The God That Failed
The 2000 election left a Democratic Party that was simultaneously angry, dispirited and divided. Populists believed that Al Gore made a terrible mistake by embracing the “New Democrats”—what we then called a group of socially moderate, culturally cosmopolitan, fiscally cautious Democrats in the ’90s—thereby failing to excite working-class whites. New Democrats, by contrast, thought Gore’s late adoption of heavily populist rhetoric had needlessly alienated whites with college degrees, costing him the election.
As this fight wore on, two important left-of-center thinkers, John Judis and Ruy Teixeira, wrote a book called “The Emerging Democratic Majority.” Although the book is known as a demographic work, the demographics discussed so extensively in it are, in fact, subordinate to the larger goal of the book: to find a way for the two factions in the Democratic Party described above to live together, and to win. Their framework was explicitly Hegelian/Fichtean: They described the “thesis” and “antithesis” as being the populist Democrats and the “New Democrats.” Their proposed synthesis: what they called “progressive centrism.”
Progressive centrism was never thoroughly fleshed out, but the basic idea was to combine the goals of populism—harnessing the power of government to do good for the “little guy”—with the New Democrats’ recognition of markets as a powerful tool for achieving those goals. Combined with an incrementalist approach, Judis and Teixeira argued, Democrats would form a new majority coalition. This coalition would be an expansion of the old “McGovern” coalition, and would consist of working-class whites, women, African-Americans and Hispanics, as well as professional whites living in what they called “ideopolises” – high-tech areas filled with state employees and professional workers.
In keeping with the progressive view that history is something with an arc that can be predicted and even bent to our will, “The Emerging Democratic Majority” was expressly grounded in realignment theory. This view of elections holds that the arc of history moves in roughly 30-year epicycles, where the country progresses through stages where different parties hold a position as the dominant “sun” party, or the pale “moon” party (to borrow the terminology of Samuel Lubell). The Democratic majority, we were told, would emerge fully in the 2000s.
This was mostly sensible, and certainly all very defensible. If this “soft” view of the Emerging Democratic Majority had prevailed, I probably would not have spent a good portion of the past eight years arguing against it. Changing demographics are absolutely an issue for Republicans to contend with, and if Democrats had stuck to the “progressive centrism” playbook, they could have built a powerful coalition indeed.
But the “hard” version of the theory that prevailed bore little resemblance to the nuanced view promoted by Judis and Teixeira. In the wake of the 2006 and 2008 elections, books like Dylan Byers’ “Permanently Blue,” James Carville’s “40 More Years,” Sam Tanenhaus’s “The Death of Conservatism” and Morley Winograd and Michael Hais’s “Millennial Makeover” emphasized the demographic shifts, with less attention paid to the limitations the Ur-text placed upon governing philosophy. Countless journal and website articles – indeed entire websites – sprouted up dedicated to a view of American elections characterized by red outposts being swamped by a blue demographic tide. There were different variations of the argument, but the central theme was the same: Republicans were doomed to spend quite a lot of time in the wilderness.
The Emerging Democratic Majority debate increasingly became an almost theological one, whose fundamental theories became nearly impossible to falsify. As it became clear that the 2010 elections were going to go poorly for Andrew Jackson’s party, we were reminded that elections were all about the economy, even as Barack Obama’s party suffered the worst loss in a midterm since 1938. This, while economy-based models were predicting that Democrats would keep the House.
After the 2012 elections, when Barack Obama became the first president re-elected with a lower vote share than he received in his first bid, it was declared a great 10th anniversary present for the theory, notwithstanding reminders from political scientist John Sides that “a realignment doesn’t take midterms off.”
After the 2014 debacle, where Obama became the first president since Ulysses S. Grant to face two midterm waves, it became a “Known Truth” that these differences could be described by a “midterm-electorate-versus-presidential-electorate” dichotomy, notwithstanding the fact that the demographic differences between the two electorates mathematically accounted for at best a portion of the differences in outcomes, and notwithstanding the fact that, as of 2007, Judis and Teixeira argued that the theory actually worked much better for congressional elections than presidential ones. The 2014 elections were enough to persuade Judis to abandon the theory, although most of the other followers carried on.
But after the 2016 elections, this god is dead. While I still believe that the “soft” version of this theory, as originally described by Judis and Teixeira, has much to commend itself (Democrats looking for a path forward would do worse than to re-read the book) and continue to hold that the “Emerging Democratic Majority” is one of the most important books on elections of the past 20 years, the hard version of the theory has little merit.
