it doesn't mean that viewers who respond to the tactic are racist, and it certainly doesn't mean that viewers who miss the implicit link to race are racist. Also, there are all kinds of sleazy tactics that campaigns can use (the most obvious being outright lying) and how one ranks particular techniques is subjective. That explicit appeals to race are recognized and rejected is good news. The next step is recognizing implicit appeals. I'd like to look at some ads in detail, and I hope we can bring "fresh eyes" to them, set aside preconceived notions, refrain from injecting incendiary terms or accusations, and simply examine the material.