The New Electorate and the Future of the Democratic Party

The New Electorate and the Future of the Democratic Party

The last two presidential elections have left Republicans reeling and Democrats crowing. Demographic changes suggest the ascendancy of the Democratic Party. Will this be the defining feature of politics over the coming decades? Has an enduring Democratic majority arrived?

In this report, we posit that this is not the case. Faith that demographics will deliver Democrats to power over the next several decades rests on seven main illusions about voters and their beliefs and behaviors, for which there is conflicting evidence. Rather, Democrats must resist complacency and develop an accurate understanding of the new electorate.


Democratic candidate Michael Dukakis won 40% of the white vote in 1988 and lost the election by 315 electoral votes to Republican President George H.W. Bush. Twenty-four years later, President Obama won 39% of the white vote and defeated Republican candidate Mitt Romney by 126 electoral votes. This data alone seems to suggest that demographics did indeed deliver the 2012 election to Democrats.

Consequently, the dominant narrative has been that the Grand Old Party needs to adapt or risk going the way of the Whigs. Aware of this criticism, Republicans have begun to consider how to position themselves better to win national elections. Less attention has been paid to the future of the Democratic coalition. Many of those who have focused on the future for Democrats have cast victory as inevitable, arguing that the Democratic Party is destined to win the presidency with the new American electorate. But obscured within analyses of 2012 is a set of illusions about voters—illusions that, if Democrats embrace, will put their majority at great risk in the future.

Can Democrats rely on demographic cohesion to deliver them to power in coming elections? Are Hispanics, Asians, and Millennials brand-loyal to the Party? Two examples provide contrasting perspectives: black voters and Catholic voters. Black voters have supported the Democratic Party in large numbers for nearly 100 years, with Democratic presidential candidates winning the black vote at least since 1936 and garnering 80%+ since 1964.1 Searing legacies of racial discrimination, the Republican adoption of the “Southern Strategy,” continuous hostility by some Republican candidates and electeds, and the championing of equal rights by the modern Democratic Party to redress past grievances has created a durable bloc of Democratic voters. Conversely, Catholics, who were once stalwart Democrats in the early 1900s, became Reagan Democrats by the latter 20th century and now comprise a bloc of swing voters.

How likely are the growing demographic groups in the electorate (Hispanics, Asians, and Millennials) to maintain support for Democratic candidates in the future? Are Hispanic, Asian, and Millennial voters more like party-loyal black voters or swing Catholic voters? To answer this question, we analyzed more than forty public opinion surveys and exit polls and examined primary and secondary source material to evaluate past trends and emerging developments. The report summarizes the demographics of the 2012 presidential election, and then unpacks the seven main illusions underlying "Demography is Destiny," the wildly popular and to date unquestioned progressive view of the 21st century electorate.

Demographics and the 2012 Presidential Election

Electoral Data

Fewer White Voters

For years, demographers have noted the declining share of whites in the U.S. population, and this trend is now evident in the electorate. Prior to 1996, white voters composed 85%–89% of the turnout in presidential elections.2 A gradual shift began in 1996 and accelerated in the 2000s. By 2004, the white share of voters fell under 80% for the first time. Simultaneously, Hispanics began increasing their share of the electorate. After an initial 5 point jump in the 1990s—from 2% in 1992 to 7% in 2000—Hispanic voters have increased their share of the electorate by approximately 1 percentage point every presidential cycle. And, more recently, the number of Asian voters has also increased, growing by 1 point between 2008 and 2012 alone, matching the rate of growth for Hispanics. While the proportion of black voters in the electorate had remained relatively stable for decades, their turnout increased in 2008 and 2012, rising to 13% of the electorate.

Composition of Electorate in Presidential Years

* From 1976–1988 the sample size of Asian voters was too small for statistically significant analysis.

Rebounding Youth Vote

Today’s young people are voting in records that match Baby Boomers, overcoming the apathy of Gen Xers, with approximately 50% of Millennials going to the polls.3 In 1976, nearly one-third of voters were 18–29 years old. That number fell by about 10 points in 1980, when the youth share of the electorate dropped to 23%. Their proportion of the electorate remained near 20% until bottoming out at 17% between 1996 and 2004. But in 2008, young people increased their participation, comprising 18% of the electorate. And by 2012 they gained another point, making up 19% of the electorate.

The impact of Millennials on national elections is amplified by their sheer size. While the Baby Boomer Generation (born 1946–1965) was large, with about 81 million U.S. adults, Gen X (born 1966–1980) was smaller, with only about 61 million. But there are more Millennials (born 1981–2000)—at 85 million people—than even Baby Boomers.4 Millennials are also the most racially and ethnically diverse generation in recent American history, which partially drives the declining share of whites in both the U.S. population and the electorate. The increase in the proportion of the electorate that is 18–29 years old and nonwhite has been driven both by relatively higher participation rates among Millennials and the size of the generational cohort.

President Obama’s Coalition

Clearly, the electorate today is different from 20 years ago, partially reflecting a new generation’s entrance onto the political scene. And President Obama and Democrats won in 2012 by running up the score with the demographic groups who are increasing their share of the electorate, even while losing men, whites, and seniors.

2012 Voting Patterns by Selected Demographic Groups

Despite losing white voters by 20 points, President Obama was successful partially because that group represents a declining share of both the U.S. population and the electorate. By winning nonwhites by 40+ points—groups which have all increased in size in the electorate recent years—Democrats were able to overcome the deficit.