It’s not just that Republicans have now won four of seven elections since the book was published, although that is, as we would sometimes note dryly when I practiced law, a “bad fact.” It’s more that it is very difficult to shoehorn into the theory this election of a 70-year-old white male with a policy portfolio that is basically the antithesis of what the “Emerging Democratic Majority” recommended. It is even more difficult to do so given that Donald Trump won in the most racially diverse electorate in American history.
Even if you could somehow account for that, I do not see how you reconcile this election outcome with the Emerging Democratic Majority theory when the incumbent president has a job approval rating in the low 50s, when the country is basically at peace, and while the economy is still growing; this is not simply an odd reaction to some global disaster. Finally, to make the theory work, you have to find a way to explain the fact that, from top to bottom, the Republican Party is the strongest it has been since the 1920s. Whatever majority may be emerging, it does not look particularly Democratic right now.
Why did the theory fail? There were problems at every stage. I’ve written extensively on this over the course of my career at RealClearPolitics, but here is a summary of the various arguments I’ve made that I think have played out well:
Realignment: Part of the problem is that the theoretical underpinnings for “The Emerging Democratic Majority” are rickety at best. Much of what is today thought of as an obvious period of Democratic or Republican dominance now looks more to me like random chance. Contingency drives elections, not “history.” As RCP alum Jay Cost and I wrote in one of my first articles here:
We can ask these types of questions for many elections. What if Bush's drunk driving arrest in 2000 had never broken, or had broken months earlier? What if the Clinton administration handled the Elian Gonzalez kerfuffle differently? What if Newt Gingrich had chosen not to shut down the government in 1995? What if Ross Perot hadn't run in 1992? What if Michael Dukakis hadn't [furloughed] Willie Horton and vetoed the mandatory Pledge of Allegiance bill? What if the recession of 1982 had lasted a few months longer? What if the 1980 election had been held shortly after the hostages were released, instead of before? What if Ford had not pardoned Nixon? What if Nixon had shaved?
If you take a long view of American politics, they are exceedingly stable; a party may jump out to a lead if the other party presides over three years of economic decline (as Herbert Hoover did from 1930 to 1932) or runs for re-election amid 9 percent growth (as FDR did in 1936), but we tend toward a 50-50 nation. Even the supposed New Deal coalition was bedeviled by contingency; in 1938, Republicans nearly won the popular vote for the House; in 1940, FDR trailed in the polls until Hitler invaded Poland; in 1942 Republicans won the popular vote for the House by four points; from 1946 to 1958 they were roughly at parity with the Democrats. Our tendency as humans to attempt to find order in chaos leads us to overlook these contrary data points, and to find continuity in the high points for Democrats. But this leads us to miss the forest for the trees.
Analysts should have been skeptical of the Emerging Democratic Majority thesis because the party dominance that proponents of the theory – especially of the “hard” version of the theory – were suggesting was essentially unprecedented in American history. That doesn’t mean that something like that couldn’t happen, or that it can’t happen in the future. It just means that we shouldn’t be surprised when it doesn’t.
The black vote: Neither of Barack Obama’s wins in 2008 or 2012 were dependent upon African-American turnout. But it certainly helped. Had the Republican nominee in 2008 received George W. Bush’s share of the black vote, and had African-American turnout resembled 2004, President Obama’s 2008 lead would have been halved. In 2012 it would have been reduced to a single point.
The possibility of a reversion-to-mean among African-American voting patterns in 2016 was always a very real one. If you look at turnout rates as reported by the Census Bureau’s Current Population Survey dating back to 2002, African-American rates have always lagged Republican rates by around five points, give or take (though if you control for socioeconomic status, African-Americans are more likely to vote than whites). This was true in 2010 as well as 2014. The exceptions were 2008 and 2012, when African-American turnout rates exceeded white rates.
Now, it was possible that we had entered a period with a “presidential” electorate and a “midterm” electorate, but it was foolish to dismiss the possibility of a mean reversion once a charismatic history-making candidate such as Barack Obama didn’t top the ticket. With the African-American share of the electorate declining to 12 percent in 2016, I think it’s pretty clear that something along these lines occurred.
Likewise, with Donald Trump winning a larger share of the black vote than Mitt Romney or John McCain did, and with the midterm electorates looking more like the electorates of 2002 to 2006, we have to take seriously the possibility of a mean reversion there as well.
Hispanics: Analysis focuses on the “fast-growing” Hispanic vote, but the Hispanic share of the electorate has actually increased glacially. It was 8 percent of the electorate in 2004, 9 percent in 2008, 10 percent in 2012, and 11 percent in 2016. If we rely on the census data for the electorate, it has been even smaller. The fact that Hispanics are increasingly adopting a “white” identity (what Reihan Salam calls “racial attrition”) may blunt this growth in the future.
