President Bush's 2004 margin-of-victory was the slimmest of any presidential incumbent (Bush: 51%; Kerry: 48%); his margin-of-victory was about three million votes cast nationally, and Bush's electoral vote victory hinged on his winning Ohio's 20 electoral votes. Bush's popular vote margin-of-victory in that state was a meager 118,601, which was just 2.1% of the total votes counted in that state. Theories of voter fraud notwithstanding, it is generally accepted by both major political parties that the GOP was able to retain the White House in 2004 because of its ability to mobilize Evangelical Christian voters and other "values" voters across the country. After having discovered in the Journal of Blacks in Higher Education that the number of active hate groups in the United States rose over 5% from 2004 to 2005, I wondered the unthinkable.
I wondered if something else might have been also operating to push the Republicans to victory in 2004. Hate. Or, its predecessor: Intolerance. To explore this possibility, I examined the 2004 presidential voting results across the United States to see if George W. Bush fared better in communities with an active hate group, or CWAHGs (pronounced SEE-wog) than he did in the state in which those CWAHGs are located.
Since I am no expert on what constitutes a hate group, let alone an expert on where they are active, I turned to a website, tolerance.org, run by the Southern Poverty Law Center, an organization with which I have no affiliation. Tolerance.org shows what hate groups are operating in which states and towns. Using that information as well as geographical information available at other internet websites, 93% of the time I was able to locate from which counties these active hate groups operate. For purposes of my research I considered each of these counties a CWAHG.
But before I reveal and discuss my results, a word about stereotyping. I am not suggesting that many people who live in a CWAHG belong to an active hate group, and I am not suggesting either that most persons who live in a CWAHG are hateful or intolerant. CWAHGs--and maybe you live in one--were merely markers I decided to examine, markers of relative micro-community intolerance. My assumptions are simple ones. If a community (usually a county) has an active hate group--like a currently operating chapter of the Ku Klux Klan--then chances are it is a location of greater relative intolerance than locations without hate groups, even nearby locations. I assume too that the more intolerant a community, the more likely that community as a whole will be attracted to political candidates and ideologies that espouse intolerance. I do not assume, however, that CWAHGs are the only places where intolerance exists. Sadly, we all will agree intolerance and hatred are ubiquitous. They are states of individuals' minds. What I set out to do was simply this: look for a reasonable micro-community marker of relative intolerance, and compare it to actual election results. CWAHGs seemed a logical choice; and, to my knowledge, this type of voting analysis has never before been conducted.
Now a brief caveat about my research methodology. Although the concept of a micro-community cannot encompass an entire, diverse large metropolis, nonetheless I included in my data all counties--even those with large cities, such as Boston and San Francisco (yes, folks, those cities have active hate groups). I know that by including these large cities--which generally favored John Kerry over George W. Bush in the 2004 election--my analyses would be somewhat slanted towards finding that intolerance had nothing to do with the outcome of the 2004 presidential election. But, my reasoning to include counties with large cities was this: I wanted to be as complete as possible; and frankly, I wanted to give our elected president that small benefit of the doubt. After all, I'm testing a disturbing hypothesis: did hate and intolerance affect who our currently sitting president is?
Here's what I found.
Of the 803 active hate groups listed at tolerance.org in 2005, I was able to determine that 753 (93.77%) of them are located in 423 counties in 44 states (this number excludes the District of Columbia because it does not have a county; and, the county locations of the remaining 43 hate groups are unknown, because they are more widely dispersed and do not operate out of a single, identifiable county). Next, I tabulated county-by-county who voted for Kerry and who voted for Bush in these 423 CWAHGs. Then, I compared the CWAHG voting results with the state's overall voting results (official voting results were used, not exit poll numbers). As I said earlier, I wondered if Bush fared better in the CWAHGs than he did overall in the state where the CWAHGs are located.
My results are shown in the table below, and here is how to read it. Let's take Arkansas as an example. In 2005 Arkansas had active hate groups operating out of 13 of its 75 counties (which is about 17% of its counties). Thus, it had 13 CWAHGs whose election results I examined (second column). George W. Bush had 54% of Arkansas' popular vote, and in 8 of the 13 CWAHGs he won more than 54% of the vote (last column). On the other hand, John Kerry got 45% of the Arkansas vote, and in 3 of the 13 CWAHGs he got more than 45% of the vote (column three). In the two remaining CWAHGs, both candidates got the same percentage of votes as was reflected in Arkansas (column four).
