What in the heck is wrong with Appalachia? I keep running into interesting correlations that tells me Appalachia should be giving far less support to Democrats at the presidential level than it actually does.
Here's the example from yesterday:
Click to enlarge.
If all of Central Appalachia behaved like the rest of the region, we'd expect to see all the points scattered near the line in the graph above. Instead, the points representing counties in parts of Appalachia go soaring off above 50%.
And Southern Appalachia does its own strange thing too. More below.
Ten Second Summary
Appalachia has been moving away from Democrats over the last 30 years, although faster rates of decline in Democratic support (in relative terms) can be found among non-Appalachian Southern whites. Still, Appalachia is far more supportive of Democrats in places than one might have expected based on several variables. Religion and history both likely play a role.
Back to Mississippi
Back in one of the first chapters of this series, I showed a plot of percent Obama versus percent white in Southern and Central Mississippi. There's a reason for that: counties in Northern Mississippi just don't fall on that line. They're a good deal more supportive of Obama than we might have expected given their racial composition. (Note that only counties where 97% or more of the population in 2000 was either African-American or white are included in the regression.)
Click to enlarge.
The same is true for Alabama. When you look at which counties don't fit on the line (shown in green below), it's a familiar pattern:
Click to enlarge.
The counties where Obama did better than expected based on the percent of the population that's white are almost all in Southern Appalachia. There's a few more along the Gulf Coast, too. Also note that the one red county is red not because whites there are far more Republican, but because it is 14% Native American, and thus an analysis based on a binary black/white universe does not work with that county.
Change, Change, Change
The pattern was even more striking in the past. Here's the graphs for the 2008, 2004 and 1988 elections:
Click to enlarge.
You can see Southern Appalachia used to be even more Democratic, given its racial composition, than the rest of Alabama and Mississippi. White support for Democrats has decreased in all parts of these states. Outside of Appalachia, support among whites in Mississippi and Alabama fell around 15-25% between 2004 and 2008 - a mere trifle compared to the 65-75% decrease between 1988 and 2008. In Southern Appalachia, support fell as well, but not as much - about 10-20% from 2004 to 2008, and 40-60% from 1988 to 2008.
Central Appalachia
In Central Appalachia we see a similar pattern. Although, as I showed in the introduction, we see far more support in parts of this region than we would have expected, this is after a substantial decline in support compared to support for Kerry. Indeed, many counties in this region showed some of the largest decreases in Democratic support at the presidential level from 2004 to 2008, both in absolute numbers and percentages.
A major difference shows up when you look at vote totals, however. In Appalachian Mississippi and Alabama, vote totals are generally a little higher in 2008 than in 2004.
In many of the counties with the greatest drop in Democratic support in Appalachian Kentucky, Tennessee, Virginia, and West Virginia, however, vote totals dropped a lot. For example, (and this is one of the most extreme examples) in Harlan County, Bush had 6659 votes to Kerry's 4332 votes. In 2008, McCain had 7165 votes (506 more than Bush) but Obama only had 2586 votes (1746 fewer compared to Kerry, a 40% drop). Harlan County is not one of the Central Appalachian counties that gives more support to Democrats than we expect. But the same pattern shows up in those counties as well - for example, Floyd County, Kentucky. This suggest that many Democrats in Central Appalachia didn't vote for McCain last year - they simply didn't vote.
Why?
We now have several questions to answer:
Why did Obama do so much worse than Kerry among many Appalachian whites?
Why do Democrats do so much better in Appalachia than we might otherwise expect?
Why are some Central Appalachians so different, politically, from their neighbors?
Why did Obama do worse than Kerry in so much of Appalachia?
First, let's assume that many Kerry voters stayed home in 2008 in Central Appalachia. Mathematically, it could be that many Bush voters stayed home and many Kerry voters voted for McCain, but that doesn't make as much sense to me. In Southern Appalachia, on the other hand, it was more of a switch to the Republican, as turnout did not decrease as much.
So why stay home or switch? Possibilities include: not a strong Obama campaign presence, and rumors Obama would eliminate the coal industry (a reason specific to coal-producing areas). Or maybe it's because he's African-American.
I can't claim to know what's in any voter's mind. However, I believe there's evidence that Obama's low support in this region was not because he's African-American. At least, not in Tennessee. Why? Because Harold Ford, in a Senate election in which race played a role, did better than John Kerry in every county in the state, including in Appalachian Tennessee. In fact, he essentially did just as well as Al Gore throughout the whole state, once home-county favorites are accounted for.
So in Tennessee at least, it's not that people wouldn't vote for an African-American, because they have in the past. Harold Ford did just as well as Al Gore, who did better than John Kerry, who did better than Barack Obama. So how else was Obama different from Kerry? Another possibility is the labeling of Obama as the ultimate "Other". Forget race - he's a Socialist Secret Muslim Foreign Anti-American Terrorist, born in -Hawaii- Kenya and raised in Indonesia!!!! As cskendrick put it (more reasonably) in a diary well worth the read:
A urbane, well-educated, well-off, well-spoken, forward-looking son of a recent immigrant who did not share the country-boy-can-survive ethos would have been a tough sell in the most affable of Appalachian moods.
