[Overview] Trump Voters I – Demographics

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa

Welcome to the first Overview at Veracities! This will be a regular feature where we summarize the current state of research on various topics we think are interesting. We’ll take as broad a view as possible to give you an overall picture of what is and isn’t yet understood. First up: the demographics of Trump support.

Something altogether unprecedented is happening in American politics, unless it isn’t. A new bloc of voters have been awakened from their quiescent slumbers, unless they haven’t. This bloc heralds the rise of a reinvigorated Republican Party, unless it threatens to tear the party apart. Much of the American public is surrendering to an authoritarian or even Fascist impulse, unless they aren’t. Trump voters are either certainly X or certainly not-X.

Many (mostly unflattering) comparisons have been made to Donald Trump since his official entry into American politics last year, from Andrew Jackson to George Wallace to modern European right-wing populists and even to Hitler. So while it isn’t clear that Trump is wholly unique, he is clearly an anomaly deserving explanation.

More specifically, I want to look closely at who is voting for Trump. What are their demographics? What are their beliefs? Their motivations? How do they see America and the world? I’ll begin by looking at basic demographic factors and then move on to broader social factors. Then I’ll talk about the ideas and ideologies that have been attributed to Trump voters and how well supported these attributions are. I’m not going to be able to disentangle the entire web of causation and correlation. I won’t be saying “23% of Trump’s support is due to a loss of manufacturing jobs”. Instead, I hope to illustrate a world and a worldview. (Spoiler: Trump supporters are right wing populists not just motivated by economic concerns.)

I’ll mainly talk about his support during the primaries, partly because that’s where the data is and partly because as nominee he’s likely to win even those Republicans who aren’t natural Trump supporters, muddying the difference between Trumpists and other Republicans. I’m not trying to explain the demographics of the Republican Party itself, so this article will focus mainly on how Trump support differs from that of other candidates, rather than from the nation as a whole. In general, he is somewhat popular among Republican groups and very unpopular among everyone else, but there are of course differences in degree.

Over time, Trump’s support among Republicans has grown from around 25% in late 2015 to around 40% in late April, which means that Trump’s coalition has broadened. The research and articles cited here come from different points in time, so their numbers may not always be directly comparable. But for creating an overall impression, hopefully that won’t matter too much. I have, however, mostly limited myself to sources from after the first primaries and before Trump became the presumptive nominee after winning Indiana.

Trump voters within the Republican Party

To date, Trump has received 10,999,775 votes, 40.65% of the total. While the absolute number is comparable to those of past nominees, because turnout has been higher this year, his share is considerably lower, and he is on track to become the first Republican nominee in decades who hasn’t won a majority of popular votes. Who, then, makes up his victorious plurality?

This Wall Street Journal infographic is an excellent introduction. It breaks down support for Trump along a number of dimensions in comparison to “Establishment Voters” supporting Bush and Rubio and “Social Conservative Voters” supporting Cruz and Carson. (How quickly time flies in campaign season!) Of the three groups, Trump supporters are least likely to have college degrees, least likely to have household incomes of over $75,000, and least likely to regularly attend church.

Only 31% call themselves “very conservative” versus 51% of social conservatives and 20% of establishment supporters. Nevertheless, they oppose same-sex marriage and support gun rights, though they are the least pro-life of the three groups.

They are the group most skeptical of the rest of the world. 81% of Trump voters say that immigration hurts America more than it helps, versus 60% of Social Conservative Voters and 45% of Establishment Voters. Trump voters are also most skeptical of free trade and most inclined to focus on “problems at home rather than overseas”.

That should give an overall picture. Poorer, alienated from institutions, focused on America. But note that he draws support from both sides of each of these divides. That will be a recurring theme: while one can say that Trump voters are more or less concentrated in a given group, his support is quite broad, at least among Republicans. The typical Trump voter is not a stereotypical Trump voter. It’s at most useful to talk about tendencies rather than make categorical statements.

Live vs. online polling

Trump does “about five percentage points” better in online polls than when talking to a live interviewer, possibly due to social desirability bias: “people are afraid to tell another human being that they support the Republican because, even though they like him, they know about his controversial statements and do not want to be judged negatively.” The gap was larger, closer to 10 points, for voters with bachelor’s or graduate degrees, and for more informed voters.



Trump is often accused of sexism and he has a history of controversial statements towards women and it’s reflected in the polls. He’s very unpopular overall, but there is a clear gender gap: 58% of men and 70% of women view him unfavorably. “The gap between his favorable and unfavorable rating among men averaged 22 percentage points in March 1-28 interviewing, compared with a 47-point gap among women.” Also, while his numbers have been declining for months, the decline has been roughly the same among both men and women; he simply started off doing better among men.

For comparison:

According to a national Quinnipiac University survey in March, only 29 percent of women have a favorable view of Cruz, while 46 percent have an unfavorable view. A CNN-ORC poll the same month found that 28 percent of the women surveyed overall had a favorable view overall of Cruz, and 58 percent had an unfavorable view.

Still, in both surveys, Trump’s numbers were considerably worse: His unfavorability rating hitting 67 percent in the Quinnipiac poll, and 74 percent in the CNN survey.

Let’s focus on the GOP. Among Republicans, Trump is the only candidate with a significant difference in how many men and women wouldn’t consider supporting him:


Of course this is just for the primary; most of these people will wind up voting in the general for whoever the nominee is, regardless of what they say now. And in any case, according to exit polls, Trump still won a plurality of women voters in most states, though consistently smaller than his pluralities among men:

Capture19(Data from here; graph created by me; not all states exit polled)

(Most primaries were pretty evenly split between male and female voters. Except for New York and Connecticut, men never made up more than 53% of the electorate.)

Alabama produced an anomalous gap of 16 points, but in general the difference in support hovered between 5 and 10%. Ted Cruz, on the other hand, polled more evenly between men and women, though generally behind Trump among both.

Overall, Trump’s poor standing with women has cost him some votes in the primaries, but will be a larger issue in the general election.


In December, a Wall Street Journal/NBC News poll found that “Among those aged 18 through 34, just 19% have positive feelings about Mr. Trump. Among those aged 50 through 64, the level rises to 35%. It’s 32% among those aged 65 and over.”

Among Republicans specifically there is also an age gap, though it’s not huge:

Capture7 (Source)

Exit polls have found a similar difference:


(Data from here; graph created by me; not all states exit polled)

The youngest voters were generally least supportive of Trump, but otherwise things are close. In no race were 18-29-year-olds most likely to vote for Trump. In only one race, Arkansas, were those aged 30-44 most likely, excluding ties. The elderly and most especially the middle-aged are clearly likeliest to support him. For the most part the gap between least- and most-likely groups is under 15 points, and often 10.

However, these groups did not make up equal shares of the electorate. Roughly 10+% of the electorate was 18-29, 20% was 30-44, 40+% was 45-64, and 20+% was 65+, with variation by state. Therefore, a large plurality of Trump voters were middle aged:


(Data from here; chart created by me; not all states exit polled; percentages may be slightly off for states where the 18-29 demographic was too small to be counted)

Again though, we can only speak of tendencies: Trump draws support from many groups of Republicans.


Trump is wildly unpopular with non-white voters overall. His net favorability with Hispanics is -65%, compared to other Republicans, who roughly break even (and to +33% for Clinton). One “tracking poll found that Trump was viewed unfavorably by 86% of black voters and 75% of Latinos.” An “NBC/Wall Street Journal survey found two-thirds of voters overall saying that they could not see themselves voting for Trump. Among nonwhite voters, the figure was 84%.”

But what about among non-white Republicans? Have they helped or hurt Trump?

It’s important to understand that a huge majority of Republican primary voters are non-Hispanic whites. In 2012, self-identified Republicans were 89% non-Hispanic white, 2% non-Hispanic black, 6% Hispanic, and 1% Asian. Trump hasn’t changed that. Of the 26 states that were exit polled, in only 5 states were non-Hispanic whites less than 90% of the Republican electorate: Nevada (85%), Georgia (88%), Texas (82%), Virginia (86%), and Florida (78%). Along with North Carolina and Ohio, these are the only states where non-Hispanic white Republicans have been analyzed separately in these exit polls.

Trump won the mainly Latino minority vote in Nevada with 42%. In every other state with available data, he lost Hispanic and non-White voters. Given that turnout is up overall while the minority share of the vote hasn’t changed much, there have obviously been more minority voters in the primaries to date. However, these voters have mainly opposed Trump. He is losing both non-white Republicans and non-whites overall.

New voters?

Republican turnout has been dramatically higher than in 2012 – some 60% – this primary season. (Although note that this doesn’t necessarily translate to better turnout in the general election.) One reason for this is simply that the contest has been seriously contested for longer than in 2012. (Democratic turnout is down 19% from 2008, though it’s just a two person race and Sanders has never had Obama’s odds of defeating Clinton.) Who are these new voters? New Republicans? Loosely attached Republicans?