REGION/STATE | # of CWAHGs | CWAHGs favoring Kerry | CWAHGs favoring neither | CWAHGs favoring Bush |
Northeast: | 16 | 6 | 1 | 9 |
... |
Connecticut | 3 | 1 | 0 | 2 |
Maine | 0 | 0 | 0 | 0 |
Massachusetts | 3 | 1 | 1 | 1 |
New Hampshire | 2 | 2 | 0 | 0 |
New York | 8 | 2 | 0 | 6 |
Rhode Island | 0 | 0 | 0 | 0 |
Vermont | 0 | 0 | 0 | 0 |
... |
Mid-Atlantic: | 54 | 23 | 6 | 25 |
... |
Delaware | 1 | 1 | 0 | 0 |
Maryland | 5 | 2 | 1 | 2 |
New Jersey | 11 | 6 | 2 | 3 |
Pennsylvania | 17 | 4 | 3 | 10 |
Virginia | 20 | 10 | 0 | 10 |
... |
South: | 219 | 78 | 8 | 133 |
... |
Alabama | 14 | 8 | 1 | 5 |
Arkansas | 13 | 3 | 2 | 8 |
Florida | 23 | 8 | 0 | 15 |
Georgia | 24 | 6 | 0 | 18 |
Kentucky | 9 | 3 | 0 | 6 |
Louisiana | 13 | 5 | 0 | 8 |
Mississippi | 18 | 4 | 1 | 13 |
North Carolina | 24 | 13 | 2 | 9 |
Oklahoma | 9 | 7 | 0 | 2 |
South Carolina | 23 | 9 | 0 | 14 |
Tennessee | 24 | 6 | 1 | 17 |
Texas | 19 | 4 | 1 | 14 |
West Virginia | 6 | 2 | 0 | 4 |
... |
Midwest/Plains: | 79 | 28 | 7 | 44 |
... |
Illinois | 7 | 1 | 0 | 6 |
Indiana | 11 | 4 | 0 | 7 |
Iowa | 2 | 2 | 0 | 0 |
Kansas | 5 | 2 | 1 | 2 |
Michigan | 13 | 3 | 3 | 7 |
Minnesota | 5 | 2 | 0 | 3 |
Missouri | 12 | 3 | 2 | 7 |
Nebraska | 3 | 2 | 0 | 1 |
North Dakota | 0 | 0 | 0 | 0 |
Ohio | 16 | 5 | 0 | 11 |
South Dakota | 0 | 0 | 0 | 0 |
Wisconsin | 5 | 4 | 1 | 0 |
... |
West: | 55 | 21 | 5 | 29 |
... |
Alaska | 0 | 0 | 0 | 0 |
Arizona | 6 | 1 | 0 | 5 |
California | 18 | 6 | 1 | 11 |
Colorado | 7 | 2 | 2 | 3 |
Hawaii | 0 | 0 | 0 | 0 |
Idaho | 3 | 3 | 0 | 0 |
Montana | 4 | 2 | 0 | 2 |
Nevada | 4 | 1 | 1 | 2 |
New Mexico | 0 | 0 | 0 | 0 |
Oregon | 3 | 2 | 0 | 1 |
Utah | 2 | 1 | 0 | 1 |
Washington | 5 | 2 | 1 | 2 |
Wyoming | 3 | 1 | 0 | 2 |
When totaling the numbers for the entire nation, the results show that in the 2004 presidential election George W. Bush had a greater margin-of-victory (or much less frequently occurring, a lesser margin-of-loss) in 56.7% of those communities with an actively operating hate group than he had overall in the states where the hate groups are located. Specifically, in 240 of 423 CWAHGs Bush scored better numbers than he did overall in the states where the CWAHGs are active. John Kerry had a greater margin-of-victory (or a lesser margin-of-loss) in 36.9% of the CWAHGs, by contrast.
Are these differences mathematically significant, or just chance occurrence? In other words, is the 56.7% versus 36.9% finding significant beyond what would be expected to occur randomly?
Yes, the margin-of-victory differences between the candidates in the CWAHGs are statistically significant (such a difference should only occur randomly every 40,000 years). Moreover, even if one were to lump those CWAHGs that favored neither candidate into John Kerry's overall tally, the percentage difference between the two candidates (Bush: 56.7%; Kerry: 43.3%) is still statistically significant (such a difference should only occur randomly every 400 years). These results mean that something was going on in 2004 with respect to Bush's better showing in communities with an active hate group, but it was not random chance. Keep in mind my methodology was somewhat skewed to find that intolerance was unrelated to Bush's re-election. Most clearly these voting results suggest that in relatively intolerant micro-communities Bush's message resonated significantly better with voters than John Kerry's message. I argue too that in relatively intolerant minds across the country President Bush's message resonated better with voters than did John Kerry's message, minds perhaps made more intolerant--even if only temporarily--by the GOP's pre-election blitzkrieg of scare ads. My voting analyses lend support to the notion that using hot-button issues that inflame a person's level of intolerance the GOP played the Hate card in 2004, and by the tiniest of margins, it worked.