Much of the commentary in 2012 also focused on the women’s vote. And President Obama won women by 11 points (a slight drop from his 13 point margin in 2008). Democrats have won the women’s vote since 1992, so this is not particularly surprising. However, the President lost white women by 14 points. The difference was thus made up with nonwhite female voters.

The final demographic piece of the puzzle is the youth vote. In the last two presidential elections, 18–29 year olds have been drawn almost exclusively from the Millennial Generation. Voters 18–29 years old supported President Obama by wide margins in both elections—34 points in 2008 and 23 points in 2012. With the exception of 1984 and 1988, Democrats have won the youth vote in every election since the 1970s. However, their margin of victory has never topped 20 points, let alone the 34 point spread in 2008. And younger voters now outnumber seniors, who have tended to vote Republican since the 1970s. However, this will likely change as the large and more evenly divided Baby Boomer Generation ages into the 65+ category in the electorate.

The electorate has changed in dramatic ways over the past 20 years. Reflecting long-term population trends, the composition of the presidential electorate has shifted—with some groups growing in size and influence. Among specific demographic groups, we also now have lopsided margins in support of one party or the other—with Democratic support at supermajority levels among voters who are increasing in the electorate. Taken together, these demographic trends and voting patterns require consideration of their long-term impact on the American political system.

Demography Equals Destiny

It may have been Stephen Colbert’s fictional alter-ego who remarked, “If there is one thing the reelection of Barack Obama proved it’s that demographic shifts are making it harder for the GOP to win nationally,” but his sentiments have become almost an article of faith to many political analysts and observers.5 President Obama’s victory in 2008 was described by proponents of this argument as a progressive triumph, embodying a 20-year shift in American politics and demographics.6 By his 2012 reelection, proclamations heralded the era in which the “McGovern coalition” came of age and could finally reign.7 Demographics were so decisive, the narrative goes, that the 2010 census might have been “the most significant event of this presidential contest.”8

The core of this argument—developed initially by John B. Judis and Ruy Teixeira in their 2002 book The Emerging Democratic Majority—is that the proportion of voters who are nonwhite, college graduates, professionals, single and working women, Millennials, and infrequent church attenders (sometimes referred to as the unaffiliated or the unchurched) is increasing—and these voters support Democrats. Conversely, white voters (especially the working class) are declining in proportion of the electorate—and these voters increasingly support Republicans. Finally, older generational cohorts—who tend to be more conservative and vote Republican—are being replaced by the Millennial generation, which proponents claim is more progressive than other generations. The result, according to many, is “an array of growing demographic groups that have aligned themselves with progressives and swelled their ranks.”9

These demographic changes matter politically since, proponents of this view argue, these voters side with liberals both ideologically and in specific policy areas. Drawing on survey data, these analysts contend that alignment between progressive ideas and the changing electorate is evident in their positions on concrete issues, including:

  • The declining salience of culture war issues, due to widespread Millennial support for marriage for gay couples, the rise of “secularism,” and “dramatic changes in attitudes towards sexuality, marriage, and gender roles…with much greater openness toward sexuality outside of heterosexual marriage and a strong belief that women are equal in every respect and should work outside the home if they wish.”10
  • Increasing support for a positive role for government, especially in economics and health care.11
  • Issue alignment in prioritizing clean energy over extractive sources, emphasizing diplomacy in national security, and increasing public investment to modernize education.12

Importantly, demographic changes and issue salience are intertwined for those who espouse this viewpoint, wherein the “current tilt of the public toward progressive ideas and priorities . . . is being accentuated by the strong support for this agenda among growing demographic groups.”13 And unlike Democratic dominance in other eras, this one is marked by a “more stable ideological plurality.”14

And these demographic groups are geographically situated to deliver Democrats to power due to a “close relationship between pro-progressive political shifts and dynamic growth areas across the country, particularly within contested states.”15 That is, demographic groups rising in prominence are located in purple areas—turning swing states blue.

The case that Democrats can turn changes in demographics into a permanent majority is bolstered specifically by Millennials. Proponents of this theory believe that Millennials are a progressive generation and will exert a profound influence over politics because they are large in size, they are the most diverse generation, and they support progressive views on some benchmark issues at levels higher than the rest of the population.16 By 2020, 90 million Millennials will be eligible to vote, which could comprise up to 40% of all eligible voters.17 Millennials now are 40% nonwhite, which proponents predict would rise to 44% by 2020, a shift that they say “should make the Millennial generation even more firmly progressive as it fully enters the electorate, since minorities are the most strongly progressive segment among Millennials.”18 Further, proponents note that even white Millennials are more progressive than whites of other generations. So under this viewpoint, the demographic changes taking place in the electorate will conjure a stable and dominant progressive majority long-term.

To be sure, proponents acknowledge obstacles to achieving progressive victories on policy issues.19 The structural design of our political system provides many opportunities for conservatives to thwart progressive goals, which is magnified by Republican control at the state level and ideological unity among Congressional Republicans. They also note that low levels of trust in the federal government and political efficacy to enact change dampen prospects for progressive success. Despite these caveats, the dominant liberal view is one of a demographic Democratic advantage that should translate into presidential victories in 2016 and beyond.

The Seven Illusions of the New Democratic Coalition

The dominant explanation that President Obama won reelection because of a demographic shift in our electorate seems plausible on its face, given the wide margins of support Democrats received from particular groups. As long as the groups “naturally” aligned with the Democratic Party comprise a large enough share of the electorate, it seems that Democrats should win, at least at the presidential level. But is the Obama Coalition likely to remain stable and reliable Democratic voters? Will they be brand-loyal to the Party going forward?