While turnout has indeed been high this cycle, it’s not clear that Trump is really expanding the party:

Republicans are seeing their voter registration rolls grow instead of shrink for the first time in years at this stage in the federal election cycle, but not by much. “The Republican Party share of the voter registration in the country is up slightly, as compared to the last tally, which was November 2014,” said Richard Winger, editor of Ballot Access News. “It’s the first time the Republicans have gone up instead of down since 2004.”

In February 2016, there were about 30.5 million registered Republicans nationally, Ballot Access News reported. That’s compared to 30.9 million in October 2014, 31.3 million in October 2012, 30.9 million in October 2008, and 29.0 million in October 2004.

Most partisans don’t vote in primaries, so increased turnout can easily come mostly from newly energized Republicans. Trump has undoubtedly drawn less attached Republicans to come to the polls, but remember that Trump has mostly only won pluralities rather than majorities, so some amount of increased turnout is due to the anti-Trump vote.

Unfortunately, most exit polls haven’t asked people if it’s their first time voting in primaries, so we don’t have a full picture of who Trump is bringing in. According to the Los Angeles Times:

In Iowa’s caucuses this year, Trump clearly did well with first-time participants. The share of voters taking part in the state’s caucuses for the first time rose from 38% in the 2012 exit poll to 45% in 2016, and Trump won the first-timers handily. In New Hampshire, by contrast, the 15% of voters casting a GOP primary ballot for the first time this year was only marginally higher than the 12% four years ago, and Trump’s share of their votes, 38%, was only slightly better than the 35% he got among repeat voters.


Polls show that moderate Republicans — not Democrats and independents, as Trump often claims — have also helped boost turnout numbers in 2016.

According to the analysis by Public Opinion Strategies, which included the most recent round of voting on Tuesday in Utah, Arizona, and Idaho, 42 percent of Republican primary voters so far this year identify themselves as “somewhat conservative,” up from 33 percent at this stage in the 2012 race.

As Trump supporters are less likely to self-describe as strongly conservative than other Republicans, this change is plausibly due to Trump bringing in new voters to the primaries.

Democratic defectors? Independents?

Anecdotes abound about voters torn between Trump and Sanders, and pundits have wondered if Trump, not a doctrinaire conservative, might attract Democrats. One poll in January found that 20% of Democrats would vote for him – but also that 14% of Republicans would vote for Clinton! However, this poll is most likely an anomaly:

According to the Reuters tracking poll, which follows responses over time to measure changes in support or opinion, more than 73 percent of whites who voted for President Obama in the last presidential election hold a negative opinion of Trump, versus 27 percent who hold a favorable opinion. It’s difficult to do a direct comparison, but this is in line with Mitt Romney’s standing among white self-identified liberals in the September before the 2012 election. Better is an apples-to-apples comparison between Trump and his nearest rival, Ted Cruz. Whites who voted for Obama like Cruz—a doctrinaire conservative ideologue—about as much as they like Trump, 28 percent favorable to 72 percent unfavorable.


A March Washington Post poll found that in a hypothetical matchup with Trump, Hillary Clinton wins Democrats 86 to 9 percent. That means, right now, that Trump does indeed gain a few more Democratic defections than Romney did. The problem is that Trump is only winning Republicans 75 to 14 percent. In other words, more Republicans are planning to vote for Hillary than Democrats are planning to vote for Trump, which helps explain why Clinton is leading in almost all the head-to-head polls.

(Trump has gained Republican support since clinching the nomination.) While this election is itself anomalous and Trump is relatively unpredictable, the weight of evidence to date suggests that Trump lacks appeal to both Democrats and independents. And in fact, votes to date back this up: “The share of the GOP primary electorate that identified as Democrat stayed steady at 5 percent, while the percentage of independents grew one point, from 26 to 27 percent.”


Again, increased turnout means that more Democrats and independents must have voted Republican, but the numbers are small overall. In no way could they be a major part of Trump’s base. His main support comes from Republicans, both party stalwarts and especially more weakly attached voters, not from drawing in non-Republicans.


Like for the GOP as a whole, support for Trump is overwhelmingly white. Relative to other Republicans, Trump voters are older and more male, but Trump often won among women and the young regardless. He may be drawing in less attached Republicans, but independents and Democrats haven’t much helped him.

Next: Where are Trump voters from? What are their ethnic backgrounds?

(Source for header image)

[Overview] Trump Voters II – Geography and Culture

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa





As you can see in these county-level maps from the New York Times, while Trump’s support rarely dips below 20% outside of the Mormon West, there is nevertheless significant regional variation in his support. (Ignore Wyoming, which has an odd caucus system. West Virginia and Nebraska voted after everyone else had dropped out.) Some states, Texas and Ohio, he lost to a favorite son. He also has notable pockets of weakness in the Great Plains and Midwest. He does well in the Deep South, mid-Atlantic, and parts of the West. For the most part, his support goes from 20-something percent to 50-something percent, a more significant spread than for the other factors I’ve looked at so far.

However, this regional map is in some ways distorting, as some states are much more Republican than others and some states much more populous than others. (A map distorted by population density would be useful.) This is a graph of Trump voters relative to voters for non-Trump Republican candidates, and to the rest of a state’s population:

image (1)

(Data from 1, 2, 3; graph created by me; Colorado, Wyoming, North Dakota, states that voted after every other candidate had withdrawn, and all territories are excluded)

The first striking thing is how small many of these primaries are, even in Republican states in a high-turnout year. (And the restrictive and time-consuming caucuses are even tinier.) But Trump’s support is clearly smaller in the Northeast than elsewhere; he’s dominating a smaller party. Many of these primaries come down to a few thousand ballots out of only a few tens of thousands cast.

But what drives this regional variation among Republicans? 

The places where Trump has done well cut across many of the usual fault lines of American politics — North and South, liberal and conservative, rural and suburban. One element common to a significant share of his supporters is that they have largely missed the generation-long transition of the United States away from manufacturing and into a diverse, information-driven economy deeply intertwined with the rest of the world.

They “compared hundreds of demographic and economic variables from census data, along with results from past elections, with this year’s results in the 23 states that have held primaries and caucuses.” What they found:

Capture (Source)

I’ll cover this more in the next section, but what “American” ancestry means is that, when asked to describe their background, some people simply say “American” rather than naming a country in Europe or wherever. In practice, the people who do so are white Americans mainly from Greater Appalachia and the Deep South. Also, I’m pretty sure that the authors are implicitly dividing whites into three groups: those with Catholic ancestry, those who report”American” ancestry (who are mostly actually White Anglo-Saxon Protestants), and others. They call this last group WASPs, but I believe it also includes people with German and Scandinavian heritage. Of these three groups, those with Catholic and “American” ancestry are the ones supporting Trump.

Brian Arbour and Jeremy M. Teigen did their own, similar analysis. (Note that their term for “American” ancestry is “unhyphenated American”.) They found similar correlations for ancestry, less education, and worse economic conditions:


Notably, these two studies differ, at least in interpretation, on the impact of race and immigration. Irwin and Katz:

Despite evidence that some individual Trump voters are driven by racial hostility, this analysis didn’t show a particularly powerful relationship between the racial breakdown of a county and its likelihood of voting for Trump. There are Trump-supporting counties with both very high and very low proportions of African-Americans, for example.

One of the strongest predictors of Trump support is the proportion of the population that is native-born. Relatively few people in the places where Trump is strong are immigrants — and, as their answers on their ancestry reveal, they very much wear Americanness on their sleeve.

Versus Arbour and Teigen:

Interestingly, Trump’s support increases as the percentage of African Americans and immigrants in a county increases. Trump and his campaign have been criticized for bashing immigrants and for being slow to disavow the Ku Klux Klan’s support. We doubt, then, that blacks and immigrants are the ones voting for him. Rather, apparently those unhyphenated white GOP primary voters are pulling the lever for Trump’s anti-immigrant and anti-minority platform and attitudes specifically because they feel threatened by those “others.” Some scholars refer to it as the “racial threat” hypothesis.

Preliminary research by one of us (Arbour) shows that regions with concentrations of unhyphenated Americans also voice more racial resentment than the rest of the country — and even more than in the rest of the South.

Unfortunately, neither article provides enough information to resolve this discrepancy. Perhaps the relationship between local diversity and Trump support is different in some parts of the country in a way that isn’t being captured by these studies, but that’s pure speculation on my part. I’m also not sure why the second study finds that Latino and immigrant populations have opposite effects on Trump support.

The Washington Post has an interesting interactive graphic that lets you see how well the candidates did in each county relative to various measures like race, education, income, and so on, with more states included than in the above two analyses. It’s a bit confusing, but if you can draw a line that fits through the red dots from bottom left to upper right, that indicates a positive correlation with Trump’s vote share. Here’s a screenshot of Trump’s support relative to the black population in a county:


The author says:

Trump’s success in heavily black counties isn’t a function of black support; this is the total population of the county that’s depicted. That strength in black counties is also why he seems to do obviously better in poorer counties — until you limit that to only the white population. Then, the apparent link drops significantly.

