Sunday, November 16, 2008

Persistent poverty and the 2008 rural vote (Part I): Indian Country

Having given some attention to the role of rural gentrification and the rural resort phenomenon on the 2008 race for the White House (here), it seems appropriate also to look at how a different cluster of rural counties voted: those which are persistently poor.

The U.S. Census Bureau uses the term “persistent poverty” to refer to those counties whose poverty rates have been 20% or higher in each decennial census since 1970. (See p. 4 of the USDA ERS publication here). Our nation’s 386 persistent poverty counties are home to 4% of our population. Of these counties, 88% are non-metro, and residents of persistent poverty counties comprise 14% of the non-metro population.

The map of persistent poverty counties from the 2000 Census is shown above (non-metro persistent poverty counties in dark purple), and it indicates that pockets of persistent poverty are in the Mississippi Delta and extending across the black belt; in Appalachian Kentucky and West Virginia; in the Rio Grande Valley; on American Indian reservations in the Plains, and in several counties in the Southwest. A few are scattered through Oklahoma and the Texas panhandle, a few more in the mountains of northwest Arkansas and southeast Missouri. Alaska has wide swaths of persistent poverty, sparsely populated, of course. The northeast has none.

Note that these are only the poorest of the non-metro counties; many others have high poverty rates. Indeed, while the 2007 poverty rate for metro counties was 11.9%, it was 15.4% for non-metro counties.

So, who did voters in these pockets of enduring poverty support in the 2008 race for President? One might expect such populations to be supports of the Democratic Party in light of its embrace of some degree of wealth redistribution to provide a baselie of welfare for all. A casual glance indicates that some regions with persistent poverty overwhelmingly supported Obama; others went for McCain. (You can see the 2008 electoral map here; select "county leaders" in upper left hand corner). Obama supporters tended to be in the Misissippi Delta, black belt, Rio Grande Valley, and in Indian Country. In short, the tended to be voters of color. The McCain voters, on the other hand, were in Appalachia and other parts of the Mid-South.

I’m going to look at some of these regions in detail in a series of posts, beginning with Indian County, in the Great Plains and into the Mountain West. While the American Indian population has a history of leaning Democratic, a look back to ’00 shows that many persistently poor counties in Indian Country voted for Bush that year. They’ve been moving back into the Democratic fold since then, and this month they helped sweep Obama to victory. That is, they helped a little bit. Even where these counties overwhelmingly supported the Democratic ticket, their populations are so small as to have relatively little impact on their states’ total vote counts.

I wrote earlier about some Montana counties that supported Obama, suggesting that the vote there was a product of rural gentrification. In fact, the Montana counties that went for Obama featured varied demographic profiles; among them were three persistently poor counties with large American Indian populations. Among the bluest are three sparsely populated counties on the Canadian border: Glacier (population 13,247), Blaine (population 7,009) and Hill County (population 16,673). Fewer than 5,000 votes were cast in Glacier County, where more than two-thirds of the land is Blackfeet Indian Reservation and one-fifth is Glacier National Park. Sixty-five percent of Glacier County voters supported Obama. In Blaine County, where 45% of the population is American Indian (and, notably, another 26% claim either German or Norwegian ancestry), 58% supported Obama. Big Horn County (population 12,671) on the state’s southern border is also persistently poor and heavily American Indian. Voters there went for Obama in a big way, too – 67%. Roosevelt County, population 10,620, on the state’s eastern edge with North Dakota, is dominated by the Fort Peck Indian Reservation but is not persistently poor. Sixty-one percent of voters there supported Obama. It is worth noting that Glacier, Blaine and Big Horn Counties all voted Democratic in ’00 and ’04, too, but they did so by smaller margins.

In the Dakotas, most of the persistently poor counties voted for Obama. In South Dakota, these include Shannon County (population, 12,466) and Todd County (population 9,040) to the south, and Dewey (population 5,972), Ziebach (population 2,519) and Corson (population 4,181) counties to the north. All five lie entirely within one or more American Indian reservation, and all overwhelmingly supported Obama. Indeed, he won almost 89% of the vote in Shannon County. In addition, Buffalo County (population 2,032), South Dakota, where many Crow Creek Sioux live, 73% of voters supported Obama. Buffalo County has the unfortunate distinction of being the poorest county in the United States with a median income of $12,692, and a per capita income of just over $5K. Interstingly, several of these poor American Indian areas supported Bush in ’00, though Bush had lost most of that support by ’04.

In North Dakota, three of five persistently poor counties opted for Obama. Sioux (population 4,044) , Benson (population 6,964) and Rolette (population 13,674) counties all supported the Democractic ticket, while Grant (population 2,841)and Sheridan (population 1,710) counties went for McCain. The populations of Sioux and Rolette counties are primarily American Indian (85% and 73%, respectively), Benson County is roughly half white, half American Indian; and Grant and Sheridan counties are 96% and 99% white, respectively. Not surprising, then, in light of the evident trend with American Indian voters, Sioux and Rolette counties have been most consistent in supporting Democratic candidates over the past two decades. The extreme population sparsity of Sheridan and Grant counties -- fewer than 3 persons/square mile-- suggests large ranching operations, which would be consistent with fiercely independent, small government ideology, although it would not necessarily be consistent with the low median income (under $25K) and the high poverty level.

Outside the Great Plains, American Indians are somewhat more dispersed. While Arizona and New Mexico have significant American Indian populations living in persistent poverty counties, those counties often have large Hispanic populations, too. Because the racial and ethnic mix of these counties differs so greatly from the more concentrated pockets of American Indians in the Great Plains, I will discuss these in a separate post.

A 2007 book, Native Vote, offers an in depth analysis of the burgeoning significance of the American Indian vote, particularly in state and local contests. The authors are Daniel McCool, Susan M. Olson, and Jennifer L. Robinson, all of the University of Utah.

No comments: