Ever since the Republican triumph in November, Democrats have been casting around for rationalizations. One theory, espoused by President Obama, blamed the party's dramatic loss on the simple fact that too many Senate races had taken place in conservative states. "This is probably the worst possible group of states for Democrats since Dwight Eisenhower," Obama remarked on Election Day. Other analysts pointed to the "six-year itch"—which often condemns the party of a second-term president to defeat during the midterm elections. Still others chalked up the results to the fact that midterm elections (with their low turnout) inherently favor Republicans, while presidential elections (with their high turnout) inherently favor Democrats. "We have two separate Americas voting every two years," wrote Markos Moulitsas Zuniga, the founder of Daily Kos. "And Democrats can win easily with the one, and Republicans can win easily with the other."

Indeed, in the wake of the midterms, the Center for American Progress optimistically predicted that, if 2016 voting patterns resemble those from 2012, the rising number of voters of color "will not only make it easier for Democrats to win states that they previously won in 2012. These demographic changes are also creating an opportunity for Democrats to win back states they lost in 2012."

None of these observations are wrong, as far as they go. It is undeniable, for instance, that the current Democratic coalition does better in presidential years than it does in midterm elections. Yet all of these explanations have a common problem: They obscure the possibility that 2014 was not an isolated event but rather the latest manifestation of a resurgent Republican coalition.

American parties routinely go through periods of ascendancy, decline, and deadlock. From 1896 to 1930, the Republican Party reigned supreme; from 1932 to 1968, the New Deal Democrats dominated; following a period of deadlock, the Reagan Republicans held sway during the 1980s. After the parties exchanged the White House, Democrats appeared to take command of American politics in 2008. In that election, Obama and the Democrats won not only the White House but also large majorities in the Senate and House, plus a decided edge in governor's mansions and state legislatures.

At the time, some commentators, including me, hailed the onset of an enduring Democratic majority. And the arguments in defense of this view did seem to be backed by persuasive evidence. Obama and the Democrats appeared to have captured the youngest generation of voters, whereas Republicans were relying disproportionately on an aging coalition. The electorate's growing ethnic diversity also seemed likely to help the Democrats going forward.

These advantages remain partially in place for Democrats today, but they are being severely undermined by two trends that have emerged in the past few elections—one surprising, the other less so. The less surprising trend is that Democrats have continued to hemorrhage support among white working-class voters—a group that generally works in blue-collar and lower-income service jobs and that is roughly identifiable in exit polls as those whites who have not graduated from a four-year college. These voters, and particularly those well above the poverty line, began to shift toward the GOP decades ago, but in recent years that shift has become progressively more pronounced.

The more surprising trend is that Republicans are gaining dramatically among a group that had tilted toward Democrats in 2006 and 2008: Call them middle-class Americans. These are voters who generally work in what economist Stephen Rose has called "the office economy." In exit polling, they can roughly be identified as those who have college—but not postgraduate—degrees and those whose household incomes are between $50,000 and $100,000. (Obviously, the overlap here is imperfect, but there is a broad congruence between these polling categories.)

The defection of these voters—who, unlike the white working class, are a growing part of the electorate—is genuinely bad news for Democrats, and very good news indeed for Republicans. The question, of course, is whether it is going to continue. It's tough to say for sure, but I think there is a case to be made that it will.

American political parties are informal coalitions of interest groups, social and economic classes, nationalities, and regions. When a voting bloc shifts from one party to another—or when a voting bloc within a party changes in size—it can often spell the beginning or end of a party's dominance. The GOP's success in the 1980s, for instance, was driven in large part by the movement of white working-class voters out of the Democratic Party and into the Republican Party—in the North as well as the South. Meanwhile, the success of the Democrats in the 1990s and in 2006 and 2008 was based on the growth of the minority vote (from 13 percent of the electorate in 1992 to 26 percent in 2008); the continuing movement of women, particularly single women, into the Democratic column; and the support of professionals, who were once the most Republican of occupational groupings. (Professionals range from nurses and teachers to doctors and architects; in exit polls, they can be identified roughly as voters who possess postgraduate degrees.)