So this agrees with the other Washington Post analysis: (white) Republicans vote Trump in counties that are more black. However, it doesn’t seem to be the main factor. “This isn’t like the Democratic contest, where knowing the percentage of the population that’s black can give you a good sense of who will win.”

Trump country is also relatively poor and uneducated. (But again, we’re just talking about relative concentration of support!) You can find profiles of these ultra-Trumpist areas. As the Times says, these regions have been suffering for a long while:

There were only weak correlations between Trump support and various measures of economic performance from 2007 to 2014, including the lingering damage from the 2008 global financial crisis. Rather, the economic problems that line up with strong Trump support have long been in the making, and defy simple fixes….

In places where Trump does well, relatively high proportions of workers are in fields that involve working with one’s hands, especially manufacturing. The decline in manufacturing employment is not a story of merely a rough few years for the economy; nationwide factory employment peaked in 1979, and as a proportion of total jobs has been declining almost continually since 1943. Forces including mechanization and trade have put employment prospects in the sector in an ever-worsening position.

Note that this analysis only shows the social conditions correlated with Trump support, it doesn’t show which voters in these counties are the ones voting for him. For instance, it could be the relatively well-to-do, while the poorest aren’t voting at all. Yes, Trump voters are poorer and less educated than other Republicans, but that doesn’t mean that they’re poor themselves. At any rate, it’s a caveat worth keeping in mind.

And as I quoted above, the relationship between support for Trump and poverty is attenuated when you only look at the white poverty rate, which weakens the case that Trump voters are themselves necessarily disproportionately poor:


At least some of the relationship between Trump support and poverty is actually the relationship between Trump support and race. Further research into how race and the economy tie together at the county level would likely be very informative.

White ethnicity

Understandably, there has been much discussion about Trump and race. But there has been relatively little discussion of Trump and ethnicity. Specifically, whether support varies by ancestry among whites. Let’s return to the gap between unhyphenated Americans/those with “American” ancestry and everyone else.


According to the Washington Post:

The people who identify this way are concentrated in Appalachia and rural areas in and around the South. They are heavily Protestant and less likely than average to have graduated college. They cluster especially along the migration path taken by highland Southerners before the Civil War — those farmers from the hardscrabble Southern hills who never had enough money to buy land in the fertile parts of the Deep South and who moved from the Appalachian highlands across the Upper South and into the near Southwest….

But are whites’ feelings of “American” ancestry why they support Trump, or is it just that Trump does well generally among downscale and rural voters? To find out, we ran a model that tests them together, using election results and county-level census data on American ethnic identity, socioeconomic status, race, religion, immigrant status and age.

Here’s what we found: When we control for those other factors, places with more unhyphenated Americans do indeed vote more strongly for Trump. For every 10 percent increase in a county’s share of unhyphenated voters, we found about 3 percent more support for Trump.

In other words, this region and these voters break for Trump above and beyond what would be predicted by other, especially economic, variables.

Outside of Appalachia, whites with Catholic ancestry also support Trump disproportionately. Discussing the New York primary, Nate Cohn says:

His best district might be New York’s 11th, which includes Staten Island and which has many voters of Italian ancestry. The model estimates he’ll receive 58 percent of the vote there.

It might seem strange to focus on something like Italian ancestry, but the number of voters with ancestry from predominantly Roman Catholic countries in Europe (like Italy, Ireland or Spain) is one of the strongest drivers of Mr. Trump’s support. Conversely, a high share of voters with ancestry from Protestant northwest Europe (like England, Germany or the Netherlands) has predicted lack of support. Unsurprisingly, Mr. Trump’s support in the area includes two prominent Catholic politicians, Rudy Giuliani and Chris Christie.

Of note, both of these groups historically favored Democrats until the realignment that began in the 1960s, while WASPs, who have supported Republicans longer, aren’t voting for the Republican insurgent.

Cohn also discusses this split further west:

Iowa, Utah, Kansas — and Wisconsin — have something else in common: a large population who report their ancestry from predominantly Protestant countries in Northern Europe.

These voters represented the base of the Republican Party for the century after the Civil War, whether it’s the old-stock “Yankees” who spread west from New England, or the German, Scandinavian and Dutch immigrants who generally settled over the same stretch of the northern part of the United States later in the 19th century.

These voters are probably the biggest problem for Mr. Trump that you haven’t heard of: He would fare about 30 points worse in counties where all of the white residents reported their ancestry from Protestant countries in Northern Europe than he would in a place where none did, according to our model. It’s the type of thing that helps separate Northern Virginia — where Mr. Trump struggled greatly — from the Boston area, where he excelled.

He further speculates that cultural “niceness”, deeper ties to the Republican party, and less racism might be the cause. Conservative pundit Michael Barone has wondered if the cause is higher social connectedness:

My first clues came from the Dutch. Heavily Dutch-American counties in northwest and central Iowa and western Michigan, around Grand Rapids, were Huckabee and Santorum territory in past years.

This year, unlike surrounding territory, they voted for Ted Cruz, with Trump a poor third. Dutch-Americans have dense networks of churches and civic groups — unusually high social connectedness….

[Robert] Putnam reports that social connectedness is highest in states with large Scandinavian- and German-American populations and in Utah. It’s lowest in — no surprise — Nevada, one of Trump’s best states.

In the 13 states highest in social connectedness, Trump has gotten just 21 to 35 percent in primaries and caucuses. In the 11 states lowest in social connectedness (except for Cruz’s Texas), his percentages ranged from 33 to 47 percent.

As we’ve seen and will see more fully, much of Trump’s base is in those parts of white America that aren’t doing well in a variety of ways. So if ethnic background is correlated with, or if there are cultural causes underlying these social problems then we should expect to see a correlation with support for Trump. It’s an issue that should be investigated more fully.


Death has been in the news lately. Last year Princeton economists Anne Case and Angus Deaton released findings that, unlike for any other group in America or in any other developed country, “The mortality rate for white men and women ages 45-54 with less than a college education increased markedly between 1999 and 2013, most likely because of problems with legal and illegal drugs, alcohol and suicide”. There has since been some debate whether the mortality rate is actually increasing or just stagnant, but everyone agrees that something is seriously wrong.

The death rate of whites is another predictor of Trump’s support. Even after controlling for factors related to income, employment, education, and urbanization, “the middle-aged white death rate in a county was still a significant predictor of the share of votes that went to Trump.”  (Note that this is measuring higher mortality in general, not the recent possible increase)


Yet again, a negative social indicator predicts Trump support.


There is a great deal of regional variation in support for Trump, but some of his highest support comes from Greater Appalachia and the Northeast. Relatedly, white Americans with Catholic ancestry and self-reported “American” ancestry vote for Trump. The areas that support Trump seem to suffer from various social ills, from higher mortality to low education to declining economies. Race and ethnicity also play a role. Living in areas with a larger black population also seems to be correlated with support.

Next: Trump, religion, and economic disaffection

[Overview] Trump Voters III – Religion and Economics

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa


The Republican Party has, in recent decades, established itself as a party for the religious. It has embraced, at least in words, a socially conservative agenda. 85% of Republicans and Republican-leaners describe themselves as Christian, 22% higher than Democrats/leaners, and 45% as evangelical/born again. So why is thrice-married casino magnate Trump, whose religious instincts do not appear to extend beyond pandering as needed, doing so well? Why, when many religious leaders have come out against him? When, in January only 5% of Republican Protestant pastors supported him?

As of March 29, Trump had won “an average of 36 percent of the vote from white “born-again or evangelical Christians,” good for a plurality in 12 states and only slightly lower than his support (38 percent) among all other Republican voters”, which would seem to indicate that religion might not be a huge factor one way or the other. However, not all self-professed evangelicals are the same:

Recently released data from the 2016 American National Election Studies (ANES) Pilot Study illustrate this. The study was conducted from Jan. 22-28, and here I focus on white respondents who called themselves born-again Christian. I divided evangelicals into people who “seldom or never” attend church services, those who “sometimes” attend (a few times a year, once or twice a month), and those who attend weekly or more often than weekly.


There is a clear negative relationship between support for Trump and church attendance. (But also note Trump’s plurality even among the most devout!) And, in fact, non-churchgoing evangelicals are in general more likely to support a Trump-like agenda. They care little for culture war issues, but greatly about economic issues. They hold less favorable attitudes towards Muslims, Hispanics, and blacks than other evangelicals.

There has been less research on Catholic Republicans, but they appear to support Trump at a higher rate than evangelicals (this is of course related to but distinct from the Catholic ancestry previously discussed):

Reporting on Monmouth University surveys has placed Catholic support for Trump at 30 percent in New Hampshire (where he took 27 percent of the evangelical vote), for example, and 44 percent in Iowa (where only 22 percent of evangelicals voted for him)…. In heavily evangelical South Carolina, where a third of evangelicals voted for Trump, Monmouth had 42 percent of Catholics doing the same….  Exit polls from much more Catholic Massachusetts placed Trump’s support from Catholics at an incredible 53 percent, four points higher than his support among evangelicals there…. [T]he Barna Group finds Trump’s net favorability among evangelicals at -38. Not only was it twice as high among Catholics, at only -19, but Catholics viewed Trump more favorably than did any other religious category Barna denominated.

Of note, Catholics attend church much less regularly than evangelicals. There doesn’t appear to be research on whether Catholic churchgoing is similarly related to support for Trump, but it seems plausible. (Also, Catholics are a heterogenous group that should be studied at a higher magnification than “Catholic”.)

Interestingly, the clearest relationship between religion and Trump is the negative one with the size of the Mormon population:


Utah, Idaho, and Wyoming, the three most Mormon states, have hosted some of Trump’s worst performances. Various theories have been posited as to why, including Mormonism’s complicated history as a religious minority. However, there isn’t a divide in Utah between Mormons and non-Mormons, but there is one between regular churchgoers and everyone else. And Mormons are the most religiously involved group in America. Perhaps, then, Mormonism matters not so much in and of itself, but in that it does a better job of getting bodies in pews. (Of course, even non-churchgoing Utahns are still not huge Trump fans relative to their counterparts in the rest of the country; there are obviously other factors. Perhaps a culture that leads to high religiosity influences even the irreligious, or maybe people just vote like their neighbors.)


Now, all of this suggests that a typical Trump voter doesn’t have strong social ties to religion, but it’s not enough to disentangle the causation. Does churchgoing change political attitudes and if so, how? Or is there some third factor causing both? Communities that are religious but not churchgoing also tend to have greater social ills. As Ross Douthat says:

HERE is a seeming paradox of American life. One the one hand, there is a broad social-science correlation between religious faith and various social goods — health and happiness, upward mobility, social trust, charitable work and civic participation.

Yet at the same time, some of the most religious areas of the country — the Bible Belt, the deepest South — struggle mightily with poverty, poor health, political corruption and social disarray.

The gains associated with religion come entirely from (or are only correlated with) religious practice rather than having a professed faith. The non-practicing-yet-religious, the “Christian Penumbra”, do more poorly on such measures than both the practicing religious and the irreligious. This is the group most likely to support Trump.

For comparison, Mormons, in addition to higher religiosity, are more socially conservative, less likely to be divorced, and have higher levels of education and income than the US average. They are more pro-immigrant than the rest of the GOP. Salt Lake City has high upward mobility. Mormonism produces – or at least happens to coincide with – social conditions that are hostile to Trumpism.

The social pathologies observed among communities that vote Trump – lower education, lower incomes, higher mortality – suggest that both likelihood to vote Trump and not go to church stem from an underlying cause. In other words, while religious factors may play a role, what’s being detected is likely not just religion, but something larger. (Though religiosity could also be a factor in its own right.)

Income and education

Here, I think that it’s useful to begin by looking at who votes for Republicans out of the nation’s population as a whole. While income may help predict who votes GOP, that doesn’t mean it tells us much about a candidate’s support within the GOP. From the 2012 election (for reference, household income in 2012 was $51,371):

2012 vote by income-01(Source)

As you can see, the average voter made more money than the average American; income and voting are positively correlated. (And since turnout is lower in primaries, primary voters may have even higher incomes on average.) Voters with above average incomes made up a majority of both Obama’s and Romney’s support. But while Romney barely won those making more than $50,000 a year, Obama had much stronger support among low-income voters, winning 63% of those making less than $30,000 a year. Presumably much of this was due to his huge margins among black and Hispanic voters.


I suspect that the sample sizes here are quite small, so take the above chart with a grain of salt. Probably some of the income gap among Hispanics is actually a voting gap between different Hispanic groups – Cubans-Americans, Mexican-Americans… – with different voting patterns. The minimal effect of income on white voters, who Romney won with 59%, seems surprising, so let’s break that down:


I’m not sure I see any real pattern among Southern whites (and some of the sample sizes are small), but the lopsided U-shaped support for Democrats outside of the South fits their usual pattern. Republicans get more support from wealthy whites, but I don’t think it’s fair to straightforwardly call either party the party of the rich or of the poor, especially when you account for other demographic factors. Still, the Republicans are the party of the more-rich, overall. In any case, this should contextualize what I’ve said about Trump voters being lower income – it’s only true relative to support for other Republicans, not to Democratic voters, and even less to the population as a whole.

Everyone other than non-Hispanic whites votes for Democrats in large numbers. Among whites, Southerners support Republicans in comparable numbers. Among white non-Southerners Republicans win the middle and upper-middle classes. (I don’t know that there’s enough data to say that income matters less among these other groups, it may just be that because white non-Southerners are clustered around the 50% mark the swings between income groups seem more important.)

As far as education goes, exit polls had Obama and Romney nearly tied, except for Obama winning 55% of those with postgraduate degrees. In general, Democrats also have U-shaped support when sorted by education. That is, they pick up a disproportionate share of the least- and most-educated, while Republicans do best among those with some college education but no degree. However, the gap is only about 10% at its largest.

So that’s the world into which Trump has entered. While, as we’ve seen, support for Trump is centered in areas that have been struggling in a variety of ways, support from merely the poor would not likely have won him the nomination. And according to 538:

The median household income of a Trump voter so far in the primaries is about $72,000, based on estimates derived from exit polls and Census Bureau data. That’s lower than the $91,000 median for Kasich voters. But it’s well above the national median household income of about $56,000. It’s also higher than the median income for Hillary Clinton and Bernie Sanders supporters, which is around $61,000 for both.

Cruz’s supporters have a median income of $73,000, essentially the same as Trump’s. But whatever the median, as the Economist shows, he has support from across the economic and educational spectra:

On average, people earning under $50,000 have made up 29% of the Republican electorate in primary states with exit polls, and 32% of Mr Trump’s voting base. However, those earning over $100,000 have accounted for 37% of the electorate and 34% of his base. In Illinois, for example, he took 46% of those earning under $50,000, but they made up only a quarter of the electorate: he won 39% of those earning over $100,000, who were two-fifths of that primary’s voters. Voters with a high-school education or less have made up 16% of the Republican electorate and a fifth of Mr Trump’s base. College graduates and postgraduates account for 43% of his support.

Trump is winning a greater share of low-income voters, but there aren’t that many of them. Lower income voters are a roughly equal share of primary voters as in 2012, so their increased turnout isn’t helping Trump.


Also, the fact that that Trump voters are older means that these numbers may overstate their socioeconomic position. A low-income recent college grad supporting Sanders may be likely to have higher lifetime earnings than an above-average income Trump voter. If Trump voters have larger families, they may feel more financial strain than these numbers indicate. Further analysis would be needed to disentangle these effects.

In March, Politico noted “In six of the statewide GOP exit polls so far, Trump was the most popular candidate among college-educated voters. In another six, he was their second-place choice. (Only in Oklahoma did Trump fall out of the top two among those with college degrees.)” So while his support is concentrated among less-wealthy, less-educated Republicans, it is by no means limited to them. He often still wins the groups he’s doing worst with.

Given that Trump kept winning handily, his broad support shouldn’t be too surprising:

Regional economic predictors seem stronger than these individual measures, which tentatively supports my hypothesis that a typical Trump voter is someone who lives in a struggling area, without necessarily being among the hardest hit themselves. Ross Douthat says:

One useful way to think about Trump’s constituency is suggested by this Alec MacGillis piece from last fall, which investigated the swing of less-well-off red-state voters, particularly in greater Scots-Irish America, into the Republican Party in the age of Obama. (A swing that pre-dated Trump, to be clear.) As MacGillis pointed out, the voters swinging rightward are not the very poor or the chronically jobless, not the people most likely to benefit from, say, Obamacare’s Medicaid expansion. Rather, they are a mix of relatively prosperous blue collar workers and what in a European context we would call the petit bourgeoisie — a coalition that straddles the working class and the lower middle class:

“In eastern Kentucky and other former Democratic bastions that have swung Republican in the past several decades, the people who most rely on the safety-net programs secured by Democrats are, by and large, not voting against their own interests by electing Republicans. Rather, they are not voting, period. They have, as voting data, surveys and my own reporting suggest, become profoundly disconnected from the political process.

The people in these communities who are voting Republican in larger proportions are those who are a notch or two up the economic ladder — the sheriff’s deputy, the teacher, the highway worker, the motel clerk, the gas station owner and the coal miner. And their growing allegiance to the Republicans is, in part, a reaction against what they perceive, among those below them on the economic ladder, as a growing dependency on the safety net, the most visible manifestation of downward mobility in their declining towns.