Neither partisan stability nor internal partisan ideological agreement is the norm in politics. Drawing upon demographic data, we assess the level of ideological cohesion and stability of the Democratic Coalition, concluding that the coalition is tenuous and in danger of dissolution if taken for granted.

Illusion #1: The new electorate is predominantly liberal

Are Hispanics, Asians, and Millennials overwhelmingly liberal? The evidence is mixed. A 2011 Pew survey found that Hispanics are nearly evenly split, with 32% identifying as a conservative (34% in the general population), 31% as a moderate (39% in the general population), and 30% as a liberal (21% in the general population).20 In this formulation, Hispanics are 9 points more liberal than the population as a whole, 8 points less moderate, and 2 points less conservative. This is logical, as Hispanics tend to be younger than non-Hispanic whites and older voters—especially those in the Silent Generation—are more conservative than later generations. In the 2012 Census Bureau Population Estimates, the median age among Hispanics (of any race) was 27.3 compared to 42 for non-Hispanic whites and 37.3 for the U.S. population as a whole.21

Hispanic Ideology 2011

But the 9-point difference in liberalism belies the fact that Hispanic voters are not reflexively liberal—in fact, they divide themselves into thirds, with nearly two-thirds identifying as moderate or conservative. For President Obama to win them 71% to 27% in 2012, he had to win nearly all self-identified liberal and moderate Hispanics, as well as garner approximately 15% of conservative Hispanic voters. The key question is how sustainable these margins are long-term.

A similar trend is evident with Asians. In Pew’s 2012 survey of Asian Americans, 24% self-identified as conservative and 31% as liberal. This represents a shift away from conservatism and towards liberalism as compared to the general population, which Pew found to be 34% conservative and 24% liberal. Thus, according to this dataset, Asians are 10 points less conservative and 7 points more liberal than the U.S. average. But notably, 37% of Asians identify as a moderate—the same as in the general population—making moderates the plurality of both the general population and Asians.22 Even if he won all of the self-identified liberal Asian voters, President Obama would have needed an additional 42% of the Asian vote to reach his 73% margin of victory in 2012. And capturing all of the moderate Asians would only have gotten him to 68%. Thus, based on ideology, the President over-performed what we might expect among this bloc of voters.

Asian Ideology 2011

The Millennial Generation tends to include fewer self-described conservatives and more self-described liberals than the population as a whole. For example, Pew analyzed data from 1996–2011 (excluding 1997, for which they did not have data) by generation. On average, for the general population over those 15 years, 20% identified as liberal, 37% as conservative, and 38% as moderate. But for Millennials, 28% identified as liberal—8 points more than the population as a whole. However, another 28% of Millennials identified as conservative, the same percent who identified as liberal. Thus, Millennials are not an overwhelmingly liberal group. In fact, moderates dominate, as a plurality, at 38% of Millennials. Of course it’s not just Millennials. Among both Gen X and the Baby Boomers, approximately 39% identify as a moderate—outstripping the share of either liberals or conservatives. It is only the Silent Generation that tilts overwhelming conservative, averaging 43%.23

Ideology by Generation

There is also growing evidence that younger voters increasingly identify as moderates. An annual survey of incoming college freshman, with over 190,000 respondents, found that the number of incoming students describing their ideology as “middle-of-the-road” (as opposed to liberal or conservative) rose by 4 points since 2008—from 43.3% to 47.5%.24 Another 2013 survey of 18–29 year olds found that 49% identified as a moderate.25 That’s a far cry from proof that Millennials are a reflexively liberal group.

Women were also crucial to President Obama’s victory. In Pew’s 2012 Values Survey, 36% of the population identified as conservative, 37% as moderate, and 23% as liberal. Women were slightly less conservative (33%) and slightly more liberal (25%) than the average, but 37% identified as a moderate, identical to the figure overall.26 Thus, women appear more liberal than the general population but still more moderate than left.

There is credence to the notion that the rising American electorate trends more liberal than other groups. Hispanics, Asians, and Millennials all display less conservatism and more liberalism than the general population. But among Asians and Millennials, the single largest ideological bloc of voters are moderates, while for Hispanics the distribution is virtually identical among liberals, conservatives, and moderates. Rather than liberal ideological dominance, we are faced with a solidly moderate coalition, who trend somewhat more liberal and less conservative than the overall U.S. population.

Illusion #2: Millennials’ ideology will be stable over the course of their lives

As we noted above, Millennials display somewhat more liberalism and less conservatism than the general population. Will this be an enduring feature of that generational cohort? Historically, ideological proclivities tend to fluctuate, even within generational cohorts. For example, the Silent Generation described themselves as 40% conservative and 14% liberal in 2000; by 2011, the number of conservatives increased by 6 points while the number of liberals increased by 2 points. For Baby Boomers during the same timeframe, the number of conservatives increased from 35% to 42% and the number of liberals changed only slightly, increasing from 18% to 19%. The share of conservatives among Gen Xers increased from 30% in 2000 to 36% in 2011, while the share of liberals fell by 3 points, from 24% to 21% (changing from a 6 point conservative advantage to a 15 point spread).27

Pew’s analysis is useful for examining the scope of ideological fluctuation. If we look at the highest and lowest percentages of generational cohorts identifying as a liberal, moderate, or conservative, we can create a range within which ideological identification has fluctuated for these groups. Just within this 15-year time frame, we see considerable variation. For example, the percent of Baby Boomers identifying as a conservative fluctuated by 10 points; that figure is 9 points for Silents. We often assume that ideology is not only stable but also deeply held and therefore not responsive to real-world events. Yet this data suggests that ideology is less genetically embedded or predetermined and more dynamic and prone to fluctuation.

ideological Variation by Generation

If you look at the conservative high points in the table above, it seems impossible for Democrats to win. If you look at liberal high points, it seems impossible for Democrats to lose. In short, people’s views change. For Millennials’ ideology to remain constant indefinitely, they would have to break the historical patterns of prior generations.