But the Democratic success of recent decades was not based only on shifts among minorities, women, and professionals. To win elections, Democrats have still needed between 36 and 40 percent nationally of the white working-class vote—which, in practice, meant totals in the twenties or even the teens in the South, and near-majorities in many Northern and Western states. At one time, unions had provided a link between many of these voters and the Democratic Party. This advantage started to dwindle in the 1970s as private-sector unions began to shrink. Nevertheless, on a promise of prosperity, Bill Clinton got about 40 percent of the white working-class vote in 1992 and 1996, and Obama got 40 percent in 2008.

Democrats also needed to hold their own among middle-class voters. Reagan and George W. Bush handily won this demographic, but Clinton won a plurality of these voters in 1992 and 1996, and Obama won voters with college (but not postgrad) degrees by 50 percent to 48 percent in 2008. In House races in 2006, Democrats split these voters, then carried them 49 percent to 47 percent in 2008.

From the 2008 to the 2012 presidential elections, Democrats maintained their core coalition—the Hispanic vote for Obama even went up 4 percentage points in 2012—but their support among both white working-class and middle-class voters began to shrink. After getting 40 percent of the white working-class vote in 2008, Obama got only 36 percent in 2012. And after winning college-but-not-postgrad voters and middle-income voters in 2008, he lost both groups to Mitt Romney, by 51 percent to 47 percent and 52 percent to 46 percent, respectively.

The drop in midterm House races was even more precipitous. Democrats slid from 44 percent of the white working-class vote nationally in 2006 to only 34 percent in 2014, and from a 49-percent-49-percent split among college-educated voters in 2006 to a 54-percent-44-percent loss among these voters in 2014. They also dropped from a 50-percent-48-percent advantage among middle-income voters in 2006 to a 54-percent-44-percent deficit in 2014. Of course, Republicans have benefited from redistricting and from the concentration of Democratic voters in metro areas, but the Democratic losses among white working-class and middle-class voters are a prime reason that Democrats have had, and will continue to have, difficulty retaking the House.

The shift of these voters played a role in some of 2014's more surprising Senate races. In Colorado, where Democrat Mark Udall was upended by Cory Gardner, Udall lost 7 percentage points among white working-class voters from his 2008 victory over Robert Schaffer; 8 points among college-educated voters; and 11 points among voters making between $50,000 and $100,000. You might be tempted to write that off as the difference between running for the Senate in a presidential election year (2008) and in a midterm year (2014)—but consider that, from 2008 to 2012, Obama lost 8 points in Colorado among college-educated voters.

In Virginia, where highly favored Democrat Mark Warner eked out a win over underfunded lobbyist Ed Gillespie, Warner's totals among white working-class and middle-class voters plummeted between 2008 and 2014. He lost 25 points among white working-class voters, 16 among college-educated voters, and 24 among middle-income voters. Here again, these results can't be dismissed entirely as a presidential-versus-midterm phenomenon. In 2008, Obama won college-educated voters in Virginia by 50 percent to 49 percent and middle-income voters by 54 percent to 46 percent. He lost both groups in 2012.

In governor's races, the Republican edge among middle-class voters helped explain several surprisingly easy victories. In Wisconsin, incumbent Republican Scott Walker increased his advantage among middle-income voters from 56-percent-44-percent in 2010 to 57-percent-42-percent in 2014. In Ohio, incumbent Republican John Kasich increased his edge among voters with college, but not postgrad, education from 59-percent-38-percent in 2010 to 64-percent-32-percent. In Illinois, a dependably blue state, the shift of middle-class voters was a key factor in Republican Bruce Rauner's win over incumbent Democrat Pat Quinn. Rauner won college-educated voters by 60 percent to 36 percent—a 12-point shift from Quinn's margin in 2010.

Overall, Democrats have continued to get a lower percentage of the vote among white working-class voters than among middle-class voters. But during the Obama years, middle-class voters have moved away from the Democrats at a comparable—and, in a few instances, such as the Senate race in Colorado, a higher—rate than white working-class voters.

And while the white working-class vote has steadily shrunk as a percentage of the electorate, middle-class voters—as defined by education and income—have grown. In the 1980 presidential election, the white working class made up about 65 percent of the electorate; by 1988, it was 54 percent; by the 2008 election, it was just 39 percent. Ruy Teixeira and John Halpin estimate that by 2020, it'll be 30 percent of the electorate. On the other hand, voters with college degrees but not postgraduate degrees went from 26 percent of the electorate in 2004, to 29 percent in 2012, to 31 percent in the last election. And according to census estimates, turnout among middle-class voters is 10 percentage points or more higher than among working-class voters. So middle-class voters are a force to be reckoned with.

The core swing voters within the middle class are whites, who make up 70 to 75 percent of this group; but the voting patterns of minorities in this income bracket don't necessarily mirror the overall minority vote. The two most rapidly growing minority groups are Hispanics and Asians. According to a Pew study of the 2012 elections, Hispanic support for Obama was 13 percentage points lower among those with a college degree than among those without a college degree; and it was 23 points lower among those making more than $50,000 than among those making less than $50,000. (There is no comparable polling among Asian voters, but they are more likely to be college-educated and have higher incomes than other minorities—and than white voters. They recoiled against Romney in 2012, probably due to his anti-immigration rhetoric, but split their vote evenly in 2014 House races.)

Some Democrats believe that the party's support among millennial voters—the 18-to-29-year-olds who first went to the polls in the 2000s—and their successors can mitigate any losses among other groups. It is true that these voters were an important part of the original Obama coalition, but they are not quite as enthusiastic about the Democrats as they once were. In 2006, 60 percent of young voters backed Democrats in House races; that number hit 65 percent in 2008, fell to 60 percent in 2012, and slid to 54 percent in 2014. Moreover, an ongoing study by the Harvard Institute of Politics has found a steady deterioration in young voters' support for Democrats since its peak in 2008. "Our recent polling," the study wrote last fall, "shows that on a wide range of issues and questions, young voters ... now look very much like the electorate at large—pessimistic, untrusting, lacking confidence in government."

To make an educated guess about whether these trends will continue, it helps to look at how the white working class and middle class have behaved historically. White working-class voters began defecting from the Democratic Party in 1968. Initially, this was in response to civil-rights legislation and Lyndon Johnson's War on Poverty. But soon their opposition to government action on behalf of blacks had crystallized into a general opposition to government spending and taxes. According to a Democracy Corps study, white working-class voters overwhelmingly agree—by 12 percentage points more than the average voter—with the statement: "When something is run by the government, it is usually inefficient and wasteful."

For their part, middle-class voters have long been mistrustful of government. In a 2010 study based on the extensive General Social Survey conducted semiannually by the National Opinion Research Center, political sociologists Leslie McCall and Jeff Manza found that those with college but not postgrad degrees exhibited more marked opposition than any other educational grouping to government spending, and to policies that promised to redistribute income from the rich to the poor.

Before the Great Depression, middle-class voters had been a stalwart Republican constituency, and they moved back toward the Republican fold after World War II. They supported Reagan in 1980 in the wake of Carter-era stagflation and the tax revolt that began in California in 1978. Reacting to the 1991 recession, a plurality narrowly favored Bill Clinton in 1992 and 1996, but they began drifting to the Republicans in 2000 and favored Bush by 58 percent to 42 percent in 2004. In 2008, in the wake of the Iraq War and the Great Recession, they supported Obama; but in 2010—angry about Obama's stimulus program and believing that the Affordable Care Act had cost too much without truly benefiting them—they once again began returning to the Republican camp.

Middle-class voters tend, on average, to be more socially liberal than white working-class voters, and they have punished Republicans for taking harshly conservative stands on social issues. In the 2012 Senate race in Missouri, for instance, Democrat Claire McCaskill, running against antiabortion crusader Todd Akin, was able to win college-educated voters by 50 percent to 44 percent after losing them by 53 percent to 43 percent to Republican Jim Talent in 2006. She edged Akin among voters with household incomes between $50,000 and $100,000 after losing them, too, in 2006. (Perhaps some of this was due to the difference between a midterm electorate and a presidential-year electorate; but some of it was almost certainly due to Akin's infamous comments about "legitimate rape," which caused him to go from leading to trailing in the polls.)

Yet while middle-class voters are generally socially liberal, they oppose candidates on this basis only when those candidates take extreme positions. And so, when Republican politicians have soft-pedaled their views on abortion or guns or immigration, middle-class voters have largely ignored these issues in deciding whom to back—reverting to their natural tendency to focus on topics like taxes, spending, and the size of government. In 2014, Democrats in Arkansas, Colorado, Maryland, Ohio, and Virginia learned this the hard way when they centered their campaigns on their opponents' opposition to abortion rights or gun control—and lost.

Middle-class voters also tend to be less populist than white working-class voters when it comes to blaming Wall Street and the wealthy for the economy's ills. As a Washington Post poll showed last October, middle-class voters are less likely than white working-class voters or professionals to agree that America's system "favors the wealthy." Many of them work for businesses where their own success is bound up with the company's bottom line. That makes them less susceptible than white working-class voters or professionals to Democratic taunts about the "1 percent."

How middle-class voters react to this sort of populism was on display in 2012. Obama did manage to make inroads into the white working class in the North by running ads deriding Romney's sharp financial practices and his opposition to the auto bailout: In Ohio, Obama lost white working-class voters by only 46 percent to 44 percent in 2012 after losing them by 54 percent to 44 percent in 2008. But what helped with working-class voters hurt with the middle class. In Ohio, he lost college-but-not-postgrad voters by 54 percent to 44 percent after having won them by 51 percent to 48 percent four years earlier.

Maybe such a trade-off would be worth it if left-wing populism could consistently win over white working-class voters en masse. The truth, though, is that Mitt Romney was a perfect target for these populist attacks in Ohio, and Democrats cannot expect to have such a useful foil in most elections. On the whole, the white working class and the middle class—animated by their distrust of government spending and taxes—have moved toward the Republicans in recent years, in the absence of some other issue (such as war or economic catastrophe or social extremism) temporarily taking precedence. And the two groups have done so largely in tandem.

There may be no better illustration of all these trends than the 2014 Maryland governor's race. I live in Maryland, but, like most national political reporters, I know very little about what goes on in my own state, county, and city. On the eve of the 2014 election, I assumed that the Democratic candidate, Lt. Gov. Anthony Brown, would easily defeat GOP businessman Larry Hogan. The Huffington Post gave Brown a 92 percent chance of winning, and Edison Research, which conducts exit polls, didn't even bother to survey Maryland voters.

Yet Hogan won fairly easily, 52 percent to 47 percent. That evening, I received my first inkling of why from my wife, who is a dentist in a middle-class part of Silver Spring. She told me that many of her patients had been complaining about their taxes going up under the outgoing Democratic governor, Martin O'Malley. My former New Republic colleague Alec MacGillis heard the same complaints when he interviewed voters in working-class Baltimore suburbs. And that included African-American voters. These concerns played into Hogan's campaign, which was focused on O'Malley's tax increases. Brown defended O'Malley, and attempted to base the election on Hogan's opposition to abortion and gun control. But when Hogan promised not to change Maryland's laws on abortion or gun control, he took those issues off the table.

Brown, who is African-American, won majority African-American Baltimore City and Prince George's County easily. And he won 62 percent of the vote in Montgomery County, which is heavily populated by professionals. Brown really lost the election in key working- and middle-class counties where O'Malley had either held his own in 2010 or defeated his Republican opponent, Bob Ehrlich. O'Malley had tied Ehrlich in Baltimore County (which consists of Baltimore city's blue-collar and middle-class suburbs) in 2010, but Hogan won the county by 59 percent to 39 percent. O'Malley had won middle-class Howard County by 54 percent to 44 percent in 2010, but Hogan won it 52 percent to 47 percent.

Some Democrats tried to attribute Brown's defeat to racism, but Brown was a bland technocrat in the mold of former Massachusetts Gov. Deval Patrick. He did not have a history of taking strong positions on racial issues and did not do so in his campaign. Undoubtedly, some voters blanched at electing an African-American governor, but that does not seem to have driven the vote.

Instead, it appears that the election hinged on taxes and the size of government—the questions to which middle-class voters so often seem to return. In early October, a Washington Post poll had showed Brown leading by 9 points—but it also revealed that for Maryland voters, the most important issue was taxes.

Several weeks after the election, in the absence of exit polls, I sought out middle-class Maryland voters who had changed their vote from Obama in 2008 to Romney in 2012 or from O'Malley in 2010 to Hogan in 2014. I wanted to find out what they were thinking. The evidence they gave me is anecdotal, of course, but it helps to illustrate the way these issues are playing out in the minds of some voters.

Jerry is in his late 50s. He is a sales representative in Southern Maryland for a multinational corporation. He has a college degree and makes about $80,000 a year. He considers himself a "moderate Democrat." He voted for Obama in 2008 and O'Malley in 2010. He says of Obama in 2008, "He was a breath of fresh air." But after Obama became president, Jerry became disillusioned. He didn't like Obama's stimulus program. "I really think Obama messed up with all the money that we were giving out," he said. He suspects that both Obama and O'Malley primarily gave the money to "their constituencies"—most notably, labor unions. In 2012, Jerry voted for Romney, whom he admired as a "businessman." In 2014, he voted for Hogan. Taxes were an important reason. "Every year I seemed to pay more with Maryland state taxes," he explained. "I am not happy with what is happening with the taxes. I don't seem to be getting anything more from them." Brown, he feared, would continue along the same line as O'Malley. "Hogan seemed to have the message," he said.

Connie is in her mid-40s, a college graduate and a paralegal at a property-management firm. She lives in north Baltimore County. She was a Democrat until a month before last November's election, and she voted for Obama in 2008 and O'Malley in 2010. In 2012, having become disillusioned with Obama, she voted for Romney. "I was disenchanted. [Obama] made a lot of promises. I have just seen our country turn around and go backwards," she said. "I work in property management. The number of young people living on entitlement programs is overwhelming to me. I have seen it increase as never before." Last November, she voted for Hogan. "I was upset with the number of taxes that I was being hit with as a single parent," she explained. "We are overspending, and someone needs to get a handle on it, and perhaps a businessman was the best person to do that." Connie supports abortion rights, but she thought Brown misrepresented Hogan's position. "Hogan is not for repealing anything," she said. She characterized Brown's attempt to paint Hogan as a foe of abortion rights as a "political jab." Hogan's antiabortion position "didn't bother me," she said.

James is in his early 30s, a college graduate and a coordinator of services at a university in Southern Maryland. He lives in Howard County. He is one of the millennial voters on whom Democrats have rested their hopes. He voted for Obama twice and O'Malley in 2010, but in 2014, he backed Hogan. "I didn't entirely like Hogan," James said. "But I liked the idea of reining in spending." He also thinks there was "some point" to Hogan's attack on Brown as a tax-hiker. "The important thing with Brown is that he was likely to spend money. That would mean more taxes," he said. James rejects the idea that Republicans are antigovernment. "Republicans are skeptical of government," he told me.

Jerry, Connie, and James are, I would argue, very typical of the middle-class voters who are moving toward the Republicans. They are not driven by any racial animus. They are socially liberal, and would probably not vote for a Republican who was openly allied with the Religious Right, but they were willing to support an antiabortion Republican who didn't make a fuss about the issue. They are not unbendingly opposed to government, like some libertarians or tea-party activists; but they are worried about overspending and taxes.

In a speech after the election, Democratic Sen. Chuck Schumer of New York advised Democrats to "embrace government" to "get the middle class going again." But if Democrats take this advice, which has some appeal within policy circles, they could continue to drive middle-class voters like Jerry, Connie, and James away.

None of this is to suggest that America is headed toward an era of Republican domination. Going forward, the country's politics is likely to remain on a seesaw. What's clear, however, is that the Democratic advantage of several years ago is gone. And the seeds of a slight Republican advantage appear to have taken root, particularly in governor's mansions, state legislatures, and the U.S. House, where Republicans sport majorities they haven't enjoyed since the Hoover-Coolidge 1920s.

In 2016, with the economy picking up, the Democrats could take back the Senate from the Republicans, who have to defend seven seats in states that Obama won. But they are unlikely to win back the House or a majority of statehouses soon. Much of the Republican edge in midterms—which really dates from 1994—has less to do with generic voting habits than with the degree to which Republicans enjoy advantages among the political grassroots—through churches and tea-party-like groups as well as business and civic organizations—that Democrats have had a difficult time countering. For decades, Democrats depended on organized labor at the grassroots level, but labor's clout seems to be receding every year.

In presidential elections, the Democratic coalition remains formidable, and the ranks of minorities and professionals—both Democratic constituencies—continue to swell. But the party may still have a difficult time winning the presidency next year. For one thing, it's tough for either party to win three terms in a row in the White House. And in the case of the Democrats in 2016, defections from the white working class and the middle class will also continue to loom large.

The White House understands that Democrats have a problem with white working-class and middle-class voters and is now calling for a "middle-class tax cut" aimed squarely at them. Yet the Democratic nominee in 2016 will still have to shoulder the size-of-government and who-benefits-from-tax-dollars grievances created by Obama's initial spending programs and by the Affordable Care Act.

The Democrats' best chances in next year's elections will come if Republicans run candidates identified with the Religious Right or the tea party or the GOP's plutocratic wing. If Republicans are smart, they will nominate for president someone in the mold of George W. Bush in 2000 or the numerous GOP Senate candidates who won last year—a politician who runs from the center-right, soft-pedals social issues, including immigration, critiques government without calling for abolishing the income tax and Social Security, and displays a good ol' boy empathy for the less well-to-do. Such a candidate would cater to the Republican advantage among the middle class without alienating the white working class.

After the 2008 election, I thought Obama could create an enduring Democratic majority by responding aggressively to the Great Recession in the same way that Franklin Roosevelt had responded in 1933 to the Great Depression. Obama, I believed, would finally bury the Reagan Republican majority of 1980 and inaugurate a new period of Democratic domination.

In retrospect, that analogy was clearly flawed. Roosevelt took power after four years of the Great Depression, with Republicans and business thoroughly discredited, and with the public (who lacked any safety net) ready to try virtually anything to revive the economy. Obama's situation was very different. Business was still powerful enough to threaten him if he went too far in trying to tame it. Much of the middle class and working class were still employed, and they saw Obama's stimulus program—which was utterly necessary to stem the Great Recession—as an expansion of government at their expense.

In the wake of the dramatic gains Republicans have made during Obama's presidency, I now read the history of the last 80 years much differently. The period of New Deal Democratic ascendancy from 1933 to about 1968 may well prove to have been what historians Jefferson Cowie and Nick Salvatore have called the "long exception" in American politics. It was a period when Americans, panicked about the Depression, put on hold their historic aversion to aggressive government economic intervention, when the middle and bottom of the American economic pyramid united against the top, and when labor unions could claim the loyalty of a third of American workers. That era suffered fatal fissures in 1968 and finally came to a close with Reagan's landslide in 1980.

It now appears that, in some form, the Republican era which began in 1980 is still with us. Reagan Republicanism—rooted in the long-standing American distrust of government, but perhaps with its roughest theocratic and insurrectionary edges sanded off for a national audience—is still the default position of many of those Americans who regularly go to the polls. It can be effectively challenged when Republicans become identified with economic mismanagement or with military defeat. But after the memory of such disasters has faded, the GOP coalition has reemerged—surprisingly intact and ready for battle.

This article originally appeared in the National Journal.