Rise in risk inequality helps explain polarized US voters

The growth in partisanship has to do with people's current income and -- importantly -- their expectations of job security, said Philipp Rehm, author of the study and assistant professor of political science at Ohio State University.

At one time, many voters were "cross-pressured" -- when looking at what they earned now and their risks of losing that income, they felt torn between Republican and Democratic policies. The result is that they were natural independents, Rehm said.

But since the 1970s, a growing number of workers have found their current incomes and beliefs about their job security have converged -- in other words, their preferences aligned completely behind either Democratic or Republican policies. Rehm calls these people natural partisans.

In a study published recently in the British Journal of Political Science, Rehm estimated that slightly more than half of Americans could be counted as natural partisans in 1968, based on their income and job security. But by 2008, the number of natural partisans had climbed to include nearly two-thirds of all Americans.

"Americans who feel cross-pressured -- liberal Republicans and conservative Democrats -- have faded in number and importance, leaving only liberal Democrats and conservative Republicans," Rehm said.

"This can help explain why politics has polarized so much over the last four decades."

Rehm said that, traditionally, many political scientists have thought that Americans' income played a large role in which political party they supported.

People with low incomes would support the Democratic Party, which favors redistributional policies -- in other words, higher taxes on the rich which could be used to help the less wealthy.

Wealthier individuals would support the Republican Party, based on its opposition to redistributional policies.

But Rehm argues that Americans' political preferences are shaped not just by their current income, but also by the risk they perceive that they could lose their current income level.

"Income is not something that is fixed for life. The more people believe they might lose their jobs, the more they would be expected to favor Democratic policies," Rehm said.

So someone who has a high income today might be expected to be a Republican, based just on that income. But if he or she also fears losing that job and income, they might also support Democratic policies. That individual would be cross-pressured, Rehm said.

The problem is that the number of these cross-pressured Americans has dropped precipitously, according to Rehm's research.

In one analysis, Rehm used data from the General Social Surveys, which have interviewed a broad cross-section of Americans from 1972 to 2006. He looked at respondents' current income and occupation. To examine their risk of losing their current income, he measured the unemployment rates for each respondent's current occupation.

Rehm also used data from the study that indicated the respondents' party identification and whether they supported redistributional policies.

Results showed that -- as Rehm hypothesized -- respondents' current income and employment risk predicted whether they would support redistributional policies and, as a result, which political party they identified with.

The next step was to see, on a national level, how Americans' income and employment risk has changed over time.

For this analysis, Rehm used data from the 1968-2008 March Current Population Surveys. This data set gives information on the wages and unemployment rates for all the major occupations over that period.

The results of the analysis showed that, over this period, income and employment risk became increasingly connected. People with low incomes increasingly had higher risk of losing their jobs. And, on the other hand, people with higher incomes increasingly had safer jobs.

The result is that more people are "natural" Democrats -- meaning they have low income and high job risk -- or "natural" Republicans -- high income and low risk. There are fewer cross-pressured people who would naturally be independents.

While many observers have talked about the growing income inequality in the United States, Rehm said this study shows that there is also growing risk inequality in the country.

"The poor are not only becoming poorer, but they are also becoming more exposed to risks to their income. And the rich are not only becoming richer, but are also less exposed to risks," Rehm said.

Rehm said he is not arguing that non-economic issues, like abortion or foreign policy, don't also affect people's choice of party identification. But he said economic issues do play a powerful role in influencing party identification, as the results of this study show.

He also notes that this study was done before the current recession and economic crisis. It's too early to say how that crisis may affect the results of this study, but it obvious that more people than ever feel at economic risk, he said.

Overall, the story remains basically the same.

"Changes in the labor market have increasingly sorted Americans along party lines," he said. "Fewer people feel the cross pressures that would create independent voters."