Illusion #3: Hispanic, Asian, and Millennial voters will remain loyal to the Democratic Party

Are Democratic voters brand-loyal? A key Illusion in the argument of Democratic ascendancy is that Hispanic, Asian, and Millennial voters will remain loyal to the Party. Black voters have remained loyal Democratic voters at least for the past several decades. Shouldn’t we expect the same brand loyalty from Hispanic, Asian, and Millennial voters?

Black Voters Party Identification

* The 2012 exit poll dataset has not been publicly released so we do not have figures for 2012 party identification.

While some Hispanic, Asian, and Millennial voters will undoubtedly remain lifelong Democrats, the evidence that each of these groups is filled with strong partisans is mixed, at best. In Pew’s 2012 Values Survey, they found that a plurality of Millennials (45%) described their affiliation as Independent—an increase of 6 points since 2008. Another 31% identified as a Democrat and 18% as a Republican.28 While currently Democrats perform better than Republicans among them, a sizeable portion of the Millennial generation has not aligned with either party. And it’s not clear that even those who do consider themselves Democrats now will remain loyal to the Democratic brand. Millennials have proven to be tricky for marketers in the private sector, in particular because they are willing to switch even from their most favored brands if they can get a better deal.29 Why would politics be any different?

While Hispanic voters have overwhelmingly supported Democrats in recent years and aligned with the Democratic Party, many define themselves as political Independents—a fact that makes them much more likely to wander between parties.30 In Gallup’s analysis of over 300,000 responses to their polls in 2012, they found that half of all Hispanics identify as an Independent, 32% as a Democrat, and 13% as a Republican. Once the Independents are asked if they lean towards one of the parties, then 51% say they are Democrats or lean that way, 24% say they are Republicans or lean that way, and 20% say they are “pure” Independents. These figures for Independents are significantly higher than among the general public, where only 39% claim to be Independents, and once leaners are removed that figure drops to 12%.31

Clearly Democrats have an edge over Republicans in Hispanic party identification right now. But President Obama’s 44 point margin among Hispanics is substantially more than any lead Democrats have in party identification. President Obama thus over-performed among Hispanics, winning substantial numbers of Independent Hispanics. These voters may not align with the Democratic Party, and could be winnable For Republicans in future elections. In this regard, Asians are very similar to Hispanics. In Pew’s 2012 Asian American survey, 33% identified as a Democrat, 18% as a Republican, and 34% as an Independent (another 14% had no preference, were unsure, or refused).32 Thus, President Obama again over-performed among this group.

The data on Millennials, Hispanics, and Asians suggests that partisan attachments appear to be neither particularly strong nor necessarily durable. While there is a vocal minority of partisan activists among these groups who will likely remain loyal partisans, the number of Independents challenges the assertion that a permanent attachment has taken root outside of the Party base.

National exit polls provide us with another layer of detail, albeit limited to those who choose to vote. But they confirm that partisanship among key groups is prone to fluctuation. In 1988, 18–29 year old voters were evenly split between Democrats and Republicans, with just under one-quarter identifying as an Independent. Since then, partisanship has varied. But the overall trend has been fewer self-identified Republicans, more Democrats, and a relatively steady and large number of Independents. The high water mark for Democrats was 2008, when younger voters identified as Democrats by nearly 20 points more than with Republicans. But even in that year, nearly 30% of Millennials identified as Independents.

18–29 Year Old Voters Party Identification

According to the exit polls, about half of all Hispanic voters identified as a Democrat in 2008. In 1980, when the Hispanic population was much smaller, more than two-thirds identified as a Democrat. The proportion has fluctuated considerably since that time, but it has fallen below 50% only once (to 42.5% in 2004). Republican identification among Hispanic voters, by contrast, started low in 1980 (only 15.2%) and then settled into a comfortable range of 20%–25% in the 1990s and 2000s. The biggest exception in recent times was again in 2004, when Republican identification jumped to 31% and President Bush won 44% of the Hispanic vote. In recent years, the number of Independents has also increased, comprising 27.7% of Hispanic voters in the 2008 exit poll, up a full ten points from 1980.

Hispanic Voters Party Identification

In 1992, Asian voters were evenly split between the two parties, with a plurality identifying as an Independent. By 2000, Asian voters had shifted slightly towards the Democratic Party, with about three in ten identifying as a Republican and the same proportion as an Independent. Since then, the number of self-identified Democrats has fluctuated, but Asian voters have largely stayed with the party, with a margin ranging from 3–15 points. The number of Republicans has hovered around three in ten, before falling to one-quarter in 2008. The share of Asian voters who are Independent has been stable, at about 35%, since 2004.

Asian Voters Party Identification

Given the lopsided margins that President Obama won among 18–29 year old (+23), Hispanic (+44), and Asian (+47) voters in 2012, we might expect higher levels of partisanship among these groups. Yet in 2008, it was only among Hispanics that a majority identified with the Democratic Party—and just barely. While these groups supported both candidate and President Obama in 2008 and 2012, they have not cemented as strong Democratic partisans, and their brand-loyalty should not be assumed.

Illusion #4: Millennials deeply believe in and Support an activist government

“If you had to choose, would you rather have a smaller government providing fewer services, or a bigger government providing more services?” This poll question has long been a proxy for ideology. While there has been some generational continuity on this question, attitudes have fluctuated over time. For example, the Silent Generation supported a smaller government by 52 points in 1996, but that margin was only 15 points just 7 years earlier. Baby Boomers have continually shifted their perspective, favoring a smaller government by 42 points in 1996, leaning towards a larger government by 12 points in 1989, and being closely divided on the issue in other years.33

Support for Bigger Government and More Services

Indeed, fluctuations in aggregate and generational opinion often appear as a reaction to the real or perceived excesses of the ruling party. In 1980, perhaps in response to the perceived big-government policies of former Democratic President Jimmy Carter, 54% favored smaller government. But after two terms of “government is the problem, not the solution” Republican President Ronald Reagan, 48% supported a bigger government providing more services. Rather than a deep and lasting political value, preferences on the size of government appear more responsive to real or perceived changes in the political environment.

In a 2007 Pew analysis, Americans overwhelmingly supported the social safety net and an activist government, with 69% of respondents agreeing both that “the government should guarantee every citizen enough to eat and a place to sleep” and “it is the responsibility of the government to take care of people who can’t take care of themselves.”34 At the time, Pew noted that this was a rebound in support for the safety net from its low point in 1994, when support for these statements was 59% and 57% respectively. But it turned out that 2007 was a high point. Just four years later in 2011, Pew found that only 56% of Americans supported those two statements—a new low.35 Rather than consistent views about an activist government and safety net spending, Americans’ views shift over time. Further, there are generational divisions over safety net support, but these also fluctuate. In 2007, 82% of Millennials agreed that the government should guarantee food and shelter; just 4 years later that had fallen by nearly 20 points to 63%. Given these fluctuations, it is foolish to believe that a generation’s viewpoint in one particular election cycle, or even two, portends their views for all time to come.

Government Guaranteeing Food & Shelter

Helping the Needy

Potentially more troubling for Democrats as the party of government, perceptions of waste and mistrust can undercut support for an activist government. Both the Silent and Baby Boomer Generations have been skeptical of the government for decades, with majorities routinely saying when the government runs something it is “usually inefficient and wasteful.” But there has been significant movement over the years—for example, between 1987 and 2011 the Silent Generation fluctuated nearly 20 points. Boomers displayed nearly as much of an increase in skepticism, with 61% saying the government is usually inefficient and wasteful in 1987 and a full 76% agreeing in 2011.

Gen Xers and Millennials have generally displayed more optimism than their older counterparts, but their views have also been malleable. In 1987, less than half (45%) of Gen Xers said the government was usually inefficient and wasteful. After fluctuating in the 1990s, that figure was 43% in 2002, but it rose a whopping 10 points just one year later and stood at 57% in 2011. While Millennials exhibit the most positive assessments about government of any generation so far, their views have shifted as well. In 2003, 31% said government was usually inefficient and wasteful; that rose to 42% in 2009 and 51% (up 20 points from just 8 years before) in 2011.36 As views about government efficiency and waste fluctuate, it’s likely we’ll see continued variation in support for expanding the government’s role—not simply reflexive support for activist government.

Government is Inefficient and Wasteful

Similarly, trust in government has also fluctuated over the years. In 1997, nearly four in ten Americans said they could trust the government in Washington to do what is right always or most of the time. By 2011, that number was down to 20%. While drops in support were evident—and tracked similarly—for Silents, Boomers, and Gen Xers, the most dramatic recent changes have come from Millennials. In 2004, 44% of Millenials said they trusted the government all or most of the time. By 2011 (less than 2 presidential election cycles later), support had dropped 18 points to 26%.37 This data suggests that even when controlling for potential cohort effects (the notion that the events and trends of a generation’s time period exert deep and lasting influence on their political leanings long-term), political values about the role of government fluctuate and are likely responsive to perceptions about real world effects as opposed to deep, quasi-genetic attachments to a worldview.

Trust in Government

Fluctuating views about the government are evident in specific policy positions as well. In 2011, 49% of Americans said the Patient Protection and Affordable Care Act (PPACA) was a good start, 30% said it was a step in the wrong direction, and 21% had no opinion. Among all generational cohorts, more people responded that it was a good start over a step in the wrong direction—each by at least 10 points! But just one year later, only 38% thought the PPACA was a good start and a whopping 34% were ambivalent. Declines in support were evident among all generational cohorts, but the largest shift came from Millennials, whose support dropped by 14 points—with increases among those without an opinion to nearly four in ten.38

Views of the Patient Protection and Affordable Care Act by Generation

* Note: In this survey, everyone older than the Boomers was labeled Seniors and lumped together, technically combining the Silent and Greatest generations.

As the data above demonstrates, Democrats cannot merely assume that the new American electorate supports an activist government—rather, they will have to demonstrate tangible benefits, modernize and adapt government to the changing environment, and avoid complacency to continue to garner support for government programs.

