2012 U.S. presidential election

Changes in U.S. Electoral Geography from 2000 to 2012: A Renewed North/South Divide?

As noted in a previous post, the presidential contest of 2000 seems to have been a watershed event in U.S. electoral geography. Up until that point, successful Democratic candidates enjoyed considerable support in many predominantly rural counties dominated by Whites, particularly in the Upper South (see the map of the 1996 election). In order for the Democrats to have carried many of these counties, southern candidates seem to have been necessary. As a result, all successful Democratic candidates from 1964 to 1996 were southerners (Lyndon Johnson, Jimmy Carter, and Bill Clinton). In 2000, the Democratic Party held to the same strategy, nominating Al Gore of Tennessee. But Gore, who is a much stronger environmentalist than most southern Democrats, lost support massively across rural, White America. Although he narrowly won the popular vote nationwide, Gore lost the Electoral College, as most states opted for George W. Bush, including Gore’s native Tennessee.

Since the 2000 election, the basic patterns of electoral geography at the county level have remained relatively constant. But a number of relatively minor shifts have occurred, which are worth examining.  Today’s post therefore compares the 2000 and 2012 U.S. presidential elections at the county level. These two contests make particularly good comparisons, as their popular vote figures were close, especially on the Republican side (in 2000, Al Gore took 48.4 percent of the vote against George W. Bush’s 47.9 percent, whereas in 2012, Barack Obama took 50.7 percent against Mitt Romney’s 47.7 percent).

The basic differences between the two elections result from a continuation of the trends that produced the map of 2000. Democratic support in the interior portion of the Upper South continues to plummet, with most of the region’s few remaining blue counties turning red. At the national level, one sees a slight intensification of macro-regional patterns, with the South trending a bit more Republican and the North trending a bit more Democratic. At the local level of analysis, however, a number of exceptions to this pattern can be seen. To make such differences more immediately apparent, I have divided the United States into three parts, juxtaposing the 2000 map of each region against that of 2012.

The paired maps of the eastern third of the country clearly show a Democratic advance in New England and adjacent areas in New York. A number of counties in northern and central Virginia also turned blue, as did a few in central North Carolina. South Carolina also has a few more blue counties, generally the result of enhanced turnout in the Black community. Northern Ohio also tended in the Democratic direction, although the opposite tendency occurred in southeastern Ohio, which is part of the central Appalachian region that has exhibited the most pronounced red shift. Similar Republican gains were made across Tennessee and in northern Alabama. A movement to the Republican side is also evident in rural counties in the northern reaches of Michigan’s lower peninsula, while a few urban counties in Indiana contrastingly moved in the Democratic direction.

The central third of the U.S. exhibits the strongest pattern of regional differentiation, with the southern half of the area trending Republican and the northern half trending Democratic. Illinois is itself spilt by this divide; note that the northern half of the state, and especially the Chicago suburbs, shifted blue between 2000 and 2012, whereas the southern half shows a red shift. The strongest move to the Republican Party on this map occurred in Oklahoma and Arkansas. In the northern Great Plains, in contrast, a handful of rural, predominantly White counties moved into the Democratic camp. Texas, however, deviates from this north/south pattern, exhibiting instead a rural/urban divide. Note that the few rural counties in Texas that voted for Gore in 2000 (outside of the heavily Hispanic south) supported Romney in 2012. Yet over the same period, the urban counties of Bexar (San Antonio), Travis (Austin), Harris (Houston), and Dallas (Dallas) all moved in the opposite direction, although only Travis gave a substantial margin of its votes to Obama.

In the western third of the U.S., the shift at the county level from 2000 to 2012 was entirely in the Democratic direction, although most of the region remains solidly Republican. In the non-metropolitan parts of the region, most blue-trending counties can be explained by such factors as migration by people from blue states seeking natural amenities (Teton County, Wyoming), or the presence of universities with large student populations (Missoula County, Montana). The two rural counties in north-central Montana that have turned blue (Blaine and Hill) both have substantial American Indian populations.