There’s a lot of evidence to suggest that these are the ur-Trump voters. Which means that — to borrow a vivid characterization from an earlier Dougherty piece (which helped inspire Kevin Williamson’s famous/infamous critique of white working class dysfunction) — they aren’t actually “Mike from Garbutt,” an opioid-addicted upstate New Yorker taking disability and contemplating leaving the unemployment line for a job at a casino. Instead, they’re probably Mike’s sister or cousin or former co-worker, who look at his situation and see a cautionary tale, an anger-inducing story about how welfare dependency political-class indifference are hastening their own communities’ decline.

What we’ve been talking about is income as a single number attached to a given voter. That doesn’t capture how a voter’s financial situation has changed over time, or how their expectations for the future have changed. From the Washington Post, this graph shows “that Trump’s gains among Republicans were larger among those who reported that their personal finances had gotten worse, compared to those whose finances had gotten better. The gap is now roughly 20 percentage points.” This is a much better predictor than income:



This matches a key attitude of Trump voters: that things are getting worse: (“Make America Great Again!”) However, there is a lot of evidence “that people’s perceptions of the economy—and even of their own financial situation—have a lot more to do with their political leanings than with objective economic facts.” Republicans say that the economy is worse when a Democrat is president and vice versa. (Overall, Americans have a positive economic outlook at the moment, by the way.) So it isn’t clear whether the above graph shows a relationship between voter’s finances or voter’s perceptions of their finances (or some combination). Perhaps questions more focused on specifics – did your income decrease over the last 12 months? – would avoid this confusion.

(This isn’t just a case of voters being convinced by Trump’s message that things are terrible though. People who were dissatisfied “about their lives, jobs, income, and economic situation in January 2015” were dramatically more likely to back Trump. The anger predates his candidacy.)

I wonder if the fact that less-educated men have seen their incomes decline over the last few decades also accounts for some of Trump’s stronger support among men. As education levels have improved too over this period, the population of people without high school diplomas today is not directly comparable to the population in the past, so the decline is probably overstated. However, an important point that I haven’t really seen made is that even if average income growth is merely stagnant or growing very slowly, that means that there is very likely a larger pool of people for whom things are in fact getting worse. Thus, if declining incomes lead to social dysfunction, it’s entirely possible that weak growth could have a larger negative impact than might be assumed from just noticing a 1% drop in growth.

To recapitulate, race predicts voting so strongly that 90+% of Republicans are white. Among whites, income plays a significant though not overwhelming factor in voting Republican. But among Republicans the effect of income on support for Trump is small relative to some of the other factors we’ve discussed. The same is true of education. And while he receives disproportionate votes from downscale Republicans, they are not the bulk of his support.


I couldn’t find good research on whether or not individual Trump voters were likely to be unemployed, or out of the labor force, but the New York Times looked at the areas that support Trump:

Likewise, a better predictor of strong Trump support than a standard-issue economic indicator like the unemployment rate is a high proportion of working-age adults who aren’t working (the correlation was strong for both men and women).

To be counted as unemployed, a person must have actively looked for work in the last month. But “not working” is a broader definition that would also include, for example, people who are discouraged by what seem like grim job prospects; who are living at home tending to the house; or who are disabled and stay home while receiving government assistance.

Nationally, 23 percent of the 25-to-54-year-old population was not working in March, up from 18 percent in 2000. The areas where Trump is most popular appear to be at the forefront of that trend.


So far we’ve covered the basic demographics of Trump voters. Next we’ll move on to their beliefs and worldviews, but first let’s review what’s been covered.

While Trump is very unpopular among non-white voters, this hasn’t harmed him much in the primary, as the Republican party is already 90% white. Among whites he does best among Appalachians and those with ancestry from Catholic countries. He does 5 to 10% worse among women. He does better among the middle aged and elderly and among the less educated and voters with lower incomes, but none of these effects are all that large individually. He hasn’t brought in many new voters to the GOP, though he has increased turnout among less attached Republicans.

Region matters a great deal. Living in areas that are socially dysfunctional matters a great deal. High mortality, “old economy” jobs, less education, low labor force participation, weak attachment to religion matter. Trump’s support in his worst counties in Utah and his best counties elsewhere differs by over 50%. Michael Brendan Dougherty summarizes:

Trump drew in the secular and moderate Northeast Republicans who had been sidelined by the conservatives. He combined those with the populist, nationalist South, particularly unchurched evangelicals. He threw in the downwardly mobile parts of Appalachia as well. The Trump coalition is a mix of Chris Christie and Jeff Sessions, with a dash of coal country thrown in.

So the prototypical Trump voter is a non-churchgoing but Christian older white male from greater Appalachia or the Northeast with some college education, maybe not doing poorly himself, but living in an area that is. (Though most Trump voters won’t fit this profile!) That Trump voters are often surrounded by social decay is the most uncertain part. More research should be done to distinguish the effects of living near poverty and living near non-whites who are poor. But we still haven’t connected the Trump voter to Trump. Why Trump? Why his populism and not Sanders’s socialism? Bill Clinton won most of Appalachia twice, so why not Hillary? Why now? It can be too easy to just say that of course this sort of voter would support Trump, to take it for granted. But why?

Next: Trumpism and Ethnocentrism

(Source for header image)

[Overview] Trump Voters IV – Ethnocentrism

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa

Racial attitudes

Okay, we’ve established who Trump voters are and the kind of environment they live in. How do we get from that to Trumpism? Economic and social malaise don’t self-evidently lead to Trumpism, even if one can sense an intuitive connection. What do these voters believe that makes them distinct? There is plenty of evidence that, above and beyond any economic factors, race and voters’ attitudes towards race matter. We’ve already seen that Trump does even worse among non-white voters than the GOP in general and he has definitely captured the support of outspoken racist whites. But what about Trump voters beyond the fringes? (The alt-right online is loud, not large.) What do they think about race and how has it influenced their votes? The racial divide must be accounted for. From one sympathetic account (which I recommend reading):

Imagine that you were born in a modest Prairie style home in small town Kentucky. Beyond the shadows of the maple trees that surrounded your house, farmland stretched down the highway, interrupted only by small Country homes with white fences, bright gardens, and sunny porches. After high school, you stayed in your hometown and had a family. But jobs were difficult to find. In your father’s day, there were plenty of good manufacturing jobs. But those had left. And the town was changing. The houses that you remembered as warm and sunny were now faded with rotting fences and unkempt lawns. Buildings that used to house hundreds of workers were empty and dilapidated with shattered windows….

Things were getting worse. You knew several people who had lost their jobs to Mexican immigrants who had just arrived in your town. You heard many more complaining about this on the radio, lamenting that immigrants had replaced them at the factory….

So, that was it—your town would continue to decay while politicians continued to pursue policies that hurt you and everyone you knew. And worse still, you weren’t even allowed to say anything about it—weren’t allowed to voice your real political opinions—without being denounced. If you were white, what did you have to complain about? You were privileged. You weren’t a victim. It was time you had gotten with the program. Embraced diversity. And stopped complaining about lower wages, fewer jobs, and changing cultural norms.

Enter Trump.

From a less sympathetic account:

For decades, Republicans relied on the Southern Strategy to win the White House. By using coded racial appeals like complaints about “welfare queens,” “crack babies” and school busing programs, they pulled whites disenchanted with the Democratic embrace of civil rights. The media inadvertently aided the Republican Party by inflating and racializing violent crime and welfare. Political scientist Martin Gilens finds that, “network TV news and weekly news magazines portray the poor as substantially more black than is really the case.” The strategy worked. Economists Ilyana Kuziemko and Ebonya Washington find that racist Southern whites leaving the Democratic Party explains nearly all of the decline in Southern White support for Democrats between 1958 and 2000. Instead of calling out this racism, many centrist Democrats succumbed to it, for instance when The New Republic famously defended welfare reform by placing a Black mother with a baby on the cover. The unwillingness of most mainstream liberals to call out dog-whistle racism has let it fester….

[T]he forces Trump released are not a joke; and the rise of white nationalism and violence against people of color confirm this. The rise of Trump isn’t just an indictment of the GOP, it’s an indictment of the unwillingness of mainstream commentators and politicians on both sides of the aisle to clearly call out racism. The problem is that now, it might be too late.

Enter Trump.

So we have two competing narratives. In the first, white residents of increasingly beleaguered areas come to feel that the wider culture not only doesn’t care about their plight, but actively despises them while valorizing non-whites and immigrants. (One study from 2014 found that whites who believed Obama had done too much for non-whites and that whites had done worse during the recession also felt “feelings of financial frustration and higher levels of blame toward the government in Washington.”) In this account economic and racial factors are intertwined, but racism per se is absent or at least minimized. Animus is not the driving force. In the second racism is “The vile core of Trump’s appeal”, plain and simple. Which, if either, of these obviously politicized narratives is best support by evidence?

