by Bill Bishop
This pattern appeared again and again as we evaluated other demographic measures.
Education. In 1970, the county groups were well balanced in the proportion of the population that had a college degree. After that, the percentage of college-educated people increased in every group—but the well-educated were especially attracted to Democratic counties (see Figure 2.3). People with college degrees increased the most in the Democratic landslide counties, where 29 percent of the adult population had at least a college degree in 2000. In the Republican landslide counties, 20 percent of those over twenty-five years of age had a bachelor's degree or higher in 2000. According to the 2000 census, in seventeen states (including the District of Columbia), the proportion of the population with an advanced degree was higher than the national average. In 2004, John Kerry won thirteen of those states, or 76.5 percent. In thirty-four states, the proportion of people with a postgraduate degree was lower than the national average. George Bush won twenty-seven of those states, or 79.4 percent.
Figure 2.2 The Separation of American Communities
Political divisions found in the 2004 election
have been growing for decades.
Percent Vote for the Republican Presidential Candidate as a Deviation from the Mean Vote
Note: Republican landslide counties: George W. Bush won by 20 percentage points or more in the 2004 presidential election. Democratic landslide counties: John Kerry won by 20 percentage points or more. Republican competitive counties: Bush won by less than 20 percentage points. Democratic competitive counties: Kerry won by less than 20 percentage points. 0 (zero) represents the average Republican presidential vote for the entire United States. Dropping below the 0% line means the county group is voting more Democratic than the nation as a whole. Above means the county group is more Republican.
Source: Dave Leip's Atlas of U.S. Presidential Elections, http://www.uselectionatlas.org.
Religion. Church members seemed to be increasingly concentrated in Republican counties. The Glenmary Research Center collects data on the number of church members in regular surveys. According to this data, from 1971 to 2000, the number of church members increased 33.8 percent in Democratic landslide counties. In the same period, the number of church members jumped 54.4 percent in Republican landslide counties. From 1990 to 2000, Democratic counties lost churchgoers, while Republican counties continued gaining (see Figure 2.4). We even discovered a difference in migration patterns between counties with a high percentage of churchgoers and those that were more secular. Only 11 percent of the people who moved out of the counties with the most churchgoers moved to the most secular counties. The reverse was true, too: only 5 percent of those who left the most secular counties migrated to the counties with the highest percentage of churchgoers.
Figure 2.3 Separation by Education
Democratic landslide counties have been
gaining citizens with B.A. degrees.
Percent of County Population with B.A. Degree or Higher (25 years of age)
Sources: Dave Leip's Atlas of U.S. Presidential Elections, http://www.uselectionatlas.org; U.S. Census of Population, http://www.census.gov.
Immigrants. The percentage of the U.S. population that was foreign-born increased in every group of counties, but those born outside the United States favored Democratic counties. By 2000, 21 percent of the population in Democratic landslide counties was foreign-born, compared to just 5 percent in Republican landslide counties (see Figure 2.5).
Figure 2.4 The Separation of Churchgoers
Republican landslide counties have gained the
most church members in the past fifty years.
Number of Church Members by 2004 Presidential Election Margin (in millions)
Sources: Dave Leip's Atlas of U.S. Presidential Elections, http://www.uselectionatlas.org; Glenmary Research Center, Religious Congregations & Membership in the United States, http://glenmary.org/GRC/grc_shopping.htm.
Race. In 1970, each of the four county groups was home to about a quarter of the nation's white population—that is, whites were distributed evenly throughout the groups. (Republican landslide counties actually had a slightly smaller percentage of the total white population than did Democratic landslide counties.) Over the next thirty years, however, whites became more concentrated in Republican counties. Democratic counties—especially Democratic landslide counties—lost shares of white population. By the time of the 2000 census, only 18 percent of the nation's white population lived in Democratic landslide counties. By contrast, in 2000, 30 percent of America's white population lived in counties that provided Republican landslide margins in the 2004 presidential election (see Figure 2.6). The real "white flight" of the past two generations has been whites moving to communities that were becoming staunchly Republican.
Figure 2.5 The Immigration Divide
Foreign-born citizens move to Democratic counties.
Percent of County Population That Is Foreign-Born
Sources: Dave Leip's Atlas of U.S. Presidential Elections, http://www.uselectionatlas.org; U.S. Census of Population, http://www.census.gov.
Figure 2.6 The New White Flight
Whites have increasingly clustered in the counties
that voted Republican in 2004.
Share of U.S. White Population
Sources: Dave Leip's Atlas of U.S. Presidential Elections, http://www.uselectionatlas.org; U.S. Census of Population, http://www.census.gov.
What About the Future?
From 1980 to 2006, Republican counties gained about 50 million people, and Democratic counties gained about 22 million people (see Figure 2.7). That means that Republican counties grew by 1 million more people a year than Democratic counties. And projections from the U.S. Census Bureau show that this trend will continue—will even accelerate—in the current century.
