Published October 17, 2017
As sure as the sun had risen in the east, upper-income, educated, private-sector-employed whites had for decades voted Republican. Even the growing importance of social issues and rising secularism hadn’t changed that fact: Republican presidential candidates still carried high-income and college-educated whites, even if their margins were lower than in prior decades.
All that changed in 2016.
Upper-middle-income, college-educated whites are increasingly unhappy with a Republican party dominated by President Trump and his agenda. Unless the party recognizes the threat and tries to keep this group under the Republican umbrella, it could find both its House majority and its dominance in statehouses at risk in 2018.
Democrats have been making inroads among these voters for quite some time. Democratic strategists Ruy Teixeira and John Judis noted this development in their important 2002 book, The Emerging Democratic Majority. They showed that Democratic candidates were winning college-educated whites who worked in the idea- and word-based industries such as publishing, marketing, computers, and software. College-educated whites in these private-sector industries tended to be more secular and less concerned about economic issues in politics than other whites. So long as Democrats remained open to private-sector growth, these people would vote Democratic based on their stances on social issues and their antagonism toward the southern-Evangelical-tinged tone of the modern GOP.
Noted tech hubs such as Boston, Austin, and Silicon Valley were not the only college-white-dominated regions to become more Democratic. Northern and western suburbs such as those found around Philadelphia, Washington, Los Angeles, and Chicago also moved noticeably to the left after 1992. While George H. W. Bush carried every suburban county in these regions by large margins in 1988, his son, George W. Bush, lost or carried them only narrowly in 2004. By 2008, the transformation was complete: John McCain lost every one of these counties except California’s Orange County.
But these trends did not hold nationwide. Other high-income, educated, white suburbs remained GOP bastions. Those around Milwaukee, Dallas, Houston, Atlanta, and Cincinnati were at least as Republican in 2004 as they had been in 1988. Communities around Jacksonville, along Florida’s west coast, and near Phoenix stuck with the GOP even as the party’s victory margins slipped nationally. Clearly not all upper-middle-income, college-educated, white voters were deserting the GOP.
That ceased to be true in 2016. Trump’s surprise victories in Michigan, Pennsylvania, and Wisconsin masked the fact that he lost tens of thousands of Romney voters in upper-middle-income places such as Detroit’s Oakland County, Philadelphia’s Chester County, and Milwaukee’s Ozaukee County. Trump lost heavily in the suburbs that had been slipping away from the GOP for decades, and he was the first GOP nominee to lose Orange County since 1936. He also lost areas around cities such as Dallas, Houston, and Atlanta that had given Romney over 60 percent of the vote just four years before.
Fortunately, Trump’s problems did not hurt other Republicans running down-ballot. Ron Johnson and Pat Toomey won close Senate races largely because they performed well in higher-income towns. House Republicans targeted by national Democrats for defeat, such as Virginia’s Barbara Comstock, California’s Darrell Issa, and Colorado’s Mike Coffman, held off their challengers even as Hillary Clinton handily won their districts. State legislators running in similar areas also tended to hold on as Republicans who deserted Trump came back in other races.
But 2017 has been a different story. According to one analysis, Democrats have run ahead of Clinton’s share of the vote in 28 of 35 special elections so far. These have not been trivial differences; Democratic candidates have run an average of 13 points ahead of Clinton’s total. This has allowed them to poach six of those 35 seats from Republicans, while the GOP has flipped none in return.
Even the most nationally prominent special election this year, the Jon Ossoff–Karen Handel race in Georgia, sent a shot across the GOP’s bow. This House district, formerly represented by Tom Price (who resigned his seat to join the Trump administration), encompasses the wealthiest and best-educated regions in Atlanta and its nearby suburbs. In 2012 it gave Mitt Romney 61 percent of its vote — but in 2016 Trump carried it by just 1.5 points, 48.3 percent to 46.8 percent, indicating that many Romney voters didn’t just go third-party but actually voted for Clinton. In a race whose turnout nearly matched the presidential election’s, Handel’s 51.9 percent of the vote suggested that she kept Trump’s voters and won back some defectors. But the more ominous statistic is that Ossoff ran slightly ahead of Clinton. This likely means he kept the Romney-Clinton voters, the sort who just eight months earlier had helped keep Republicans in control of the House and Senate by coming back to the party.
A continuation of this trend could spell disaster for Republicans in 2018. House Republicans hold a majority of 24 seats, but 23 of them represent districts that Clinton carried. If Romney-Clinton voters stay Democratic, as they apparently did in the Georgia special election, then the vast majority of these seats could be lost. These voters are also crucial to Republican hopes of winning back the Virginia governorship this fall and retaining statehouses in swing states such as Michigan, Ohio, Florida, and Wisconsin. If Republicans lose upper-middle-income voters, their hopes in these states will rest even more on the Obama-Trump voters whose habitual loyalty is to the Democrats.