Illusion #5: The culture wars are over

In one area, many Democrats are certain they’ve already won the Millennial Generation for good: the culture wars. Proponents of this view argue that gay equality is a proxy for the culture wars broadly, and that Millennials aren’t buying what the Right is selling. They note the vast differences between Millennials and Silents, as well as between whites—particularly white males—and nonwhites and women. But beyond gay equality, the evidence that the culture wars are dead is mixed, at best.

Indeed, culture war issues continue to emerge and be divisive because they represent a constant tension between freedom and order, between a license to behave as one wishes and the value of structure and stability undergirding society. For example, President Nixon responded to the permissiveness of the 1960s by emphasizing law-and-order. The permissiveness associated with the 1960s was itself a reaction to the law and order priorities of the Eisenhower Administration. This tension is ongoing. There is no final winning or triumphant side. Rather, conservatives will continue to emphasize tradition and order while liberals emphasize modernity and freedom. Who is ascendant in balancing these interests will vary, largely in reaction to real or perceived unbalances going forward.

Two Different Types of Culture War Issues: Marriage v. Abortion

While views of all Americans—old and young, Democratic and Republican, Evangelical and unaffiliated—have changed recently toward support for marriage for gay couples, public opinion surrounding other hot-button culture issues has remained largely stable, not reflecting a surge in the liberal position. This is particularly true on the issue of abortion.

Gallup has tracked American’s support for legal abortion since 1975. While there has been some variation over the past 38 years, more notable has been the remarkable stability in the country’s views. For example, in 1975, 21% of respondents said abortion should be legal under any circumstances, 54% said legal only under certain circumstances, and 22% said illegal in all circumstances. In 2012, 28% of respondents said abortion should be legal under any circumstances, 52% said legal only under certain circumstances, and 18% said illegal in all circumstances.39

Legality of Abortion, 1975-2012

Whereas opinions on abortion seem relatively stable, positions on marriage for gay couples are in significant flux. Support for marriage is certainly more pronounced among Millennials than among the Silent Generation. However, since 2004, we have seen major shifts in viewpoints among all groups, including the oldest Americans. Between 2004 and 2011, overall public support for marriage equality rose 16 points. But while one-quarter of that change stemmed from younger voters replacing older ones in the population, three-quarters of the shift in public opinion has come from people of every age changing their minds on the issue. And these trends are evident in all demographic subgroups between 2004 and 2011:

  • Republicans +10.6 points;
  • Independents +15.1 points;
  • Conservatives +13.6 points;
  • Moderates +21 points;
  • Evangelicals + 8.3 points;
  • Weekly church-goers +12 points;
  • People without a high school degree +9.2 points;
  • People born prior to 1940 +13.1 points; and,
  • People born between 1940 and 1949 +14.4 points.40

While the country as a whole is moving towards greater acceptance of marriage equality for gay couples, with Millennials leading the way, public opinion surrounding abortion doesn’t exhibit the same tendencies—even when comparing generational cohorts. For example, in 2012, one poll showed that 49% of Americans favored marriage equality (45% opposed), including 68% of Millennials (and only 31% of those 65 and over). While 54% of Catholics supported marriage, that figure rose to two-thirds for Catholic Millennials. And while 73% of white Evangelical Protestants opposed marriage for gay couples, only 69% of white Evangelical Protestant Millennials held similar views.41 In fact, younger voters were more supportive of marriage in every demographic group.

By contrast, on abortion we find relative stability in the percent favoring legality across generational cohorts—and if anything, perhaps even a rightward shift among some subsections of Millennials. For example, 56% of Americans and 54% of Millennials believe abortion should be legal in all or most cases. Similarly, 48% of Catholic Millennials and 53% of all Catholics believe abortion should be legal in all or most cases. White Evangelical Protestants overwhelmingly oppose abortion, with 64% saying it should be illegal in all or most cases. Among white Evangelical Protestant Millennials, a whopping 88% believe abortion should be illegal in all or most cases.42 Divisions arise when comparing views on marriage and abortion in Pew’s analysis as well. In their pre-election analysis in 2011, they found considerable stability across generational cohorts in views towards abortion with big divergences on marriage.43

Allowing Gays to Marry & Abortion Should be

Marriage and abortion should not be conflated as interchangeable “culture war” issues. While the former has seen rapid acceptance throughout American society, the latter is still a battlefield, even among Millennials.

Changing Nature of Culture War Issues

Over the years, the menu of so-called culture war issues has evolved. They tend to arise every few years and are unpredictable. Sometimes issues have faded from the public eye entirely. (Remember flag burning, term limits, prayer in schools, banning books, and burning draft cards?) On other issues, public opinion seems to have durably shifted. While interracial dating and interracial marriage were divisive topics for decades—in 1958, only 4% approved of the former—now there is virtually unanimous support, with 86% approving of marriages between blacks and whites.44 Women working outside of the home used to be controversial—in 1986, only 45% felt it was good for marriages and 24% for children. But by 2012, 75% said a woman working outside of the home was good for marriages, 52% for children and 81% for society in general.45

Other culture war issues are cyclical, with both public opinion and their prominence in public debates waxing and waning over time. For example, support for gun safety measures was well over 50% in the 1990s, peaking at 67% in 2000 before falling below 50% in 2009.46 In 2001, 47% of Americans said strengthening the nation’s gun safety laws was a top priority, but in 2013 that number had fallen to 37%.47 Reducing crime was a top priority in the latter part of the 20th century, with well over 70% agreeing. By 2011 though, only 44% of Americans said crime was a top priority. Concerns on that front seem to be rising somewhat, with 55% now saying that crime reduction should be a top priority for Congress and the President.48 And in 2007, 55% said dealing with illegal immigration should be a top priority, but in 2013 that fell to 39%.49

The 1960s were characterized by political struggles over war, racial integration, and social structures. The Voting Rights Act and affirmative action—both legacies of the battle for civil rights—have been weakened by recent Supreme Court decisions and could be severely limited in the coming decade. But these issues do not necessarily resonate with today’s younger generation like they did 50 years ago. Among younger Millennials, 63% do not believe that race or gender will impact their future career prospects.50 And in a recent survey, only 23% of those under 40 years old supported using race as a factor in university admissions, with a whopping 62% strongly opposed.51 Rather than assume the culture wars are over, we should acknowledge the ways in which old issues can recede and new issues can emerge.