Andrew Sullivan and several other commentators have argued that the Republican Party is increasingly keyed to the American South, and in particular to the old Confederacy. The map analysis provided above indicates a tendency in this direction, but only a relatively slight one. As Karen Cox recently argued in the New York Times, the conservatism of the South is often exaggerated, as many of its urban counties continue to support Democratic candidates—although it is notable that only ten percent of Whites in Alabama voted for Obama. At the same time, in the North many counties remain staunchly Republican. In central Pennsylvania, only semi-metropolitan Dauphin County (Harrisburg) gave a majority of its votes to Obama; most counties in this region overwhelmingly supported Mitt Romney. Even Centre County, home of the massive Pennsylvania State University, is colored red, albeit in the lightest shade of the color possible (Romney took the country by twenty votes, out of some 67,000 cast).

Owning to its conservative nature, central Pennsylvania is sometimes referred to as “Pennsyltucky,” which the Wikipedia tells us is a “portmanteau constructed from “Pennsylvania” and “Kentucky”, implying a similarity between the rural parts of the two states. It can be used in either a pejorative or an affectionate sense.” In 1992, Democratic strategist James Carville similarly referred to central Pennsylvania as “Alabama without the Blacks.” A 2008 blog posting by Brian Schaffner in the Center for Congressional and Presidential Studies, however, disputes this characterization, arguing that the region actually has more cultural features in common with the rest of Pennsylvania than it does with Alabama. Surveys show, for example, that many more central Pennsylvanians regard political comedian Jon Stewart favorably than do residents of Alabama, just as many more Alabamans than central Pennsylvanians shop at Walmart. I suspect, however, that voting behavior makes a better yardstick in this regard than shopping patterns. Still, Schaffner’s analysis is intriguing.

(“Pennsyltucky”  Image credit here)


Iowa, Minnesota, and the Anomalous Zone on the U.S. Electoral Map

In recent U.S. presidential elections, rural counties have tended to vote heavily for the Republican candidates. As a result, most of the United States is shaded red on county-level electoral maps. Most of the low-population counties that do support Democratic candidates fall into one of several categories. In some cases the explanation is clearly demographic; the heavily African-American belt stretching from southeastern Virginia to western Mississippi, the heavily Hispanic areas of northern New Mexico and southern Texas, and the scattered Native American counties in the northern Great Plains and the Southwest have long been Democratic strongholds. Location also plays a role, as rural counties in New England and on the Pacific Coast that are closely connected with nearby metropolitan areas also tend to support Democrats. Most of the other seemingly anomalous blue rural counties can usually be explained by their specific characteristics, such as the presence of a major ski resort (Blaine County, Idaho), a local economy based (or recently based) on unionized mining (Silver Bow County, Montana), or the existence of a large college (Athens County, Ohio).

One large block of largely rural, Democratic-voting territory, however, does not fit into any of these categories. In what might be called the “Upper Mississippi River Valley Anomaly,” one finds a substantial cluster of mostly agricultural, deeply religious, heavily Caucasian “blue” counties. In this area focused on southwestern Wisconsin, southeastern Minnesota, eastern Iowa, and northwestern Illinois, almost all counties opted for Barack Obama. A few counties in western Minnesota and in the eastern Dakotas, located just outside this contiguous zone, also exhibit the same pattern.

The voting behavior in few of the counties in this zone can be explained by their particular characteristics. Dark-blue Dane county in Wisconsin, for example, is the site of the main campus of the state university, and the same is true in regard to Johnson County in Iowa. Strongly Democratic Mower County in Minnesota, in contrast, contains a massive meatpacking facility, and is therefore a labor stronghold. Most of the counties in this area, however, are typical farming communities, dominated by Whites and slowly losing population. Consider, for example, Mitchell County Iowa, population 10,776 (having dropped from 14,121 in 1940). As the Wikipedia article on the county notes, “Mitchell County was the whitest county in the country, at 99.27%, to vote for Barack Obama in the 2008 Presidential election. Obama garnered 56% of the vote, while John McCain received 44%.” Obama’s margin over Mitt Romney in 2012 was not so wide, but the President still took Mitchell County with room to spare.