Much of the research I’m citing is drawn from one study in particular, “The newest American National Election Studies 2016 pilot survey[,] a 1,200 person internet survey performed by YouGov between January 22 and 28 of 2016 that includes incredibly detailed questions about race and racism.”

It’s possible that some of the measured differences between Trump supporters are really just “differences in how Trump supporters feel they ought to answer surveys” rather than actual differences in hostility. For example,”it might be that Trump supporters are simply more willing to express their dislike of Muslims and Transgender people in a survey”, while others just hide their views better. But this is the data we have.

To avoid getting tied down in discussions of what is and is not racism, I’m not going to use the term and instead just look at responses to specific questions. Readers can draw their own conclusions. Instead, I’ll use the term “ethnocentrism” in a broad sense, covering not just open racism, but also subtler negative attitudes, and a sense of ethnic solidarity.

Some people have argued that “Any political movement that was straightforwardly based on economic distress would find its greatest support among nonwhite Americans.” Thus, Trumpism can’t be, because blacks and Latinos are economically worse off than whites and yet Trump is historically unpopular with both groups. And the recession hit black Americans very hard. However, a political movement based on social decline could find greater support among white Americans, even though white living standards are higher than those of blacks and Hispanics. Remember that the rise in mortality was limited to non-Hispanic whites. According to The Atlantic:

Likewise, the groups that have been affected most viciously by these market trends in the U.S., African Americans and Latinos, have not suffered the dramatic increases in death by suicide or substance abuse that whites have. It may be that changes in the economy have affected these workers in different ways. For instance, whites are more likely to be employed in the declining manufacturing sector than African Americans or Hispanics—and for that matter, they’re more likely to live in the rural communities devastated by this most recent, post-NAFTA era of deindustrialization. Furthermore, whites are less likely to be union members than African Americans (though not Asians or Hispanics).

“Make America Great Again” may not be a slogan that appeals to people for whom, recession aside, things aren’t always getting worse.

Certainly social decline isn’t the only reason for the racial gap. Trump’s racially charged rhetoric on top of decades of estrangement and hostility between minorities and Republicans would have divided voters by race anyway. But it does mean that Trump voters might not just be motivated by race. Stagnation and ethnocentrism aren’t mutually exclusive possibilities, though since one makes Trump voters look bad and the other makes them seem sympathetic they tend to be framed in opposition to each other.

One study found that both attitudes toward race and toward the economy predicted support for Trump. In addition to asking about financial security, they also asked “whether they thought it was more of a problem that African Americans and Latinos are ‘losing out because of preferences for whites’ or whether whites are ‘losing out because of preferences for blacks and Hispanics.'”

The odds that a person who feels strongly that whites are losing out supports Trump are more than three times higher than for a demographically and financially similar person who feels blacks or Hispanics are losing out or that neither group is losing out more. Likewise, the odds a person who says he or she is struggling financially supports Trump are about twice as high as someone who says he or she is comfortable or moving up economically….

Those who voiced concerns about white status appeared to be even more likely to support Trump than those who said they were struggling economically, but the results did not clearly show which concern was more important among Trump’s coalition.

chart2 (Source)

One of the pollsters also said, “What was striking to me in analyzing the data is that even after controlling for a variety of demographics and attitudes… believing whites are losing out continued to be a key predictor of Trump support…. Its importance persisted under a wide range of scenarios.” So even controlling for financial status racial attitudes matter. Clearly Trump is not wholly a phenomenon of economic anxiety. Race matters.

(There’s polling that finds support from Trump voters for more incendiary racial beliefs than I’m discussing here, but I don’t trust Public Policy Polling, a partisan firm, enough to cite them. I think the main points stand with the data I do discuss. Check out Snopes for more details.)

White Americans

What sort of racial anxiety might we be talking about? It isn’t just about simple hostility towards non-whites. Molly Ball, discussing Trump’s complaint at a rally “Why are they allowed to do things that we’re not allowed to do?”, says:

It was a potent summary of the identity politics that seem to form a significant part of Trump’s appeal: the idea that they, the others, enjoy privileges, resources, and status to which we are denied access. It is a sentiment I have repeatedly heard from the dozens of Trump supporters I have met over the past eight months I have spent covering his campaign. More complicated than the overt bigotry of, say, the Ku Klux Klan, it is a form of racial resentment based on historic white entitlement and a backlash to the upsurge in leftist identity politics that has marked American politics in the age of Obama…

Trump’s supporters have told me that minorities commit crimes with impunity, that illegal immigrants get benefits at higher rates than Americans, that gays and Muslims are afforded special status by the government. They lament that Confederate symbols, and the people whose heritage they represent, are sidelined while diversity is celebrated. They don’t understand why Democrats can campaign on overt appeals to the interests of blacks and women and Latinos, but Republicans are deemed offensive if they offer to represent the interests of whites and men.

There is support in the polls for this. Voters “who think their identity as whites is extremely important…. perceive a great deal of discrimination against their race…. [and] who think it’s extremely likely that ‘many whites are unable to find a job because employers are hiring minorities instead'” are all far more likely to vote for Trump:

tesler_trumpwhites (Source)

The numbers at the bottom show the sample size of each group. By themselves, these numbers don’t necessarily give us a great sense of how much support Trump is getting from these more race-focused voters. They could be a smaller portion of his support. However, other polls suggest that these voters make up a large portion of Republicans and of Trump supporters particularly. According to a Quinnipiac poll:

There is a wide partisan division among American voters on the statement, “The government has gone too far in assisting minority groups.” Agreement is 45 percent among all voters, 72 percent among all Republicans and 18 percent among Democrats. Agreement is highest among Trump backers, 80 percent.

Black Americans

Salon offers further reporting on the findings of the ANES survey. Salon hardly has a sterling reputation these days, and the article is definitely pushing an agenda (this is the unfavorable piece I quoted above), but I don’t have any particular reason to think that the data they’re drawing on is incorrect. They found that Republicans had a more negative view of blacks than Democrats, and Trump voters were more negative still:

CohenMc1 (Source)


The ANES survey also included a feeling thermometer test, which asks respondents to place their feelings for different groups on a scale from 0 (very cold) to 100 (very warm)…. [T]he relationship is strongest for Trump support, which shows that  negative feeling toward Blacks is most closely associated with his supporters. The model controls for age, gender and education.

As Trump has expanded his support in the months since this poll was taken and become the presumptive nominee, I would expect that Trump voters would become more similar to Republicans as a whole, since many Republicans will now be supporting Trump. But we can see here the base Trump started with.

The authors discuss further results from the survey here, and they also criticize this Vox article which found that racial resentment didn’t much vary among supporters of different Republican candidates.

One of the authors also Tweeted this from the survey:

% of whites who support Trump who think government favors Blacks: 51%
% of whites who don’t support Trump who agree: 32% (ANES 2016 Pilot)


In January, the Washington Post found:

Trump barely leads among Republicans who say immigrants strengthen America. But he holds a 33-point lead — commanding a solid majority — among Republicans who strongly believe immigrants weaken society.

Pew found that Trump supporters consistently take the harshest stances against immigrants, and not just illegal immigrants:


They are also likeliest to support deporting illegal immigrants, though only a plurality of 42% do. But note the near-even split on whether undocumented immigrants should be given legal status conditionally.

The Washington Post looked at this same study:

59 percent of registered voters nationwide think that an increasing number of people from different races, ethnic groups and nationalities makes the United States a better place to live; only 8 percent say this makes America worse. But among Trump backers, 39 percent say diversity improves America, while 42 percent say it makes no difference and 17 percent say it actually makes America worse. Supporters of GOP rivals Ted Cruz and John Kasich were significantly more upbeat on diversity….

Other somewhat-related attributes may be as or more predictive of whether somebody will support Trump: approval of deporting undocumented immigrants, strong feelings that the government is dysfunctional, and support for banning Muslims from entering the United States. (Authoritarian child-rearing attitudes, believed by some to be closely related to Trump support, were less predictive.)

Trump voters are the most hostile to diversity, but still over twice as many Trump voters think diversity improves America as weakens it. Unfortunately, these studies doesn’t really show how much hostility to immigration is driven by economic vs. cultural concerns. Opposition to immigration doesn’t necessarily stem from ethnocentrism or racism, but it is of a piece with Trump supporters’ attitudes towards other groups.


Using a survey from November, Michael Tesler found that concerns about Muslims and Islamic terrorism are also related to support for Trump:

tesler_trumpmuslim3 (Source)

In December, after the shooting in San Bernardino, Trump proposed a total ban on Muslims entering the United States “until our country’s representatives can figure out what is going on”. In March a poll found that 50% of Americans agreed:

While 71 percent of Republican voters supported the ban, 34 percent of likely Democratic voters and 49 percent of independents also did, according to the new poll by Morning Consult.

Although support for the ban is highest among Trump’s supporters, at 84 percent, there is also support for a ban among those who support fellow GOP candidates Ted Cruz and John Kasich — at 65 percent and 48 percent, respectively.

