The labor economist Stephen Rose offered a more illuminating analysis in a 2006 essay. Rather than focus on income in any given year, an approach that treats graduate students as members of the non-working poor, Rose looked at households’ average income over a 15-year period. He found that the fraction of adults between the ages of 26 and 59 with a direct material interest in programs for the poor is no more than 23 percent, smaller than is commonly understood.
That fraction may help explain why white voters without college degrees — a definition of the white working class that has been embraced by many political scientists — have grown more likely to vote for Republicans, most markedly in the South, not coincidentally home to some of the country’s lowest-cost metropolitan areas. If we assume that these downscale Republicans don’t look to government for their economic security, at least until retirement, it is easy to understand why they feel free to vote according to their values. They may well resist tax increases designed to pay for programs for other people, especially those seen as undeserving.
Over time Republicans have become ever more reliant on these voters. Exit pollsters classify voters by the last level of formal schooling they received. In 1992, Republican congressional candidates did seven points better among college graduates than among voters who had attended college but not received a degree, and did twelve points better among college grads than among voters who had earned high-school degrees but not attended college at all. These were typical results for the period: The 1996 differentials were the same. But they are narrower now. In the 2010 elections, Republican congressional candidates did only five points better among college grads than among voters with some college education, and only six points better among college grads than among just-high-school grads. Republicans actually do better among white voters without college degrees than among white college graduates; in 2010 they won the white college vote by 19 points and the white non-college vote by 30.
All educational groups swung away from the Republicans between 2004 and 2008 and then back toward them in 2010. But the swings were a little stronger among voters with less formal schooling. Republican congressional candidates did three points better among holders of post-graduate degrees in 2010 than they had done in 2008. They did ten points better among college grads, nine points better among voters with some college, and nine points better among high-school graduates. Multiply these numbers by the percentage of the electorate represented by each group, and Republican gains among voters without college degrees accounted for 4 percent of all votes cast, while gains among voters with college degrees accounted for 3.6 percent. Since these groups were each roughly half the electorate, voters with less formal schooling accounted for a disproportionate share of Republican gains.
The data suggest that conservatives ought to focus more intently on devising an agenda that addresses the concerns of lower-middle-class voters. Formulating such an agenda would have three advantages.
First, it would help the Republicans to keep and build on their existing working-class vote, which, as we have seen, is especially likely to swing between the parties. Second, it would indirectly help Republicans win upper-middle-class votes. Bereft of a policy agenda that appeals to lower-middle-class voters, Republicans often seek their votes using a cultural message that sounds strident and anti-intellectual to college-educated voters. If Republicans are perceived to be dividing voters into two categories, “real Americans” and “latte-sippers,” voters who fall in the second group understandably recoil. Third, such an agenda might help Republicans make modest inroads among working-class Hispanic voters and even black voters — at least among the considerable number of these voters who make political decisions based on pocketbook concerns rather than identity politics.