Monday, December 15, 2008

Persistent poverty and the 2008 rural vote (Part III): The Mississippi Delta and the black belt

Most non-metro, persistent poverty counties are in the South. (See a USDA ERS report here). Among this particular group of counties, Obama drew support primarily from counties with significant Black populations. For example, most of the non-metro, persistent poverty counties that lie along the lower Mississippi River Delta—in Arkansas, Mississippi, and Louisiana—supported Obama. Most of these counties also supported Kerry in ’04 and Gore in ’00, as well as Clinton in the '90s Presidential contests.

Among the delta region counties in Mississippi that did not support Obama are Warren County, Mississippi (population 49,644), where McCain edged out Obama with 51% of the vote. Interestingly, Warren County, where blacks are 43% of the population, is the only Mississippi county bordering the river that does not share the persistent poverty designation. The poverty rate is nevertheless high there, at 18.7%. Persistently poor counties farther east in the state tend to have higher percentages of white residents, and they tend to vote Republican. 2008 was no exception.

Like Mississippi, Louisiana is a study in contrasts in terms of how its 24 persistently poor, non-metro parishes voted. Persistent poverty was not a predictor of how any given parish voted; race of voters was. McCain easily won most of these parishes, with Obama carrying only those where blacks are a majority of the population. East Carroll (population 9,421; 67% black), Madison (population 13,728; 60%), and Tensas (population 6,618; 55% black) parishes on the Mississippi River (and Mississippi state line) supported Obama by 64%, 58%, and 54%, respectively, but McCain won 60% of the vote in Concordia Parish (population 20,247), just south of those three and only 38% black.

Persistently poor parishes in other parts of the state—all with minority black populations—also supported McCain. In Sabine Parish (population 23,459; 18% black), McCain won a whopping 75% of the vote, while in neighboring Natchitoches (population 39,080; 38% black), McCain garnered a more moderate 53%. Tangipahoa (population 100,588; 28% black) and Washington (population 43,926; 31% black) parishes in the southeast part of the state are also persistently poor, with significant but not majority black populations. They voted for McCain 65% and 66%, respectively. Even in Claiborne Parish (population 16,851) on the Arkansas state line, where blacks are 47% of the population, McCain won 55% of the vote. Thus, a race-linked trend is evident throughout the state’s persistently poor, non-metro parishes.

A quick look at a smattering of persistently poor non-metro counties across the South—from Mississippi to North Carolina—reveals similar results. As a general rule, only counties with majority black populations supported Obama. Because most persistent poverty, non-metro counties in the South have significant black populations, a map highlighting these counties looks very similar to a map showing the counties Obama won. I’ve reproduced the persistent poverty map above; the electoral map is available here (click on “county level” in upper left corner). The so-called black belt is evident in both. What’s stunning to me is that in most of these counties, unless blacks are a majority of the population, McCain probably won. In many, many counties, the percentage of voters supporting Obama was within a couple of percentage points of the proportion of black residents in the county.

What’s equally stunning—at least to me—is that race of voter appears to be a strong predictor of party support at the Presidential level in elections over the past 20 years. If you slide your cursor over the tab at center left on the county-level ’08 map, you will see that most persistent poverty, non-metro counties supported the same party’s Presidential candidates in the ’92, ’96, ’00, ’04 and ’08 races. (The greatest exceptions are seen in Louisiana, where persistently poor white voters, along with most of the state, supported Clinton in ’92 and ‘96).

White people in rural, long-time pockets of poverty tend to vote Republican. Black people living in rural, long-time pockets of poverty tend to vote Democratic. So, a good argument can be made that race of the candidate did not drive voter choice among these populations. Race of the voter—long party-aligned—did. While Obama pretty clearly attracted black voters in the South, in fact, the Democratic Party has long had a good hold on most of these voters. On the other hand, most poor white voters in the South appear to have long been loyal Republicans. This seems consistent with some recent analysis published over at the Daily Yonder. They found that “[r]ural Republican communities on average have lower incomes and less education than rural Democratic communities,” divisions that are growing as people migrate. The Yonder study does not analyze on the basis of race, but its report suggests that being poor and rural doesn’t necessarily lead to a preference for Democratic administrations. Also, the Yonder's analysis is of the nation's rural places, and so it surely reflects the phenomenon of rural gentrification, which is not a major force in the South.

For analysis of the Southern vote in the ’08 race, see stories in the Daily Yonder here, here and here. See also Adam Nossiter’s NYT story about Southern voters here. My other posts on this and related topics are here and here. Read Part I of this series here, and Part II here.

1 comment:

Anonymous said...
This comment has been removed by a blog administrator.