The final potential reason I will bring up is simply a continuation of prior trends. Essentially, whites in Appalachia have a strong Democratic tradition that has been on a slow decline over the last thirty years that is accelerated in areas with a strong Baptist presence (not necessarily caused by Baptism itself). In Southern Appalachia, the Republicans' racist Southern Strategy has also played a role. The 2008 election is simply consistent with prior trends, and the splotch of purple showing decreased Democratic performance is not necessarily due to anything specific about Obama. Let's look at the voting records of selected nearly all-white counties to see how this plays out:
Click to enlarge.
On the left, we see two counties in Central Appalachia with very few Baptists and the state of West Virginia essentially following the country in voting preferences until about 1980. After 1980, the country gradually swings towards Democrats, while Central Appalachia slowly slides the other way, although still maintaining relatively strong support for Democrats. In the middle, we see a different pattern post-1980 in counties with large numbers of Southern Baptists: a surge of support for Clinton, followed by a precipitous drop as the faith-based Republican Permanent Majority get out the vote program is put in place. On the right, we see the effect of Civil Rights and the Reagan Revolution. In the Roosevelt era, white Southerners were strongly Democratic; after a wild ride of third party candidates and Nixon mania, Carter pulled the old Democratic coalition together one last time before the plunge. After 1980, we see a Baptist and Appalachian pattern in the two counties shown, according to their characters.
Why is Appalachia unusually Democratic?
Part of the answer to this question lies in perspective and the way I've set up the narrative of this diary. When we're looking at Alabama and Mississippi, whites in Appalachia are far more Democratic at the presidential level than whites in the rest of these two states. Comparing Appalachian Pennsylvania to the rest of that state shows a different story. What we're really seeing is a geographic/political/cultural boundary that follows the physical geography - the Appalachians on their Eastern and Southern side - instead of the political geography - state lines. (On the other side of the Appalachians, the boundary of political behavior is more diffuse, as we saw in the previous diary, and indeed extends to some degree well past the Appalachians, through the Upland South and Ozarks).
The real question then becomes, Why is Appalachia behaving as a political unit? That is a more complicated question to answer, rooted in history, class, and economics, and I would refer you again to cskendrick, and also Genius at Wrok.
Why are there large differences among "American" Central Appalachians?
In the previous diary, and in the Introduction, we saw some very large discrepancies in the voting behavior of those who label their ancestry as "American" or who do not chose a label. In some parts of Central Appalachia, this group more or less evenly split their vote; in other parts, it was a landslide in favor of McCain.
In Western Kentucky, for instance, the Northern half was fairly supportive of Obama; the Southern half had low support. There's no clear difference in variables such as race, education, and income. But one variable does stand out: religion, as shown above. Take a look at the table comparing two Appalachian counties in Kentucky:
Click to enlarge.
There's two major differences - population density and religion. The county that is more rural gave much more support to Obama. The county that is more Baptist gave much more support to McCain. This pattern holds throughout Appalachian Kentucky: maps show the patterns of religious membership and population density have a rough relationship to the pattern of places with relatively high Democratic support. And the churches we're talking about are predominantly Baptist.
Note that those not claimed by a religious body are not necessarily non-believers; rather, their denomination may not have been surveyed, or they may be religious without attending formal services. For instance, home-based bible study groups.
What we see here may be the effect of the institute of Baptist religion on politics. There could be something else at play as well: Why do we see the particular pattern of Baptist adherence in the maps above? Is it related to the geography and population density? There could easily be some underlying factor not captured by the numbers above the results in resistance to the Baptist faith, lower population density, and support for Democrats.
However, on the statewide level, there is a strong relationship between the percent of whites who voted for Obama and the percent who call themselves evangelical or born-again. This relationship is described in a diary that is essentially the next part of this series, although it was published last year.
For more on religion, we'll wait until tomorrow.
__________________
This diary is the thirteenth in a series taking a close look at the 2008 electorate and exploring three themes: diversity within demographics, progressive feedback loops, and demographic change.
Tomorrow: Why Republicans Should Be Really, Really Scared - Religion
Cross posted at Open Left.
Diaries in this series (updated list):
Looking Back
Alternate History
Why Republicans Should Be Really Scared
African-Americans – We Are Not All of Us Alike
East and South Asian Americans – Diverse and Growing
West Asian Americans – Rapid Change
Native Americans – Increasing Participation
Islander Americans – In Need of More Representation
Native Alaskans – An Economic Factor?
Latino Electorate – Increasing Influence
European-Americans – Tribal Politics Persist
“Americans” – You Might Be Surprised
Appalachia – Surprisingly Democratic
Why Republicans Should Be Really, Really Scared
Why Republicans Should Be Really, Really, Really Scared
A Few More Tidbits
Related 2008 electorate diaries:
Maps: Blue America and the Changing Electorate
Maps: Obama and White Evangelicals