“What is at stake in the conflict over representations of the future is nothing other than the attitude of the declining classes to their decline—either demoralization, which leads to a rout….or mobilization, which leads to the collective search for a collective solution to the crisis. What can make the difference is, fundamentally, the possession of the symbolic instruments enabling the group to take control of the crisis and to organize themselves with a view to a collective response, rather than fleeing from real or feared degradation in a reactionary resentment and the representation of history as a conspiracy.”
Pierre Bourdieu, The Bachelor’s Ball (p. 189)
“We’ll let you guys prophesy/ We gon’ see the future first.”
Frank Ocean, “Nikes” (Blonde)
In 2008, Democratic Presidential Nominee Barack Obama outperformed his predecessors John Kerry and Al Gore with virtually every single demographic group, handily defeating his Republican rival John McCain.
Many progressives came to believe that these results were not a fluke, that Obama’s coalition represented the future: an “Emerging Democratic Majority” that stood to reshape American politics as we know them.
The logic was simple: most of those who are young, college-educated, women or minorities lean left. Older white men lean right, but whites were declining as a portion of the electorate due to immigration and interracial unions. Therefore, as the older generation passes away and a younger, more diverse, and more educated cohort steps into the fore, America will become more progressive in an enduring way.
Right now, these predictions are not looking so good. In a virtual inversion of 2008 (only worse), Republicans comfortably control both chambers of Congress. They also dominate state legislatures and governorships nationwide —bodies which arguably matter more to people’s everyday lives than the federal government.
Meanwhile, Democrats lost perhaps their best chance in a generation to fundamentally reshape the Supreme Court. And the new Republican Administration seems committed to rolling back many of the signature accomplishments of the most impactful Democratic President since LBJ.
In the midst of such a bleak reality, it may be tempting to hold onto the faith that the Emerging Demographic Majority thesis remains essentially sound: Trump is an anomaly, certain to self-destruct, ushered into power as a final, desperate act of defiance by a segment of the population that knows its time is up.
However, in this instance, optimism would be ill-advised: the electoral trend actually seems to be going the opposite direction. If anything, it seems as though progressives may be on the “wrong side of history.”
From Ballot Counting to Exit Polls
The Democratic coalition rapidly deteriorated after the 2008 election. The Democrats’ 2010 midterm losses were historic: they lost the House in the most sweeping Congressional reversal of the preceding 62 years. The hole only got deeper in 2014, as the Senate also came under Republican control. Between 2008 and 2016 there was a dramatic downward trajectory across presidential races as well:
In 2008 Barack Obama beat John McCain by 192 Electoral College votes and 8.54 million popular votes. In 2012 he beat Mitt Romney by 126 electoral votes and 3.48 million popular votes. Obama’s margin of victory, while objectively comfortable, represented a 34 percent decline in his electoral dominance as compared to 2008, and a 59 percent decline in his the size of his popular vote lead.
In 2016, Hillary Clinton won the popular vote by 2.87 million. Even if she had won the presidency on this basis, it would have marked another steep decline in Democrats’ margin of victory: down nearly 18 percent from Barack Obama’s 2012 performance, and 66 percent as compared to 2008. It would have marked the narrowest popular vote margin of any winning candidate since the 2000 election (for comparison, Bush won by 3.48 million votes in 2004).
However, Clinton’s popular vote lead came overwhelmingly from densely-populated and left-leaning states like California and New York. Relative to Barack Obama, Clinton under-performed in key Midwestern states, ultimately losing the Electoral College by 74 votes and costing the Democrats the White House. All said, the Democratic Party is in its weakest electoral position since the Civil War.
Exit-polls are a great resource for understanding why the Emerging Democratic Majority thesis has failed so spectacularly over the last ten years. In fact, they are specifically designed to help pundits and analysts make sense of electoral outcomes and produce narrative frames.
Relying on New York Times exit-poll data from the last three midterm (2006, 2010, 2014) and presidential cycles (2008, 2012, 2016), we can identify longitudinal trends across demographic dimensions such as gender, race, age, income, educational attainment and ideological alignment. Patterns which persisted across all cycles are analyzed below. As one might imagine given the Democrats’ breathtaking electoral collapse, there is basically nothing but bad news:
In terms of gender, Democrats have been doing worse each cycle with both men and women. This year proved no exception: the first major-party female candidate for president, running against someone widely perceived as a misogynist, captured the Democrats’ lowest share of female voters since 2004.
And although Trump also got a lower share of female voters than his last three Republican predecessors (Mitt Romney, John McCain, George W. Bush), he nonetheless won over a majority of white women. Among men, the picture is even worse: Democrats are down 8 percentage-points since 2008, with Republicans picking up the bulk of those votes (up 5 percentage-points).
With regards to race, Democrats are down 6 percentage-points from 2008 among white voters, while Republicans are up 3 percentage-points. For blacks, Democrats are down 7 percentage-points and Republicans up 4 percentage-points.