Recent moves by Democratic governors to outlaw the death penalty demonstrates the ability of old issues to resurface. Increasing public support for medicinal and recreational use of marijuana—coupled with the costs associated with the War on Drugs—may thrust legalization into mainstream debates, conjuring a new front in the culture wars. And whether related to marriage for gay couples or contraception coverage in employer-provided health care, religious liberty issues are bubbling to the surface and will likely spark heated debates in the coming years, focusing on individual Free Exercise claims rather than the Establishment Clause (prayer in schools) fights of the 1980s and 1990s. Despite the distinct views of Millennials and their increasing share of the electorate, the culture wars aren’t going away. They are simply shifting ground.

Furthermore, unidirectional progressive movement on some issues does not guarantee the same in other areas. Perhaps more frequently, attitudes ebb and flow. Consider marijuana legalization. In 1978, 31% of the general public supported legalizing marijuana, including a record high 49% of Baby Boomers. Overall support fell to 16% in 1990—and only 18% among Baby Boomers.52 Twenty-three years later, support reached 52% among the general public, including a new high of 50% of Boomers.53 The counterculture generation may have popularized marijuana, but they have resisted being defined by it.

Marijuana Legalization by Generation

As mentioned above, prior battles over religious liberty have focused primarily on limiting government actions that constitute or suggest the establishment of a national religion. But the landscape is now shifting to questions of free exercise of religious beliefs and even swung towards accusations of government hostility to religion. Free exercise, in the past more typically employed to protect minority religious practices, is now being used frequently by Catholic and Evangelical Protestant groups to claim that the government is infringing on their rights (e.g., contraception coverage in the PPACA). We know that Millennials support the current understanding of the Establishment Clause and would not likely warm to complaints about the lack of prayer in schools. But how they will respond to issues framed as infringing on an individual’s free exercise of their own religion remains to be seen.

The debate in 2012 over contraception and birth control coverage in PPACA is demonstrative of this point. Overall, 55% of Americans supported the Administration’s decision that most employer health care plans must provide no cost contraception and birth control, with only 40% opposed to the requirement. However, when asked about exemptions for religiously-affiliated colleges and hospitals, support for the contraception requirement dropped to 49%, with 46% opposed. Among Millennials, support fell from 65% to 58%. And if churches or other places of worship were included, 57% of Millennials opposed the requirement for contraception coverage while only 36% supported it.54 The report of the death of the culture wars may be an exaggeration.

Illusion #6: Hispanic and Asian voters are homogeneous

President Obama received overwhelming support from Hispanic and Asian voters in 2012. Yet neither group has necessarily had cohesive political identity historically. Rather, their lopsided support for Democrats and internal cohesion appears as a recent phenomenon and one that may not be a constant.

The battle over immigration reform is illustrative. The marches and other social movement actions that became widespread in 2006 united disparate segments, particularly within the Hispanic community. Evidence for this can be found, for example, in pre- and post-movement surveys, whereby policy differences among 3rd generation Hispanics and more recent arrivals virtually vanished.55 While Asians and Hispanics appear internally cohesive at this moment in time, are they to remain united in political outlook and partisan support in the future? Two facets of these communities bear consideration.

First, the “Hispanic” and “Asian” labels are not sticky within these communities. Less than one-quarter of Hispanics primarily use the label “Hispanic” or “Latino.”56 Fewer than twenty percent of Asians self-identify as “Asian” or “Asian American.”57 Rather, they are likely to identify based on their ancestral country of origin or simply call themselves Americans. By considering all Asians to be similar, we mask their diverse experiences, languages, and traditions. And we fail to consider the implications of the immigrant experience and integration into the U.S.

Second, a considerable range of opinion is found within these communities, with patterns emerging based on time in the U.S. Hispanics and Asians born in the U.S. or who are a third generation (or longer) American tend to display attitudes more in line with the national average. For example, 81% of recent immigrants and 72% of 2nd generation Hispanics support a bigger government with more services, but only 58% of third-generation Hispanics agree with that statement.58 As people integrate into communities, they may also assimilate prevalent values. In doing so, these communities could become swing voters, like Catholics, rather than a more homogeneous group who overwhelmingly support one party.

Illusion #7: The presidential map will continue to favor Democrats

In detailing why the key growing demographic groups are so important in elections, the proponents of the idea that demography equals destiny for Democrats note that these growing groups of voters are geographically situated in purple states. Thus, they say, these contested states will turn reliably blue in upcoming elections, locking up the Electoral College for the foreseeable future for the Democratic Party. But that analysis also assumes that once states go blue, they won’t ever go red. That is, progressive dominance spreads but never retreats. And it presumes that swing states which have historically leaned toward the Democratic Party will continue to do so in the future—regardless of their own demographic shifts.

