Contrary to the views of those who see polarization as almost entirely an elite phenomenon, the deep divide between the parties in Washington and in many state capitols is largely due to the fact that Democratic and Republican elected officials represent electoral coalitions that differ sharply in their social characteristics and political orientations. The roots of polarization are in our changing society—and above all the growing racial and ethnic diversity of the American population.
Due to large-scale immigration from Latin America and Asia and higher fertility rates among nonwhites, the racial and ethnic makeup of the American population has undergone a major transformation since the 1960s. Nonwhites comprise a growing share of the overall population and of eligible voters. This demographic shift has had very different effects on the two major parties, however.
The impact of growing racial and ethnic diversity on the American electorate and on the composition of the Democratic and Republican electoral coalitions is readily evident in the figure above. Not only has the nonwhite share of voters in presidential elections quintupled since the 1950s — and more than doubled since the 1990s — but the racial divide between the Democratic and Republican electoral coalitions has widened dramatically.
Before the 1965 Voting Rights Act resulted in the rapid enfranchisement of African American voters in the southern states, both parties’ electoral coalitions were overwhelmingly white. Today, the Republican electoral coalition remains overwhelmingly white. Nonwhites made up only 10 percent of Romney voters according to the 2012 national exit poll. But the nonwhite share of Democratic voters has increased fairly steadily since the 1960s and that trend has accelerated since 1992. Nonwhites comprised 45 percent of all Obama voters in 2012, and a majority of Obama voters under age 40.
The political significance of increasing racial diversity reflects the reality that, despite much progress in race relations over the past half century, American society remains deeply divided along racial lines. In many ways, the United States is still a segregated and unequal society. African Americans and Latinos continue to experience significantly worse health outcomes, poorer educational and job opportunities, inferior housing, higher unemployment and lower incomes than white Americans. They are also much more likely to encounter hostility and prejudice in their interactions with public and private bureaucracies. These differences in life experiences and opportunities are reflected in sharply differing views on issues such as taxation, spending on social services and the proper role of government — as well as major differences in party identification and voting behavior.
The growing dependence of Democratic candidates and office-holders on nonwhite voters, along with a Republican strategy of appealing to white voters unhappy with the Democratic Party’s racial and economic liberalism, has contributed to an ideological and regional realignment within the white electorate. Conservative whites in the South and elsewhere have moved increasingly into the Republican camp, while moderate-to-liberal whites in the Northeast, Midwest, and Pacific states have moved increasingly into the Democratic camp.
There is every reason to expect these trends will continue. Census data indicate that the nonwhite share of newly eligible voters will continue to grow for many years. Yet despite the threat that this trend poses to the future viability of the Republican Party in national elections, the influence of the ultra-conservative, anti-immigration tea party movement makes it unlikely that the GOP will be able to successfully appeal to this growing nonwhite electorate. As a result, the racial divide between the parties’ electoral coalitions is likely to increase over the next several election cycles.
Race has certainly not been the only factor behind rising partisan polarization. Another crucial component of the ideological realignment of the past thirty years has been a growing religious divide between the parties. This is not the traditional divide between Protestants and Catholics that dominated American politics in much of the country during the first half of the 20th century. Instead, it is a divide between the religiously observant and non-observant.
In addition to appealing to white voters unhappy with the racial liberalism of the Democratic Party, Republican leaders starting with Ronald Reagan sought to appeal to evangelicals and other religious conservatives within the white electorate who were unhappy about the Supreme Court’s Roe v. Wade decision legalizing abortion and other perceived threats to traditional values. The result may not have been a “culture war,” but it certainly was a widening partisan divide between religious and non-religious white voters, as the figure below shows.
The movement of religiously observant white voters into the Republican camp accelerated after the GOP made circumventing Roe v. Wade a key plank of its national party platform in 1980. And the partisan divide between observant and non-observant white voters continued to widen in the 1990s and 2000s as gay rights and same sex marriage became salient issues.
By 2012, 69 percent of white voters who reported attending religious services at least once per week identified with the Republican Party compared with only 41 percent of white voters who reported rarely or never attending religious services—the largest divide ever recorded. Some 75 percent of religiously observant whites voted for Mitt Romney in 2012 compared with only 46 percent of non-observant whites.
Much attention has been paid by political commentators to the potential impact of growing income inequality on partisan polarization. When it comes to explaining partisan polarization in the contemporary American electorate, however, income differences appear to play a much smaller role than either race or religion. According to data from the American National Election Study, the correlation between family income and party identification among all voters in 2012 was a very modest 0.13. As family income goes up, voters are a bit more likely to back the Republicans, but not strongly so. The correlation among white voters was meager and statistically insignificant 0.03.
As the figure below shows, at every level of family income, religiously observant whites were much more likely to vote for the Republican presidential candidate in 2012 than non-observant whites. In fact, non-observant, upper-income whites were much more likely to vote Democratic than religiously observant lower-income whites. Indeed, family income had no relationship at all with vote choice among religiously observant whites. A remarkable 80 percent of observant whites with family incomes below $15,000 voted for a Republican presidential candidate who famously expressed no interest in appealing to 47 percent of the electorate including, presumably, those living below the poverty line.
The 2012 election results once again revealed the existence of an electorate deeply divided by race, religion and ideology. Those divisions are found among general election as well as primary voters. The Democratic and Republican electoral coalitions are more distinctive now in terms of social characteristics and political outlook than at any time in the past sixty years. There is no disconnect between representatives and represented — today’s elite polarization is not imposed on a centrist electorate.
The deep partisan divide in Washington clearly reflects a deep partisan divide within the American electorate, and for this reason, it is unlikely to diminish any time soon.