Among Democrats, more Hillary Clinton supporters agreed with the travel ban, at 37 percent, versus 27 percent of Bernie Sanders‘s supporters.

Pew found that a smaller number of Americans support profiling Muslims in America, but again that Trump supporters were most likely to be in favor:

Section2_6 (Source)

(And according to left-leaning polling firm Public Policy Polling, 62% of Trump supporters say that Obama is Muslim, versus 54% of Republicans overall.)

Yet again Trump voters take the hardest line, though again they aren’t complete outliers either.


I’ve gone into unusual detail in this section, because race and related topics are obviously particularly sensitive, and I wanted to be sure that any conclusions reached were backed by a wide variety of data.

Obviously Trump draws support from voters that have consistently less friendly attitudes towards various outgroups and are defensive about the prospects of their white ingroup. Probably all of the polls I cited here are more or less picking up on a single factor: ethnocentrism. Combining various questions from the ANES survey into one measure of ethnocentrism, Kerem Ozan Kalkan made this graph:

kalkan_fig2 (2)

I’m not going to speculate if or where these attitudes shade into racism. Certainly by the time you hit David Duke. But if we can separate racist actions from the attitudes that lead to them, I think that ethnocentrism is the right way to understand the underlying attitude that is motivating Trump supporters. This ethnocentrism includes more negative attitudes towards blacks, immigrants, Muslims, but it isn’t just that. There are other associated traits, such as having a strong group identity and hostility to trade and foreign aid (which I’ll discuss later). Again, my purpose here is neither to condemn nor condone, just understand more precisely.

I’d encourage readers to look at the charts above again. Trump is winning the more ethnocentric wing of the Republican Party (which itself itself has a much stronger tendency towards white ethnocentrism than the Democrats), but on a given question many voters who don’t give the ethnocentric answer support Trump and many voters who do don’t support Trump. While I think that this factor is one of the most important ones when understanding Trumpism, it isn’t the only one, or even the only major one. To say that racism is the driving force is too simplistic, but so is to pretend that race has no importance, or that race as a factor can be wholly reduced to economics. Whether economic malaise can drive people to greater ethnocentrism is beyond the scope of these posts.

These polls can’t capture everything about race. If a given Trump supporter doesn’t express racist attitudes, that doesn’t mean there aren’t deeper social forces helping to sort things by race. Race can influence how we vote even if we’re not saying so to pollsters, whether or not we’re not aware of it.

Next: The alienation and anger of Trump voters

(Source for header image)

[Overview] Trump Voters V – Isolation

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa

Disaffection and powerlessness

Americans are angry. According to NBC last year:

Overall, 49 percent of Americans said they find themselves feeling angrier now about current events than they were one year ago. Whites are the angriest, with 54 percent saying they have grown more outraged over the past year. That’s more than Latinos (43 percent) and African-Americans (33 percent).

Seventy-three percent of whites said they get angry at least once per day, compared with 66 percent of Hispanics and 56 percent of blacks.

The poll also found Republicans are angrier than Democrats. Sixty-one percent of Republicans say current events irk them more today than a year ago, compared to 42 percent of Democrats.

Republicans, perhaps expectedly, have it worse than Democrats. Here’s how one researcher working in Wisconsin described the mood:

Even before the Great Recession, people told me that they were working hard to make ends meet, and their taxes continued to climb. Where was their money going? They didn’t believe it was coming back to them.

Here’s how they saw it: Working class whites told me that although they could not afford health insurance for their own families, they were nevertheless expected to pay taxes so that public employees could have health care. To add insult to injury, they told me, those folks did not even seem to work very hard. They sit at desks, rather than working with their hands. They shower before work, not afterward. And public school teachers and university faculty got summers off!

Remarkably, Republicans are three and a half times more likely to say that life was better for people like them 50 years ago than worse. And Trump voters had the most dismal outlook of all. 75%!

overview_1 (Source)

Among Republicans, Trump voters are also most likely to say that “the economic system in the country unfairly favors powerful interests” (61%), and to say that “business corporations make too much profit” (43%). (Democrats are more likely still.)

Trump voters feel powerless. According to a Quinnipiac poll:

There is widespread agreement, 76 percent, with the statement, “Public officials don’t care much what people like me think.” Agreement is 84 percent among all Republicans and 68 percent among Democrats. Ninety percent of Trump supporters agree, the highest of any candidate.

There is a lower level of agreement, 64 percent, with the statement, “The old way of doing things no longer works and we need radical change.” Agreement is 71 percent among all Republicans and 58 percent among Democrats. Agreement is highest among Trump supporters with 83 percent.

The Atlantic reported something similar:

RAND tested several queries to clearly divide Trump’s support from his rivals. For example, they found that Trump crushes Ted Cruz among voters who both strongly believe that “immigrants threaten American customs and values” and among voters who “strongly favor” raising taxes on the richest American households. But voters who agreed with the statement “people like me don’t have any say about what the government does” were 86.5 percent more likely to prefer Trump. This feeling of powerlessness and voicelessness was a much better predictor of Trump support than age, race, college attainment, income, attitudes towards Muslims, illegal immigrants, or Hispanic identity.


Incidentally, out of everyone, Trump voters are most likely (40%) to have “a very great deal or good deal of confidence in the public’s political wisdom”, which has got to have something to do with one’s preferred candidate taking over an entire party.

Attitudes toward government

Let’s delve in to the sources of anger towards government specifically. Pew found that voters overall were most upset with politicians in general and the system as a whole rather than over specific issues:

Section1_13png (Source)

And Trump voters are more angry about government and politics than anyone else. Essentially no Trump voters are content:

Section1_14 (Source)

What is fueling this discontent? The results of this December focus group with Trump supporters (an in-depth, in person discussion with a few dozen voters) run by Republican pollster Frank Luntz are probably less reliable than those of larger polls, but they give a sense of the spirit of Trump voters. They describe America today as:

“slumbering,” “confused,” “volatile,” “incompetent,” “selfish,” “resilient,” “at a precipice,” “corrupt,” “stalled,” “angry,” “innovative,” “vulnerable,” “divided,” “full of opportunity,” “stagnant,” “becoming very vile” and “hijacked by criminals.”

For Obama, “A handful used profane words or harsh phrases that are unpublishable here. Out of 29 people, only three believe that Mr. Obama is a Christian and only four think he loves America.” The media is:

“kowtowing,” “own agenda,” “focused on the negative,” “biased,” “slanted,” “liberal,” “out of touch,” “partisan,” “stupid,” “left-wing”.

Trump voters think that the government is doing too much by an overwhelming margin, though a bit less than other Republicans:


But as we’ll see, this general anger at government mostly doesn’t extend to specific programs.


Wendy Rahn and Eric Oliver looked at whether Trump voters are populists (which we’ll return to). Some of the questions asked are relevant to Trump voters’ attitude towards expertise:

Anti-elitism. What separates populists from authoritarians is their alienation from political elites. We measure this with statements like “It doesn’t really matter who you vote for because the rich control both political parties,” “Politics usually boils down to a struggle between the people and the powerful” and “The system is stacked against people like me.”

Mistrust of experts. Populists often fear not just political elites and billionaires, but anyone who claims expertise. We measure this with questions like “I’d rather put my trust in the wisdom of ordinary people than the opinions of experts and intellectuals” or “Ordinary people are perfectly capable of deciding for themselves what’s true and what’s not.”

They found that among Republicans, Trump voters are by far the most anti-elitist and least likely to trust experts:


Trump voters are also more likely to believe in conspiracy theories. 40% believe that “President Obama is hiding important information about his background and early life,” compared to about 30% for Cruz and Kasich voters. “50 percent of Trump’s supporters said Clinton definitely knew about the attack in Benghazi beforehand. Just 43 percent of Cruz’s supporters and 40 percent of Kasich’s supporters agreed.” According to liberal pollsters PPP in September, “61 percent don’t believe President Obama was born in America” versus 44 percent of Republicans.

The gap between Trump supporters and other Republicans shrinks when considering less obviously political questions. 52% of Trump supporters said that vaccines possibly or definitely have been shown to cause autism, and 21% said that the Sandy Hook school shooting could have been or was faked, but for both of these questions supporters of other candidates gave similar answers. So Trump voters are more conspiracy-minded than other Republicans, but the gap is not especially large.

For comparison, “A majority of Democrats, for example, told Fairleigh Dickinson that it was at least possible that President George W. Bush knew about the terrorist attacks of Sept. 11, 2001 before they occurred.”

Interestingly, a slight majority of Trump voters believe that global warming is happening, more than among Cruz voters. However, of those voters 55% say that global warming is not man-made but natural.


It would also be interesting to look at how Trump voters follow the news. Unfortunately, the only thing I could find was this disreputable-seeming survey from October, commissioned by “video startup Wibbitz”, which found that Trump voters were the most likely to get their news from television.


