What happened to President Obama’s majority?
The surface-level answer to this question is simple. Donald Trump won over a supermajority of white working-class voters while holding on to many traditional Republicans, allowing him to win the Electoral College. At the same time, down-ballot Republicans built on their solid showings in 2010 and 2014, which helped them retain control of Congress and significant power in the states.
But there’s a deeper dimension to this question. For the past eight years, election wonks have been arguing about whether Democrats have an unassailable majority in presidential politics. The theory was straightforward: As the electorate becomes more racially diverse and educated, the Democratic base would grow and the party would build an invincible majority. The theory wasn’t just Democratic wishful thinking. Many good analysts from across the political spectrum bought into it at the time. So the question isn’t just why Democrats lost, but how such a widely respected theory went so wrong.
One way to tackle this question is to trace these arguments back to their source – a 2002 book called “The Emerging Democratic Majority” by John Judis and Ruy Teixeira. That work argued that Democrats could weld together a durable, winning coalition of minorities, professionals, working women and white working-class voters by championing a socially and fiscally liberal set of policies called “progressive centrism.” Many wonks read this book, saw the elections results in 2006, 2008 and 2012 and concluded that a durable Democratic majority had indeed arrived.
And that’s where many political observers made a key mistake – Barack Obama built a winning coalition, but it wasn’t the one that Judis and Teixeira described. The authors’ imagined coalition probably would have looked more like John Edwards than Barack Obama, and it might not have had some of the vulnerabilities that led to the Democrats’ recent electoral collapse.
The Emerging Democratic Majority and the Obama Coalition Are Very Different
Judis and Teixeira summarized their vision of a majority by saying, “It is fair to assume that if Democrats can consistently take professionals by about 10 percent, working women by about 20 percent, keep 75 percent of the minority vote, and get close to an even split of white working-class voters, they will have achieved a new Democratic majority.” But President Obama’s majorities in 2008 and 2012 and Hillary Clinton’s popular vote win departed from this blueprint in critical ways.
Obama and Clinton relied more on minority voters than Judis and Teixeira recommended. Obama won about 85 percent of the nonwhite vote in 2008 and about 80 percent of nonwhite voters in 2012, according to the American National Election Studies (ANES). Exit polls show similar numbers: Obama won roughly 80 percent of the nonwhite vote in 2008 and 83 percent in 2012. And in both elections, his vote total was boosted by historic levels of black turnout. The jury is still out on Clinton’s exact level of support with nonwhite voters -- exit polls, county-level data and post-election polls tell somewhat different stories -- but initial evidence suggests she may have also performed at or above the benchmark that Judis and Teixeira set.
But Obama’s and Clinton’s overperformance with nonwhite voters helped mask growing Democratic weakness with the white working class.
This graphic shows two important trends. The first might be familiar to some readers -- it shows the Republican candidate’s lead with white working-class voters in the last 10 elections, per the exit polls. Generally Republicans have performed well with this group: Bill Clinton was the only Democrat in the time period covered to win them, and typically the Republican candidate prevailed with this group by double digits regardless of whether he prevailed in the general election.
The second trend is less well-known but arguably more important. It shows Republican performance with non-college-educated whites adjusted for the popular vote margin (e.g. John McCain lost the popular vote by about seven points and won white working-class voters by 18 points, so his adjusted lead is about 25 points). Put simply, it shows Democratic problems with the white working class have been growing ever since Bill Clinton’s name left the ballot. (Note that exit polls rather than ANES were used because ANES hasn’t released 2016 data yet. ANES data from 1980 to 2012 shows the same shape as this graph.)
This trend strongly suggests that Obama and Hillary Clinton failed to hit Judis and Teixeira’s benchmark of 50 percent of the white working class. It’s important to note here that exact definitions of working class can skew these comparisons -- Judis and Teixeira often define “working class” and “professional” according to category of job rather than education level. But Democrats lost non-college-educated whites by almost 40 points. Given those numbers, it would be hard to come up with any reasonable definition of white working class – occupational or otherwise – that ends with them giving equal support to Clinton and Trump.
It’s a bit harder to compare the Obama coalition to Judis and Teixeira’s benchmarks when it comes to professionals and working women. Those terms are both occupational, so comparing them to commonly used educational or income-based vote breakdowns is dicey. That being said, Clinton and Obama both performed well with women and college-educated voters. Clinton came close to winning (or, depending on the data source, won) college-educated white voters while winning college graduates as a whole. That’s a significant achievement for a Democratic presidential candidate -- ANES shows that Republicans have won college-educated white voters from at least 1952 through 2012. Additionally, Clinton and Obama have managed to rebound with women after John Kerry nearly lost them in 2004. This is probably due in part to increased Democratic strength with working women and single women -- key parts of the coalition that Judis and Teixeira described.
In other words, Obama didn’t lead Democrats into the era of dominance described by the “Emerging Democratic Majority” authors. He led the party to a different majority by relying more heavily on the “ascendant” pieces of his coalition (minorities, women and well-educated voters) while losing strength with the white working class. So it’s worth asking: What would a Democratic Party that followed the 2002 roadmap look like? And would it have been able to avoid some of the problems that Obama’s coalition encountered?