The Upper Mississippi River Valley Anomaly dates back roughly a dozen years. In the early 1980s, the area was heavily Republican, overwhelmingly supporting Ronald Reagan. In the 1990s, many counties in the region turned to Democratic Bill Clinton, but so too did many other parts of the country with similar demographic and economic characteristics. By 2000, however, a distinct atypical zone had emerged in this region. As result, Iowa in particular has become a divided state, red in the west and blue in the east. The same spatial divide can be seen in primary elections; Republicans in western Iowa, for example, were less inclined to support Romney’s nomination than those in the east, favoring instead the more conservative candidates such as Rick Santorum and Rick Perry.

It is difficult to find correlations that might help explain this phenomenon. As the paired political and demographic maps of Iowa indicate, more of the blue and light-red counties in the state have larger cities and towns than the deep-red counties, but the linkage is not strong. Nor can the Iowa vote in general be explained on the basis of religion, as the state’s main religious divide separates the south from the north, not the east from the west.  (As can be seen in the maps, southern Iowa has lower rates of church attendance and a higher proportion of Methodists than the deeply religious, strongly Lutheran north.) Religion does seem to be a factor, however, in the strongly Republican northwestern corner of Iowa, where three counties are dominated by members of the historically Calvinist Reformed Church.

Similarly, factors that might help explain the existence of the larger Upper Mississippi River Valley Anomaly are difficult to locate. The agricultural foundations of the local economies do not seem to play a role, as the “anomalous zone” includes part of the Corn Belt and part of the Midwestern dairy belt. Issues of ethnic heritage also seem to be minor. The zone in question is heavily German, but so too are surrounding areas. Much the same can be said in regard to Norwegian background; many of the farming counties that have been voting for Democratic candidates in recent elections were heavily settled by Norwegians, but so too were a number of decidedly Republican counties in North Dakota.

Minnesota is also the key site of another intriguing anomaly in U.S. electoral geography. Over the past few decades, the state as a whole has been solidly blue, not opting for a Republican candidate since Richard Nixon’s landslide victory in 1972. Yet a number of Republican strategists thought that Romney had a good chance of taking Minnesota in 2012. In the end, Obama won the state by more than a seven percent margin, but that margin was still relatively thin when compared with those found in other solidly blue states.

To the extent that Minnesota might be regarded as a purple-trending swing state, the answer is located in suburbs. In the northeastern and Pacific Coast states, the suburbs have switched in the past three decades from Republican to Democratic bastions. As conservative historian Vincent Cannato notes, “The GOP needs to run better in affluent suburban counties. If the party of business and economic growth can’t do well among the most prosperous and economically vibrant parts of the nation, then something is seriously wrong.” But in Minnesota, and in neighboring Wisconsin as well, the suburbs remain red. In fact, they have moved in the Republican direction over the past few decades. Consider, for example, the paired maps of the greater Minneapolis/St. Paul metropolitan area posted here. In 1980, when Ronald Reagan took 489 out of 538 electoral votes nationally, Jimmy Carter took almost the entire metro area. In 2012, however, Obama lost most of the suburban counties, while easily wining the mostly urban counties of Hennepin (Minneapolis) and Ramsey (St. Paul). The northern suburbs of Minneapolis and St. Paul are represented in Congress by Michele Bachmann, one of the most conservative members of the House of Representatives. Bachmann won reelection this year, although narrowly.

When examined over the past three decades, Minnesota thus exhibits dual anomalies, with affluent suburban counties trending Republican and the rural, farming counties in the southeast (and to some extent in the far west) trending Democratic. (The mining and logging counties of the northeast have remained in the Democratic camp over this period.) Such patterns are difficult to explain. If any readers have any ideas, I would love to see them.


The Geography of the 2012 Illinois Republican Primary

The geographical patterns in the recent Republican presidential primary in Illinois are quite clear. As can be seen by comparing the two maps, Mitt Romney triumphed in most urban and suburban parts of the state, doing particularly well in the Chicago metropolitan area, whereas Rick Santorum did very well in rural counties, particularly in the southeastern part of the state, an area with a somewhat southern cultural background. Exceptions to this pattern include Rock Island County in the northwest, Madison County in the center-west, and Tazewell County to the south of Peoria, which are all relatively urban or suburban, yet voted for Santorum.