Birthrates are higher in Republican areas than in Democratic areas. This phenomenon has been described as the "liberal baby bust" by USA Today. In 2004, New York Times columnist David Brooks wrote that the higher birthrates in Republican areas were part of a "natalism" movement. "They are having three, four or more kids," Brooks wrote of America's "natalists." "Their personal identity is defined by parenthood. They are more spiritually, emotionally and physically invested in their homes than in any other sphere of life, having concluded that parenthood is the most enriching and elevating thing they can do. Very often they have sacrificed pleasures like sophisticated movies, restaurant dining and foreign travel, let alone competitive careers and disposable income, for the sake of their parental calling." Some observers of the trend Brooks described predicted that the baby advantage in Republican areas would lead inevitably to larger Republican majorities.6
It's a plausible theory, but to this point, it isn't the primary reason red counties are gaining population faster than blue ones. People are born, it's true, but they also die. And it happens that death rates in Republican counties are also higher than in Democratic counties. When deaths are included in the calculation, it turns out that the natural increases in population—births minus deaths—account for very little of the growing difference in population between Republican and Democratic counties. Republican counties gained about 28 million more people than Democratic counties between 1980 and 2006, but only 2.9 million of that increase was due to natural increases in population. (Immigrants aren't included in these calculations.)
Figure 2.7 The Republican Population Shift
The greatest population increases have taken place
in counties voting Republican in 2004.
Total Population by County Type (in millions)
Sources: Dave Leip's Atlas of U.S. Presidential Elections, http://www.uselectionatlas.org; U.S. Census of Population, U.S. Census estimates, http://www.census.gov.
Instead, almost all of the Republican county population jackpot was because of domestic migration. In absolute numbers, Republican counties were the winners in the Big Sort. In fact, we found that the Republican counties with the str
ongest majorities were the most attractive to those who moved. Meanwhile, from 1990 to 2006 alone, 13 million people moved from Democratic to Republican counties.
Many more people moved to Republican landslide counties than to Democratic landslide counties, but they were considerably poorer, earning on average only three-quarters of the income of migrants to Democratic landslide counties. In 2003, the individual income of people moving from another state into a Democratic landslide county averaged $30,492, according to IRS figures. Those moving from another state to a Republican landslide county had an average income of $22,939. And the more Republican the county, the poorer the migrant. Those moving to competitive Republican counties earned on average $25,120, almost $2,200 more than migrants to landslide counties.
Migrants with the highest incomes were those moving from a Democratic landslide county in one state to a Democratic landslide county in another state. Their incomes were 37 percent higher than the national average. Migrants with the lowest incomes were those moving between Republican landslide counties within a state. They earned 30 percent less than the national average. There is simply no telling what the consequences will be of this kind of economic sorting.
The Colorado Twist
Seen in light of the Big Sort, some puzzling dimensions of American politics begin to make sense. For example, traditionally Republican Colorado has become increasingly Democratic over the years. In 2004, Kerry cut Bush's margin of victory there to half of what it had been in 2000. A Democrat won the U.S. Senate seat in 2004, and the party swept both houses of the state legislature. Democrat Bill Ritter won the governor's race in 2006 based in large part on his promotion of alternative energy. Ronald Brownstein of the Los Angeles Times, writing in 2006, found that the language of politics in the state had shifted. People were more concerned with government services than with low taxes or abortion. "The whole rhetoric has changed in the past four or five years," a Democrat in Denver told Brownstein.7
Colorado has become more Democratic overall—but not all of Colorado: some parts of the state are just as Republican as at any time in the past half century. Over the past two decades, however, people from other states have flowed into Colorado. When we tracked these migrants, we learned that the Colorado counties with the highest inflows of people from other states were also the counties where support for Democratic presidential candidates was growing. The counties least affected by migration from other states had grown slightly more Republican since the 1980s. In addition, these politically opposite parts of Colorado were attracting people from entirely different places. The people moving to the fast-growing counties around Denver were three times more likely to have come from "blue" counties outside Colorado than the people moving to the slower-growing (and heavily Republican) counties along the Kansas, Oklahoma, and Nebraska borders. The county that sent the most people to Colorado between 1981 and 2004 was deeply Democratic Los Angeles County, California.8
The migration of people from Democratic counties elsewhere in the United States was turning Colorado into a tightly contested state. But because the Big Sort works at the community level, although the state as a whole grew more politically mixed, the divisions between Republican and Democratic areas within the state widened. Colorado's political story in the coming years will be one of expanding cultural and political division between the fast-growing Democratic counties of Denver and Boulder and the increasingly Republican counties in other parts of the state. In that sense, Colorado is a microcosm of the nation, where governments are being called on to reconcile the demands of communities that have less and less in common. It's a chore made harder by the peculiar psychology that is the special property of like-minded groups.