The Emerging Democratic Majority Might Have Looked Like a Cleaned-Up John Edwards
If Democrats really wanted to follow Judis and Teixeira’s script, they probably should have nominated someone like John Edwards in 2008. Obviously, Edwards himself would have been a disastrous choice for the presidential nod. Late-breaking revelations about an extramarital affair (while his wife was terminally ill), a child from that affair and subsequent indictments for violating campaign finance law would (and should) stop any presidential campaign in its tracks.
But Edwards (sans the moral failures) seems like the sort of politician who could weld together Judis and Teixeira’s majority. Edwards was an economic progressive. His memorable “Two Americas” speeches focused on economic inequality between the wealthiest set of Americans and the rest of the country, and his campaign often emphasized increasing access to health care, improving education and hammering predatory lenders. He was liberal on abortion and LGBT issues (in the context of his time), but he was for the death penalty and at least tried to appear deferential to gun owners. Maybe most importantly, Edwards was able to take liberal policy positions without projecting cultural cosmopolitanism. If you roll all of that together, you get something like what Judis and Teixeira described – a fiscal and social liberal who could still make a credible appeal to white working-class voters.
A candidate like Edwards would probably not do as well with racial minorities as an Obama-era Democrat, but he or she would have other advantages. For example, a North Carolina county-by-county regression model shows that in 2016 Hillary Clinton outperformed Edwards’s 1998 showing in highly educated and racially diverse areas, but Edwards’s vote share went up as the rate of college education among whites went down. Clinton narrowly lost the state, but if Edwards’s 1998 county-wide vote shares were applied to the 2016 electorate, he would have barely won.
The same principle might have carried over to other states. Judis and Teixeira rated Kentucky, Louisiana, Arkansas and Texas as “Leaning GOP but Competitive” and West Virginia, Missouri, Florida and Ohio as “Leaning Democratic” in 2002. Some of these ratings now seem absurd. Louisiana, West Virginia, Kentucky, Texas, Missouri, Ohio and Arkansas were all far from competitive in 2016. An Edwards-style Democratic Party certainly wouldn't have carried all or even most of these states, but it might not have lost all of them either.
A Democratic Party in Edwards’s mold may have been less willing to embrace cultural cosmopolitanism (and thus may not have won California by 30 points), but it might have had better luck down-ballot. As Sean Trende and I recently noted, the geography and political divisions in the United States tend to give rural whites a significant amount of power. Key Democratic constituencies – Hispanics, African-Americans, Asians and liberal whites, to name a few – are often concentrated in safe states, urban centers or both. Many Hispanics live in Texas, California, New Mexico, Arizona and Nevada – none of which, with the exception of Nevada, can be properly categorized as swing states. A large proportion of African-Americans live in blue city-states like Maryland or deeply Republican Southern states. Asians, a growing Democratic group, often live on the already blue West Coast. Moreover, racial minorities and college-educated whites frequently live near the core of cities, allowing them to be packed into a relatively small number of congressional districts. Working-class whites, on the other hand, are spread more evenly across the states and congressional districts – giving them more power in down-ballot races.
Put differently, an Edwards-style Democratic Party would likely have to tolerate some dissent on cultural and social issues and at times settle for incremental rather than sweeping implementation of progressive economic policy. But they might have had a better chance of holding ancestrally Democratic congressional districts in the Midwest and North Central regions of the country.
But maybe most importantly, this version of the Democratic Party might not have been as vulnerable to a populist as the Obama coalition. Trump aimed much of his platform and personal appeal at working-class white voters – he emphasized building a border wall, renegotiating trade agreements and giving voice to a group of people who felt they had lost cultural standing as well as economic opportunity. If the national Democratic Party had more cultural appeal to working-class whites, they might have been able to stop the bleeding enough to hold states like Pennsylvania, Florida, Michigan, Wisconsin or North Carolina.
This Isn’t Prescriptive or Perfect
None of the arguments here are prescriptive for Democrats going forward. “The Emerging Democratic Majority” was written to help the party regain power in the mid- to late- 2000s and potentially hold it for a decade or two. At this point we’re in the back half of the 2010s. The country’s economic and cultural situation has changed, and both parties need to continuously adjust their playbooks if they want to gain or maintain power.
I’m also not claiming “The Emerging Democratic Majority” outlines a perfect strategy. It’s possible that in some other universe, Democrats found a morally cleaned-up Edwards or Edwards-like figure, ran him or her in 2008, and ushered in an era of Democratic dominance. Or maybe the Republicans of that universe found a hole-in-the-Death-Star-like flaw in that strategy and dismantled Democrats even earlier than 2016.
But Democrats in our universe didn’t try the Judis and Teixeira strategy. So we’ll never really know if they could have built a durable majority in the late 2000s and 2010s – all we know is that Obama didn’t.