It’s not hard to determine why upper-middle-income whites are now in play. Data from the Voter Study Group’s December 2016 national poll showed that Romney voters who switched to Clinton were much closer to Democratic positions on hot-button issues such as immigration, feelings toward Muslims, and trade. Indeed, the Romney-Clinton voter was more liberal than other Romney voters on every issue examined by the New America Foundation’s Lee Drutman. These voters may not have been partisan Republicans as conservatives understand the term, but they were crucial to the GOP voter coalition.
Data from the Voter Study Group’s July 2017 poll suggests these voters are steadily becoming partisan Democrats. Romney-Clinton voters despise Trump; 87 percent disapprove of him while only 2 percent approve. And this feeling is spreading down the ballot: About half of Romney-Clinton voters say they will vote Democratic for Congress in the 2018 midterms. Only about one-third say they will vote Republican; a further one in six are unsure.
Upper-middle-income areas also saw many Romney voters who did not back Trump but instead voted for a third-party candidate, usually Gary Johnson or Evan McMullin. The 2017 Voter Study Group survey shows that these voters approve of Trump by only a 49–40 ratio. As we saw in Georgia, they are much likelier to return to the party in congressional races: The same survey found that over 60 percent said they would vote GOP in the midterms. But tellingly, over a third said they were unsure which way they would vote. Between them and the Romney-Clinton voters, it is clear that the upper middle class is now up for grabs.
Conservatives cannot be oblivious to this. Republicans simply cannot have a national majority unless a large share of these voters returns to the fold. More important, there is little likelihood of majorities in key swing states without their votes.
They say that if you’re in a hole, stop digging, so the first item on the GOP agenda should be to stop doing unnecessary things to drive these voters away. That means the soon-to-be-released tax bill needs to do something tangible for them. Unfortunately, it will likely make them worse off.
The problem comes from the way the GOP wants to change itemized deductions. The administration and the Republican congressional leadership need to close some loopholes to finance the large corporate-tax-rate reduction they seek. So far, their target is the federal deduction for taxes paid to state and local governments, which mainly helps people making $100,000 or more. Republican proposals eliminate the deduction entirely but offset the loss by nearly doubling the standard deduction. This would mean a heavier tax burden for many who itemize their deductions and a lower burden for many who don’t.
Those with very high itemized deductions wouldn’t benefit from the higher standard deduction at all, and would simply lose the entire value of the state-and-local-tax deduction. Other taxpayers would see their itemized deductions fall low enough that they’d just take the standard deduction instead, as they don’t give enough to charity or pay enough in mortgage interest to make itemizing worthwhile under the proposed system. Both sets of taxpayers would have more of their income subject to taxation than they do now.
Conservatives have been told that all this will mainly hurt voters in high-tax blue states, but this just isn’t so. County-level data from the IRS show that upper-middle-income suburbanites in key swing states are very likely to see tax hikes because of these changes. Taxpayers who annually earn between $100,000 and $200,000 in counties with large numbers of upper-middle-income voters deducted an average of around $25,000 in 2014, the last year for which data are available. Taxpayers earning more than $200,000 deducted an average of more than $50,000 a year.
Without the state-and-local-tax deduction, a lot more of these taxpayers’ income will become taxable, and most of them are in the 25 percent bracket. The GOP tax reform would give these individuals a rate cut, but it would not be enough to make up the difference. Anyone currently deducting more than $26,000 would be at risk of seeing his taxes rise because of the GOP tax bill.
Such people vote in very large numbers in the key swing states of Florida, North Carolina, Pennsylvania, and Wisconsin. The 2016 exit polls showed that $200,000-plus voters were a large share of the vote, ranging from 7 percent in North Carolina to 10 percent in Pennsylvania. Add in voters making between $150,000 and $200,000 and it’s clear that 10 to 15 percent of voters in these four states could see tax hikes under the GOP’s tax-reform proposals.
Republicans are already having too much trouble retaining these voters to make such a careless error. These voters already are leery of Trump. If the GOP hikes their taxes, their anger will spread to Trump’s party.
There are some easy, if expensive, alternatives. Instead of keeping the mortgage-interest and charitable deductions itemized, the Republican tax reform should make them available to all taxpayers. Doing this — making them what tax pros call “above-the-line deductions” — would allow upper-middle-income taxpayers to get the benefit of the doubled standard deduction and deduct their other expenses.
Republicans could even adopt a populist tone to advocate limiting the value of the mortgage-interest deduction to, say, $500,000. Such a limit would not harm many upper-middle-income taxpayers, but it would cause much wealthier taxpayers to pay more. Since they would also have their rates cut substantially, this would likely only reduce the amount of their tax cut. Removing the ability to deduct mortgage interest for a second home would also largely hit the highest earners, bringing back more revenue to finance corporate-rate cuts without burdening upper-middle-income voters.
Recovering these voters’ trust will take time and effort. The current tax-reform proposals will drive them away at exactly the time their loyalties are being tested. Better not to risk a backlash next November.
– Mr. Olsen is a senior fellow at the Ethics and Public Policy Center. He is the author of The Working Class Republican: Ronald Reagan and the Return of Blue-Collar Conservatism.