Social scientists are overwhelmingly politically liberal, which would not be a problem if they could keep their politics from tainting their science. Unfortunately, it often seems they cannot. See Jussim, 2012, for many such examples involving unjustified claims about stereotypes — such as the oft-repeated but repeatedly disconfirmed claim that stereotypes are "inaccurate."
A recent case is an editorial that appeared in the NYTimes Sunday Review of 6/10:
The author, a graduate student in economics at Harvard, performed an interesting study but then reached an entirely unjustified conclusion about the likely role of racism in the upcoming presidential election. He mined Google search data to: 1) identify areas where there were unusually large numbers of anti-Black racist searches (e.g., searches using racial slurs); and 2) found that the 2008 vote for Obama was depressed by an average of 3-5 percent in those areas. Mining Google search data for this type of information was creative and may provide a new tool for research in the social sciences. Furthermore, high community levels of racism depressing the pro-Obama vote by 3-5 percent is consistent with lots of other research (see references at end).
The problem is that the author did not stop at this conclusion, which is quite reasonable, and went on to reach a conclusion that is unjustified: "If my findings are correct, race could very well prove decisive against Mr. Obama in 2012."
This rises to an unacceptable level of social scientific misinterpretation on both logical and empirical grounds. We all have the right to our opinions; if, however, we claim our opinions are based on "science" then it behooves us to at least remove obvious sources of political taint from our "science."
First, the logical issues. The author focused exclusively on disadvantages to Obama due to anti-black racism. He failed to consider: Obama's racial advantages; McCain's racial disadvantages; or McCain's racial advantages. If one wishes to reach conclusions about net racial (dis)advantage, it is obviously necessary to assess all four racial effects. It is appalling that neither the author nor anyone at the Times editorial staff caught this simple logical oversight.
In fairness, the author did briefly consider the advantages that accrued to Obama because of his race, but summarily dismissed them: "Yes, Mr. Obama also gained some votes because of his race. But in the general election this effect was comparatively minor."
Which gets to the second flaw: It behooves social scientists making such proclamations to actually know the data. Fortunately, the relevant data are widely available and summarized in the final chapter of my recently-published book. Here are some excerpts:
• Obama received a higher proportion of the white vote than did Kerry or Gore (Observationalism, 2008).
• Since WWII, there have been six Presidential elections where neither candidate was incumbent (as in 2008). On average, the Democrat lost those elections by 4 percent. Obama won by 7 percent, i.e., 11 percentage points better than a typical Democrat in such elections.
• Nonetheless, about 5-7 percent of the voters did not vote for Obama because he is black. This was the conclusion reached by national surveys conducted by Yahoo/Stanford University and Gallup before the election (Gallup, 2008).
• But, what about the proportion of the vote that McCain lost because of his race? This is the type of question that almost never occurs to many social science researchers concerned about issues of racism, sexism, and bigotry. Fortunately, however, it did occur to the Gallup organization (Gallup, 2008), who asked voters whether they were more or less likely to vote for Obama or McCain because of their race. They found that about 6 percent said they were less likely to vote for Obama because of his race. However, they also found that 9 percent said they were more likely to vote for him because of his race; and that 6 percent said they were less likely to vote for McCain because of his race (they also found that 7 percent said they were more likely to vote for McCain because of his race).
So, when you put all this together:
1. Racial preferences did play some small role in the election
2. There was little or no net disadvantage for Obama because he is black.
Why do so many social scientists have the type of blind spot to phenomena such as pro-black bias or anti-white bias revealed in that NY Times editorial? There is no way to know for sure in any particular case. There is, however, a strong contender: politics, especially the type of leftwing politics that dominates the social sciences. Biases against stigmatized groups so swamp the radar of many social scientists that simple questions such as "how many votes did Obama gain because of his race" and "how many did McCain lose because of his race" do not even come up, and, when they do, they are summarily dismissed without examination of data (as in the case of the NYTimes editorial). Political blinders can lead to distorted conclusions that reflect social scientific incompetence more than they reflect the data.
Image credit: MySafetySign.com
Gallup (October 9, 2008). Obama’s race may be as much a plus as a minus. Retrieved on 6/26/12 from: http://www.gallup.com/poll/111049/obamas-race-may-much-plus-minus.aspx
Jussim, L. (2012). Social perception and social reality: Why accuracy dominates bias and self-fulfilling prophecy. NY: Oxford University Press.
Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion. New York: Pantheon
Haidt, J. (2011). The bright future of post-partisan social psychology. NYTNYTRetrieved 6/26/12 from http://www.edge.org/3rd_culture/haidt11/haidt11_index.html.
Observationalism (November 9, 2008). Selected exit poll comparisons, 2000-2004-2008. Retrieved on 6/26/12 from: http://observationalism.com/2008/11/09/selected-exit-poll-comparisons-20....