Moreover, I’ve long believed that analysis of what motivates Hispanic voters misses the mark. White and liberal analysts are far too reductionist when it comes to these voters, and for some reason have decided that immigration reform is a make-or-break issue for them. This ignores an awful lot of contrary evidence, such as the fact that a majority of Hispanic voters told exit pollsters in 2008 that immigration reform wasn’t important to them, or voted Republican anyway. It ignores the fact that sizeable minorities of Hispanics voted for anti-illegal immigration candidates such as Jan Brewer and Sharron Angle. It ignores the fact that a large number of Hispanic voters backed Propositions 187 and 209 in California, and so forth.
I was always skeptical (though not entirely dismissive) of the idea that Hispanic voters were on their way to voting like African-American voters. Given that Donald Trump has likely out-performed Mitt Romney among Hispanics, I think it is safe to say that 27 percent represents something of a floor for Republicans. It could be the case that Republicans will suffer further erosion here over time, but given that, over the long term, the Hispanic vote has gradually become more Republican (Bill Clinton, Michael Dukakis, Jimmy Carter and George McGovern all won larger shares of the Hispanic vote than Obama did in 2012), and that Hispanics become more Republican as they move from the border to the burbs, and that Hispanic immigration has for now leveled off, it may also be the case that the Republican share of this vote will grow.
Whites: I have written extensively about the Republican voting trend among white voters, especially among working-class whites. That is obviously an incredibly salient point in the wake of this election, where whites without college degrees voted like Hispanics, but with the impact Hispanics would have if they constituted 40 percent of the electorate. It is true that there weren’t enough working-class whites to win the election for Trump, as many asserted during the campaign. But it was closer than a lot of people think.
I’m not going to rehash everything here; it is pretty well covered in the links. I will just make two points. First, mocking the GOP as the Party of White Voters was, from an electoral perspective, extremely short-sighted. White voters are still 70 percent of the electorate (probably more). Winning around 60 percent of those voters will win a party an awful lot of elections. If Trump were to bring college-educated whites back into the fold, that share will grow.
Second, this chart should have really scared Democrats a lot more than it apparently did.
Women: Here, I can be brief. Analysts are right to examine the gender gap – the distance between the male share of the vote and the female share of the vote – but they are wrong to make predictions based upon it. As I wrote earlier this year, the gender gap giveth, but it also taketh away. We see this on full display in 2016. The 24-point spread in 2016 was actually the largest on record. But like the year with the second-largest spread (2000) and the third-largest spread (1980), it ended in Republican victory. In fact, looking at the years with the four smallest gender gaps in history (1976, 1972, 1992, 2008) we may reasonably ask ourselves if perhaps large gender gaps tend to hurt Democrats.
Overreach: The major theme of my book is that all party coalitions fall apart because, well, governing is hard and it inevitably forces parties to choose among members of their coalition. More importantly – and this is where I think realignment theory isn’t just wrong but also counterproductive – parties see their wins as a sign that they’ve finally “won” at politics. But this hubristic take is always wrong, and usually destructive. Such hubris destroyed the Republican coalition in 1910 when they thought they had won a mandate to pass the self-serving Payne-Aldrich tariff. It weakened the Democratic coalition in 1937 when FDR believed he had a mandate to pack the Supreme Court and pass the Third New Deal. It destroyed the Republican coalition in 2005 when George W. Bush famously quipped that he had earned political capital and intended to spend it.
I have little doubt that a belief that demographics would save them at the presidential level led Democrats to take a number of steps that they will soon regret, from going nuclear on the filibuster to aggressive uses of executive authority. But one thing deserves special attention. A good deal of e-ink has been spilled describing the ways in which the culturally superior attitudes of the left drove Trumpism. This too, I think, derived from a belief that history had a side and that progressives were on it, combined with a lack of appreciation of just how many culturally traditionalist voters there are in this country.
Consider these factoids: In 2004, white evangelicals were 23 percent of the electorate, and they cast 78 percent of their vote for fellow evangelical George W. Bush. In 2012, they were 26 percent of the electorate, and gave Mormon Mitt Romney 78 percent of the vote. In 2016, Donald J. Trump, a thrice-married man who bragged about sleeping with married women and whose biblical knowledge at times seemed confined to the foibles of the two Corinthians, won 81 percent of their vote. Notwithstanding the fact that I have been assured repeatedly that these voters represent a shrinking demographic and that Republicans had maxed out their vote share among them, they were once again 26 percent of the electorate.