Granted, blacks and whites are shrinking of a share of the electorate, while Hispanics and Asians are growing — but there has been a lot of volatility in how these latter groups vote. For instance, despite the fact that Trump’s campaign was largely premised on targeting immigrants and foreigners (particularly those from Mexico), while demonizing China — Trump outperformed Romney with both Latinos and Asian Americans.
Meanwhile, racial and ethnic “others” are both growing as a share of the electorate and consistently trending away from the Democrats — down a full 10 percentage-points since 2008.
Looking at age groups, 30-44 year olds are the only age group (or demographic category, period) moving away from Republicans in all three cycles. But even many of these votes have shifted to third-party candidates: the Democratic share was stagnant between 2008 and 2012, and actually fell in 2016.
Meanwhile, young people (under 30) are slightly growing as an electoral force, but Democrats are down 11 percentage-points among these voters, while Republicans are up 5 percentage-points from 2008.
Americans 45-64 have been growing even faster as a share of the electorate (up 3 percentage-points since 2008), but Democrats are down 6 percentage-points with these voters, while Republicans are up 4 percentage-points.
In terms of education, among people who lack a 4-year degree (whether they have only some college or none) Democrats are down 7 percentage-points from 2008 while Republicans are up 5 %points.
Those with BA degrees are growing as a share of the electorate (up 4 percentage-points since 2008), but are swing voters: going for Obama in 2008, Romney in 2012, and Clinton in 2016. There are differences within this group along racial lines too. For instance, Trump won a plurality of white voters with a 4-year degree.
Post-grads lean consistently left, but have been more or less stagnant in terms of their levels of support and their electoral share.
With regards to income, Democrats’ support among the poorest Americans (earning under $30k per year) has been evaporating rapidly: they are down 12 percentage-points from 2008 while Republicans are up 9 percentage-points. Granted, there are fewer Americans in this category than in 2008 or 2012 (thanks Obama), but these declines in electoral share have been offset by increases in the percentage of Americans who earn more than $100k per year-–and they consistently lean Republican. This year was no exception.
In terms of religion, Christians comprise roughly 75 percent of the electorate. Since 2008, Democrats’ support among Protestants declined by 6 percentage-points, and with Catholics by 9 percentage-points. Meanwhile, Republicans gained 4 percentage-points with Protestants and 7 percentage-points with Catholics.
Granted, Americans are increasingly straying away from these denominations, and a growing number of Americans consider themselves part of other faiths or no faith at all. However, these changes seem to be of little benefit to Democrats either: since 2008 the party declined 12 percentage-points among the non-Christian religious, and 7 percentage-points among the non-religious.
Meanwhile, Republicans actually ticked up 7 percentage-points with the former group and 3 percentage-points from the latter. Indeed, although evangelicals turned out in record numbers to support Trump, he also won more non-Christian and non-religious voters than any Republican since the 2000 election.
Ideologically speaking, the pool of those who identify as moderates are shrinking, as Americans increasingly define themselves as either liberals or conservatives.
While the portion of Americans who claim to be liberals is growing much faster than those who think of themselves as conservatives (4 percentage-points v. 1 percentage-point)— this ideological alignment has not translated into party affiliation. On the contrary, Democrats have been doing worse, every single cycle and across the ideological spectrum: down 8 percentage-points among moderates and 5 percentage-points among both conservatives and liberals.
As the number of Americans who identify as Democrats has been falling each cycle (down 3 percentage-points from 2008), Republicans and Independents have been growing (up 1 percentage-point and 2 percentage-points respectively). Simultaneously, support for Democrats among those who identify as Independents has been plummeting — down a full 10 percentage-points from 2008.
So far, we have been discussing the trend in presidential election years. However, this same dismal picture holds with regards to the midterm elections: Democrats progressively lost ground with blacks, Hispanics, Asians, men, young people, those aged 45-60, liberals, moderates, conservatives, Protestants, the less educated, and among all income groups above $30k. Meanwhile, Republicans consistently gained with all of these groups.
A Reality Check
Despite these trends, many popular narratives about the 2016 election seem to reinforce the Emerging Democratic Majority theory—for instance, the common trope that Trump was ushered into power by old, white, economically-threatened men.
In reality, Trump actually did worse than Romney among whites and seniors, while he outperformed his predecessor among blacks, Asians, Hispanics and young people.
Similarly, while Democrats lost a lot of ground among lower-income Americans, it would be a mistake to interpret these as Trump’s base: he won a plurality of every income bracket above $50k (as of 2015, U.S. median household income was around $56.6k).
All said, there is not really any demographic category one can point to where Democrats seem to be doing progressively better. However, there are lots of them on the Republican side:
Granted, Democrats still pull in nearly 90 percent of the black vote, 2/3 of Hispanic or Asian votes, and majorities among racial and ethnic “others”; they continue to capture a majority of women and young people. While the exit polls show that Republicans have been consistently chipping away at this coalition, the trend does not suggest the GOP will actually win majorities from any of these groups anytime soon.