3. THE PSYCHOLOGY OF THE TRIBE
Can two walk together, except they be agreed?
—AMOS 3.3
"ANTOINE KILLED A brownish snake, two feet long, in the house, at the foot of the staircase," Secretary of State John Quincy Adams wrote in his diary during the early days of Washington, D.C. Hogs nosed through trash discarded on the side of the road, and carcasses of animals putrefied in the stagnant water collecting in brickyard excavations. There were no lights and few roads that were little more than trampled-down cow paths. The city was hard to find. Abigail Adams lost her way on the trip from Baltimore back to the White House. She stumbled about Maryland for a few hours before hiring a vagrant "to extricate us out of our difficulty." Those who came to the newly built city were anxious to leave. During Washington's first three decades, nearly one in five U.S. senators resigned every two years. Better to give up public office than to live in a "cosmos of evil and immorality."1
Washington was, from its beginning, a politically segregated city. In his forty-year-old study, The Washington Community, historian James Sterling Young mapped three Washingtons, one created for each of the three branches of government. The nine members of the U.S. Supreme Court lived in the same house until 1845. Executive branch workers gathered in one section of the city, near the White House, while congressmen were bunched together nearer the Capitol. "Men whom the Constitution merely separated into different work groups separated themselves into different societies," wrote Young.2
Congressmen lived in boarding houses. They formed eating clubs around common tables, and they slept together, two to a room. Young tracked the membership of these new boarding house communities and found that the residential segregation that marked the entire city was repeated in the houses. Men from one state or region would board together, finding comfort in their similar cultural ties, political outlooks, and, no doubt, culinary proclivities. "Legislators had a decided aversion to sharing their mess table, their living quarters, and their leisure hours with colleagues from regions other than their own," Young wrote.3 Washington had been created as the common ground of the nation, an intentionally heterogeneous society consisting of men gathered from across the new country. Without plan or foresight, however, the city had been transformed into an archipelago of culturally homogeneous and politically insular fraternity houses.
The homogeneity of the boarding houses crisply reflected the country, where communities were isolated by rivers, mountain ranges, and vast distances. The cultural segregation in early America was enforced by the lack of mobility, whereas today it's the ease with which Americans are able to move that has created political segregation. Even though we know much more now about the psychological effects of living in like-minded groups, the founders understood the dangers of self-segregation in ways we do not, and they sought to temper those influences. The research on the psychology of groups began more than one hundred years after the nation was formed. In scores of experiments, social psychologists learned about the power of groups to shape opinion and snuff out dissent. But without the benefit of science, the founders made an instinctual decision to embrace difference. It's not at all clear now that even with all of our knowledge, we are willing or able to make the same choice.
"He Has Betrayed Those with Whom He Broke Bread"
The residential segregation of the early-nineteenth-century boarding houses extended to the Capitol. The floor of Congress was intended to be the place where men with real cultural and regional diversity could meet, mingle, and come to a national consensus. But as Young tracked the votes in the early meetings of Congress, he found that the boarding houses and eating clubs became voting blocks. "For members who lived together, took their meals together, and spent most of their leisure hours together also voted together with a very high degree of regularity," Young discovered.4 They were a nascent form of political parties—coalitions magnetized by regional interests and bound by residential solidarity. Young found that in three out of four House votes from 1807 to 1829, no more than one congressman would bolt from the boarding house or eating group caucuses. Members who lived in the same houses voted unanimously in just about half of the 116 roll call votes that Young analyzed over five sessions of Congress.5 Young even found evidence that boarding house companions sat together on the floor of Cong
ress. "They transformed a national institution into a series of sectional conclaves," he wrote.6
A member who voted against his tablemates risked political retribution and, worse, social exile. Representative Stephen Van Rensselaer of New York initially agreed with his eating group to support William Crawford when the 1824 presidential election had to be decided in the House of Representatives. Van Rensselaer later crossed the boarding house, however, and voted for the eventual winner, John Quincy Adams. One of Van Rensselaer's boarding house fellows wrote that he avoided Van Rensselaer on the House floor after the vote and refused to shake his hand when the tearful legislator asked for understanding back at the boarding house. "Other gentlemen of the mess" also shunned the legislator after the vote. "We let him continue with us, sit at the same table with us, but we do not speak to him," the man wrote. "He is beneath anything but contempt ... He has betrayed those with whom he broke bread."7
The block voting of the boarding houses thwarted compromise or even debate. President Thomas Jefferson observed that legislators "are not yet sufficiently aware of the necessity of accommodation & mutual sacrifice of opinion for conducting a numerous assembly." Legislators came to work "in a spirit of avowed misunderstanding, without the smallest wish to agree," Jefferson wrote.8 Far from achieving the ideal of deliberation and debate—a mixing of representatives sent by diverse communities—these legislators lived, and voted, in a segregated fashion.
