Many a postmortem of Mitt Romneys defeat has focused on his poor showing among Hispanics and argued that Republicans wont do better in national elections until they find a way to appeal to this growing voting bloc, especially by modifying the partys stance on immigration. Yet these analyses often rely on faulty data that overstate the impact that Hispanic voters have on elections. They also typically ignore the fact that, as exit polls show, Latinos are almost certainly voting, like everyone else, on major issues—especially the economy—not on narrow ethnic lines. Latinos are just like other voters, history suggests: theyre more likely to vote for Republicans when the party puts forward a good candidate with broad appeal.
Most of the analyses that Ive read begin by noting the rapid growth of Americas Hispanic population. But one-third of adult Hispanics are not U.S. citizens and consequently cant vote. Even Latinos who are citizens dont vote as reliably as whites or blacks do, and as a result, their population growth rate doesnt translate into commensurate voting power. According to U.S. Census data for the 2010 midterm elections (the most recent national data available), adult Hispanics numbered 32.5 million in the U.S. population, but only 10.9 million were registered to vote and only 6.6 million actually voted (up from 5.6 million in the 2006 midterms). By contrast, of the 155.5 million adult white residents in the United States in 2010, 104 million were registered to vote and 74.3 million did vote. In other words, nearly half of the countrys adult whites participated in the 2010 elections; only 20 percent of adult Latinos did.
Sure, as Latinos become more assimilated into American society, their participation rates may increase—but their voting patterns will probably change as well. One recent analysis warns that Latinos share of the population by 2050 will be so large as to permanently damage Republicans prospects. Such scenarios, however, assume a static electorate that, in 40 years, votes the same way it does today. If in 1940, say, I had constructed a similar chart projecting the growth rate of the countrys Italian-American population, based on its having a higher birthrate than that of the Anglo-American population, I could have issued the same warning to Republicans. Americans of Italian descent were voting heavily Democratic back then. By 1980, they had become a key component of the Reagan coalition.
The argument that Latinos are moving away from Republican candidates relies on the contrast between the number of Hispanic votes that successful GOP candidates like Ronald Reagan and George W. Bush received and the less impressive results of John McCain in 2008 and Mitt Romney this year. On the surface, that argument seems compelling, but it ignores a couple of important points. For one thing, it generally overstates Bushs numbers among Latino voters, thanks to some bad exit polling in 2004 long since discredited. Today, the media persist in claiming that Bush received somewhere between 41 percent and 45 percent of the Latino vote against John Kerry, a sharp contrast with the 27 percent that Romney captured. Yet subsequent academic analyses of the 2004 elections estimate that Bush actually received about 35 percent of the Latino vote. That wasnt a historic high, as Bush political advisor Karl Rove still claims; rather, it was in line with what Reagan, another victorious Republican, had achieved.
Also, theres the stubborn fact that McCain and Romney did worse than Reagan and Bush among many other demographic groups, including traditionally strong Republican ones. Romney, for instance, won the vote among those who say they attend religious services regularly, but not by as wide a margin as Bush did. Once we have better data, the larger issue in the 2012 election may turn out to be a sharp decline in white voters, which cannot be explained merely by demographic shifts. Its possible that 5 million or more eligible whites didnt vote, perhaps because of a lack of enthusiasm for either candidate.
Whats more likely than race to account for Hispanic voting trends is income, a decisive factor in this election. The Obama campaign did a good job of portraying Romney as a Wall Street multimillionaire whose policies would favor the rich. Despite some conservatives belief that the Republican Party is capturing blue-collar America, Romney lost decisively among lower-income voters, who continue to vote Democratic in large numbers. Hispanic households fit into this demographic group: on average, their incomes are about 35 percent lower than the national average. Even more to the point is that Romney did terribly among voters who earned less than $50,000 a year, capturing just 38 percent of their votes—and over 60 percent of Hispanic households fit that income profile.
The media like to focus on race and ethnicity in voting patterns, because those factors seem to demand great changes in parties political strategies. Since your race or ethnic background doesnt change, the thinking goes, its the party that must change to reach out to you. That this emphasis on race is often mistaken doesnt mean that Republicans dont need to work with Democrats on, for example, comprehensive immigration reform that fixes our broken system. Todays status quo—more like a stalemate—isnt good for anyone.
But in most cases, income is a far better determinant of voting patterns than race is (blacks are an exception, for historical reasons). The voting of ethnic groups evolves significantly as their incomes change. The ancestors of millions of todays ethnic voters came to America in the great immigration wave of the early twentieth century and voted reliably Democratic for generations. Over the last 30 years or so, their descendants voting allegiances shifted significantly. Many were first attracted to the Republican Party by an optimistic presidential candidate who campaigned on a convincing pro-growth agenda. That won over voters in 1980; it would do so today, too.