But here’s the rub: Republicans actually don’t need to outright win (or even come close to winning) any of these demographic categories in order to deny Democrats victory. If minority turnout is low, Republicans win. If Democrats fail to capture 2012 levels of blacks, Hispanic and Asian voters, they lose.
It doesn’t really matter if lost votes go to Republicans or Independents–the outcome is the same. The Democrats’ current coalition presents a very narrow path to victory. And to the extent that Republicans actually do rally the white vote (again, Trump did not), the Democrats’ margin for error more-or-less vanishes.
Changing Demographics, Persistent Problems
It is true that demographics have shifted over the last decade in ways that would seem to favor Democrats (i.e. the electorate is more educated, urban and racially-diverse). However, the theoretical composition of the electorate doesn’t matter nearly as much as who actually turns out to vote on Election Day:
Democrats rely heavily on irregular voters to win national contests, particularly during years with presidential elections. However, irregular voters tend to stay home unless they are inspired; races between the lesser of two evils tend not to be sufficiently compelling. And even when these voters truly believe in a candidate or cause, they can be easily discouraged from going to the polls.
Democrats’ ability to motivate and mobilize this base has been in freefall since 2008. Midterm participation dropped every cycle of Obama’s tenure, with rates in the 2014 election among the lowest in 70 years. Participation in the presidential elections dropped every cycle as well, with the 2016 election pulling in the fewest voters in 20 years.
Then there’s still a basic math problem for forming a winning coalition: LGBTQ Americans comprise up to 4 percent of the electorate, Muslim and Jewish Americans combined amount to around 3 percent, blacks and Hispanics around 10-12 percent each, Asian Americans 4 percent. And of course, there is some overlap between these categories so they are not completely additive.
Meanwhile, whites amount to no less than 70 percent of the electorate (and often more). This means Democrats could get 100 percent of the votes from all other groups combined, and still not be anywhere near a majority overall unless they pulled in at least a third of the remaining white vote. However, Democrats do not have unanimous support from any of these populations, meaning they’ll need to compensate by drawing in more whites.
This reality seems unlikely to change in the foreseeable future: 2008 probably marked the highest levels of turnout and support that Democrats can plausibly aspire towards with black people. 2012 may have marked the “floor” for Republicans among Hispanic voters, given the party actually gained in 2016 with Trump as their nominee.
However, these estimates have so far assumed that populations are evenly distributed across the country. Instead, minority votes tend to be concentrated in relatively safe states and voting districts. Most of the “favorable” demographic shifts for Democrats have occurred in regions that are basically non-competitive as well. So long as this trend holds Democrats stand to benefit little, if at all, in terms of Congressional seats or Electoral College votes, regardless of how many more Americans happen to fall into Democratic-leaning categories.
Therefore, to win statewide or national races, Democrats would have to capture an even larger share of the white vote than the raw electoral share data would suggest—particularly in suburban and rural areas which tend to have higher turnout (despite their lower populations).
Finally, demographic changes present an opportunity to Democrats only to the extent that they can indefinitely maintain or expand their current levels of support among the educated, Asians or Hispanics, etc. Should these groups grow more ideologically diverse as they expand, as populations typically do— or should citizens of mixed race or ancestry come to identify as “white” and vote more like “whites” than “minorities”—Republicans would actually benefit more than Democrats from the projected population shifts.
Adjusting for relative participation rates, internal disagreement and uneven geographic distribution–in practice, a winning Democratic coalition would likely require a ratio of at least one non-minority white for each minority (i.e. LGBTQ, Jewish, Muslim, black, Hispanic, and/or Asian) constituent.
Yet Democratic support among white voters has plummeted in every election since 2008. This trend is not sustainable if progressives aspire toward any kind of majority coalition in any foreseeable future.
Obama’s election was not the first time Democrats have prophesied a permanent majority—similar claims were made prior to the ascendance of Nixon, and then again just before Reagan took the country by storm. This track record alone should inspire deep skepticism about deterministic and epochal political predictions.
Democrats have largely tried to blame their most recent loss on the idiosyncrasies of the 2016 cycle: Russian meddling, Comey’s letter, Hillary Clinton’s gender, or the popularity of third-party candidates Jill Stein and Gary Johnson. And to be sure, Trump’s candidacy and campaign were exceptional.
However, Democrats have been profusely bleeding voters since 2008, in every single cycle, and across a broad swath of demographic categories. Viewed in this context, 2016 seems less like an aberration than the culmination of a long-running trend. If steps are not taken to rectify the situation, 2018 and 2020 may prove to be more of the same.
Democrats need to fundamentally and urgently rethink their platform, messaging and outreach. But Republicans should hardly grow complacent with their apparent advantage either: In U.S. politics overwhelming majorities tend to be unstable, and nothing is truly inevitable until it actually happens.