While Democrats may display dominance in presidential voting in some key states, the data does not suggest that all of the so-called “blue firewall states” have turned permanently blue or are solidly progressive. Both Michigan and Pennsylvania have voted Democratic in every presidential election since 1992. Yet at the state level, the governorship and both chambers of the legislature are under Republican control. Wisconsin has voted Democratic in every presidential election since 1988, but there, as well, the state government is currently under unified Republican control—despite President Obama winning the state twice. And in deep blue liberal New Jersey, the GOP may be making a comeback, led by popular Governor Chris Christie. Indeed, Republican governors have been elected in many “blue” states in recent years, including Maine and New Mexico. Even California had a Republican governor, despite being a reliable Democratic state in presidential voting since 1992. Assuming that every one of these states will stay in the Democratic column despite all the other changes in the national electorate is short-sighted.

It is also not clear that purple states are becoming less “swingy,” or leaning toward Democrats in future elections. Ohio is still a presidential prize, voting for the victor in every year since 1944 (except 1960). And Ohio is currently under unified Republican control. Florida, Virginia, and North Carolina—often labeled the “new South”—all voted Democratic in 2008 (and Florida and Virginia in 2012 as well). But Florida and Virginia are also under unified Republican control at the state level. Rather than progressive or Democratic dominance, the picture in these states is muddied, with many current splits between blue presidential voting and red state governments.

There are two implications of this ticket splitting. First, it does not seem that a single party or ideological view is necessarily dominant in these states (consider, e.g., Walker-Obama voters in Wisconsin). Second, electing Congressional Democrats even from blue presidential states is not guaranteed, which could have a major impact on both Democratic success in governing and how voters perceive government and the Democratic Party.

While it is true that the white, non-Hispanic population of the U.S. is shrinking—and that this group tends to vote Republican in the modern era—the demographic trends are not as pronounced in several purple states. Between the 2000 and 2010 U.S. censuses, the percentage of the population that was Hispanic increased by 3.8 points to 16.3% nationwide, while the share of the white, non-Hispanic population fell by 5.4 points to 63.7%.59 But in competitive states such as Wisconsin, Ohio, and Iowa, demographic changes are occurring at a much slower pace than the national average. If the rising demographic groups are concentrated in a handful of states, their impact will be limited in national elections.

Population Changes by Race and State, 2000–2010

A similar phenomenon is occurring when it comes to the distribution of age cohorts. In recent elections, younger voters have tended to back Democrats, while seniors have been reliable Republican Party voters for the past 30-plus years. But the Democratic Party may have an emerging age problem in some traditionally blue states, where the percentage of people 65 and over is increasing, and often at a faster pace than the national average.60 Purple states, such as Michigan and New Hampshire, are aging faster than the nation as a whole. And if demography truly equals destiny, then some blue states could be put back into play, since both Maine and Minnesota are aging more rapidly than the national average. Given the recent pattern of older voters supporting Republicans, the state picture muddies the impact of the rising Millennial Generation on national elections.

Population Changes 65+, 2000–2010

Democrats have also fared better than Republicans among union members, winning them by approximately 60% to 40% over the past few presidential elections. However, union density is also on the decline in several blue states, again cutting into a demographic group that has tended to support Democrats.61 Some of this is due to restrictive (e.g., “right-to-work”) legislation aimed at undercutting the ability for unions to legally form (e.g., Wisconsin). But declining union membership is also likely driven by the changing economy, which boasts fewer traditional manufacturing positions.

Union Density, 2000–2012

So it seems that demographic changes may be pushing the electorate in both directions—especially in many of purple or traditionally blue states. And it’s unclear which direction will win out in the future.


Hispanic, Asian, and Millennial voters have leaned Democratic for a few cycles. Democratic prospects for winning future presidential elections rests upon whether the current coalition is sustainable and loyal to the Party brand. We have two conflicting models in black and Catholic voters. But it’s important to note that in terms of these rising demographic groups, Republicans aren’t standing idly by but are actively trying to ascertain how to win them over. And history suggests they will have opportunities to capture at least some of these voters, which could be enough to propel them to the White House.

Consider this: Ronald Reagan won 489 electoral votes and 44 states in 1980, and 525 electoral votes and 49 states in 1984. George H. W. Bush won 426 electoral votes and 40 states in 1988. In the late 1980s the Democratic Party appeared dead. Then came Bill Clinton, a pro-reform, pro-business, pro-trade Southern Democrat. The Democratic Party won two presidential elections in a row—marking the first time since FDR. Given how aware Republicans are of their party’s problems with younger and nonwhite voters, we can only assume that they will course-correct at some point and become nationally competitive again.

If Republicans continued to attract large numbers of white, older, and male voters, they wouldn’t need to win other groups of voters outright. Indeed, many Democratic voters will likely remain brand loyal. But shaving off a few points among key groups could vastly improve Republican chances at winning national elections. Former President George W. Bush did just that in 2004, losing Hispanics and women narrowly while winning reelection. Indeed, if we look just at the national popular vote, President Bush would have won reelection in 2004 even if the electorate mirrored the racial and ethnic distribution of 2012. In fact, if President Bush would have won racial and ethnic groups by the same percent as he did in 2004, Bush would have won the 2012 election 49.68% to 49.02%.62

Demographics may not be destiny after all.

Appendix A

Exit Poll Citations

Exit Poll Citations

  • American Electorate220


Get updates whenever new content is added. We’ll never share your email with anyone.