Two points demand attention. The first, which “demographics-is-destiny” types typically gloss over, is that Trump received more votes from white evangelicals than Clinton received from African-Americans and Hispanics combined. This single group very nearly cancels the Democrats’ advantage among non-whites completely. This isn’t a one-off; it was true in 2012, 2008 and 2004.
Second, you may wonder why this group voted in historic numbers for a man like Trump. Perhaps, as some have suggested, they are hypocrites. Perhaps they are merely partisans. But I will make a further suggestion: They are scared.
Consider that over the course of the past few years, Democrats and liberals have: booed the inclusion of God in their platform at the 2012 convention (this is disputed, but it is the perception); endorsed a regulation that would allow transgendered students to use the bathroom and locker room corresponding to their identity; attempted to force small businesses to cover drugs they believe induce abortions; attempted to force nuns to provide contraceptive coverage; forced Brendan Eich to step down as chief executive officer of Mozilla due to his opposition to marriage equality; fined a small Christian bakery over $140,000 for refusing to bake a cake for a same-sex wedding; vigorously opposed a law in Indiana that would provide protections against similar regulations – despite having overwhelmingly supported similar laws when they protected Native American religious rights – and then scoured the Indiana countryside trying to find a business that would be affected by the law before settling upon a small pizza place in the middle of nowhere and harassing the owners. In 2015, the United States solicitor general suggested that churches might lose their tax exempt status if they refused to perform same-sex marriages. In 2016, the Democratic nominee endorsed repealing the Hyde Amendment, thereby endorsing federal funding for elective abortions. Democrats seemingly took up the position endorsed by critical legal theorist Mark Tushnet:
The culture wars are over; they lost, we won. . . . For liberals, the question now is how to deal with the losers in the culture wars. That’s mostly a question of tactics. My own judgment is that taking a hard line (“You lost, live with it”) is better than trying to accommodate the losers, who – remember – defended, and are defending, positions that liberals regard as having no normative pull at all. Trying to be nice to the losers didn’t work well after the Civil War, nor after Brown. (And taking a hard line seemed to work reasonably well in Germany and Japan after 1945.) I should note that LGBT activists in particular seem to have settled on the hard-line approach, while some liberal academics defend more accommodating approaches. When specific battles in the culture wars were being fought, it might have made sense to try to be accommodating after a local victory, because other related fights were going on, and a hard line might have stiffened the opposition in those fights. But the war’s over, and we won.
Perhaps comparing evangelicals to the Japanese in World War II was a bit much, and helped push evangelicals into a defensive crouch. Before my Democratic friends warm up their keyboards to protest “but we’re correct,” let me say that on some of these issues I agree with you! My point here is descriptive, not prescriptive. An aggressive approach to the culture wars and the sneering condescension of the Samantha Bees and John Olivers of the world may be warranted, but it also probably cost liberals their best chance in a generation to take control of the Supreme Court. That’s a pretty steep price to pay. It may well be that Democrats would be better able to achieve their goals if they were less, for lack of a better word, fundamentalist about those goals. Henry Clay famously declared that he would rather be right than president; he at least got his way on the latter.
Of course, none of this is to say that the Democrats are shut out of power. Far from it. Trump’s majority is fragile; a percentage point difference in three states, and then we are having a quite different conversation right now. Republicans are in charge, and they own America from top to bottom. The global economy is shaky, and a recession in the next four years is likely. Iran, China, and Russia all pose threats to global stability. There is always a possibility that a scandal could topple an administration.
If Republicans govern well – and that is a big if – then they will have an opportunity to build a strong majority of their own. Fusing Trump’s vote share among whites without college degrees with Romney’s vote share among whites with college degrees would almost guarantee Trump a return to the White House in 2020. But more than that, a class-based appeal would probably appeal to African-Americans and Hispanics, and remake the way we view politics. History could remember Trump as it does William McKinley, taking power at the end of the Long Recession and inaugurating a period of peace and prosperity. If, however, Republicans govern poorly, the House and Senate will fall, and history will remember Trump as we do Jimmy Carter; a footnote who was sadly unprepared for the job.
Ultimately, though, I think it is foolish to predict what history’s judgment will be. If the last few elections have done nothing else, it has been to convince me that history has no arc; it bends toward nothing; we are certainly ill-equipped to harness whatever power it has. Rather, it simply meanders like a lazy river; we are carried along by the current, and we label what we hope is around the next bend “justice.”
Correction: In an earlier version of this article, Henry Clay's quote was misattributed to Daniel Webster.