A common complaint among pro-Trump (or at least pro-Trump voter) articles is that his supporters are looked down on as “mental midgets and xenophobic troglodytes who’ve crawled out from their survivalist caves in order to destroy the Beltway Establishment” and that this contempt from elites is helping fuel support for Trump. Unfortunately, I didn’t find research on this question in particular. “Do you feel culturally looked down upon?” would be interesting to ask, though I’m not sure that polling would be enough to establish that contempt is a cause. After all, it could be that Republicans/Trump voters support whom they do for some other reason and are looked down on by the left-leaning media ultimately for the fact that they have different political views. (Remember that people discriminate by party more than they used to.) That is, partisanship can cause derision (both ways, of course), so saying that Trump voters feel they’re held in contempt isn’t enough to show that contempt is a cause. It’s not implausible – the idea that partisan hatred magnifies existing divisions seems likely to me – but it has to be shown.

What voters care about

Exit polling gives us another way to look at what is motivating people to vote for whom. Pollsters asked voters which of four issues was most important to them:


(Data from here; graph created by me; not all states exit polled)

As you can see, voters who cared most about immigration were consistently the most likely to vote for Trump. He more often than not won pluralities among voters caring about the other three issues, but clearly immigration is his strongest issue. However, immigration isn’t actually the most important issue among Republicans. Only an average of 11% of Republicans named immigration, versus 35% for economy/jobs, 23% for terrorism, and 29% for government spending. Thus a plurality of Trump voters actually cared most about the economy, not immigration. In fact immigration provided Trump with the smallest amount of voters (though still important to his success):


(Data from here; graph created by me; not all states exit polled)


We’ve already established that Trump voters are disconnected from organized religion. Whites are less likely to be in unions than blacks, and union membership is in general decline. The less educated (who disproportionately vote Trump even if they don’t make up the bulk of his support) have weaker social ties. And here we see high levels of hostility towards government and the media, and a strong sense of powerlessness. This isolation, in multiple senses of the word, is, I think, another one of the foundations of Trumpism.

Next: How “conservative” are Trump supporters? What are their actual ideological motivations?

(Source for header image)

[Overview] Trump Voters VI – Conservatism and Policy

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa


Well before Trump, and well before Obama, Republicans had been getting more conservative – certainly rhetorically. “[S]elf-described conservatives [made] up a greater share of Republicans in 2012 (68%) than they did in 2000 (59%)”. Until roughly yesterday Tea Party intransigence was the order of the day, including a two-week shut down of the federal government. So why has Trump, who 45% of Republicans consider a moderate rather than a conservative, won?

At least in part, it’s because there exists a large bloc of “conservative” voters who aren’t in practice that conservative:

[A]fter analyzing survey data from the 1950s, Converse found that the public was largely “innocent of ideology.” When asked their likes and dislikes about the political parties and presidential candidates, relatively few used ideological concepts or terminology. The majority could not define terms like “liberal” and “conservative” or could define them in only in vague terms.

Moreover, people’s views on various political issues didn’t “go together” the way that liberalism or conservatism would predict. Knowing people’s view on one issue didn’t really help predict their view on another issue. This was particularly true among voters with less formal education — exactly the group that has been most attracted to Trump….

Similarly, recent work by James Stimson and Christopher Ellis (discussed here) as well as Sunshine Hillygus and Todd Shields has found that it is common for voters to have political views out of step with the mainstream of their party.

In particular, Ellis and Stimson show that it is self-identified conservatives who are particularly prone to this because so many of them take liberal positions on key questions like the size of government. They find that this group — “symbolically conservative” but “operationally” liberal — actually comprises a larger share of the electorate (nearly 25 percent) as of 2008 than it did in 1974.

In fact, when it comes to “whether the government should do more, less, or the same in lots of different policy areas”:

[A]lmost 30 percent of Americans are “consistent liberals” — people who call themselves liberals and have liberal politics.  Only 15 percent are “consistent conservatives” — people who call themselves conservative and have conservative politics.  Nearly 30 percent are people who identify as conservative but actually express liberal views.  The United States appears to be a center-right nation in name only.

These “conflicted conservatives” are of course conservative for other reasons, whether social conservatism or cultural affiliation with the right. But, given that the Republican party has been led by its business and libertarian wings toward a more economically conservative policy than its base would naturally support, there was an opening for someone who could operate outside the normal party apparatus (and thus avoid the need to prove ideological loyalty) to exploit.

There’s evidence for this. Plenty of polls found that Trump did well among moderates and exit polling bore that out. What’s striking is that he often did as well or better among self-described “somewhat conservative” voters. It was only with very conservative voters that he lagged, though not always by much:


(Data from here; graph created by me; not all states exit polled)

In most elections somewhat conservatives were the largest bloc (40+% of voters), and outside of the Northeast moderates were always the smallest bloc, so a large plurality of Trump supporters were somewhat conservative:


(Data from here; graph created by me; not all states exit polled)

Economic conservatism

So if neither Trump nor his supporters are consistently conservative, what policies do they support? Given that Trump has switched his positions on a whole bunch of issues, I’m not sure that polling is a very good guide to what policies he will actually pursue. Nevertheless, polling may tell us something more about what animates his base. Let’s look at polling from the 2012 election to start. “Majorities of GOP primary voters were willing to cut only four things: unemployment benefits, spending on housing, spending on the environment, and foreign aid.” And other than for foreign aid those majorities were hardly overwhelming:


Clearly even before Trump there was a large Republican constituency to preserve government programs. And according to the Washington Post, he has disproportionate appeal to those who want to increase federal spending (on a number of specific programs rather than in general):


The authors say:

There are two important takeaways. First, Trump does significantly better (15 points) among those who want spending increased rather than decreased.  Second, the numbers at the bottom show there are far more Republicans who want spending increased on these issues than decreased.

RAND’s long-term survey of select voters also found that voters who favored raising taxes on high incomes were far more supportive of Trump. He is tapping into a large group. “51 percent of Republican primary voters strongly or somewhat favor increasing taxes on individuals who make more than $200,000 a year”.


Although Trump has called Obamacare a disaster, he’s (vaguely) signaled support for some sort of universal coverage even last year. His supporters haven’t followed his lead, though. 84% don’t believe that the federal government has the responsibility to make sure everyone has health care, very similar to other Republicans. (The numbers for Democrats are almost exactly reversed.) On the other hand, 73% of Trump supporters don’t want Social Security benefits reduced, not too different from any other group of voters. So there’s no real divide between Trump voters and Republicans, just between Republicans and Democrats.

This broad support of current government programs and government spending needs to be reconciled with the fact that a mere 1% of Trump voters said they were “basically content” with the federal government. Let’s bring back this chart:

Section1_13png (Source)

The chart is for all voters, not just Trump supporters, but it makes sense that Trump supporters would be angry at politics and politicians, rather than government per se. Trump voters may not be angry about what government does (at least spending-wise) so much as angry at how it does it.

There is another possible reason for the disconnect between anger at “government” and support of government I’d like to propose: people have hugely distorted views of federal spending, so they might have favorable attitudes towards the programs that actually make up most of the government, but unfavorable attitudes towards a small number of programs that they think are eating up a huge amount of money, like foreign aid and welfare. Some amount of voters’ anger may simply be due to bad information. A 2011 CNN poll (in line with many other such surveys), found a great deal of ignorance as to how the government spends tax dollars:

According to the poll, on average, Americans estimate that foreign aid takes up 10 percent of the federal budget, and one in five think it represents about 30 percent of the money the government spends. But the actual figure is closer to one percent…. According to our poll the public estimates that the government spent five percent of its budget last year on public television and radio…. On average, Americans think the federal government spent 10 percent of its 2010 budget on pensions and retiree benefits; the OMB figures indicate the real number is about 3.5 percent.

A sizeable minority would like to see food and housing assistance for the poor on the chopping block, but Americans’ estimates of how much the government spends on those programs are three to four times higher than the actual price tag….

But the public, once again, overestimates the amount of military spending. They told us 30 percent in our poll. In reality only 19 percent of the 2010 budget went towards military spending, according to 2010 OMB figures.

(When people are told how much is actually spent on foreign aid, support for cuts drops by half.)

If voters understood government better, perhaps they would be less angered by it, though complaints about process and politicking would remain.

Notably, Trump voters are also more likely to feel positively about unions:


This fits with the general pattern of Trump voters not being doctrinaire economic conservatives.

Social conservatism

Trump voters are socially conservative relative to the country as a whole, but not relative to other Republicans. Ted Cruz was the candidate who captured the social conservative vote:


Salon looked yet again at the attitudes of Trump voters towards various groups, this time looking beyond race/ethnicity/religion. They asked “respondents to place various groups and political figures on a scale from 100 (very warm or favorable feeling) to 0 (very cold or unfavorable feeling).” (The graph’s title is unfair. A majority of Trump supporters report positive feelings toward blacks and Hispanics, for instance.) Relative to other Republicans, Trump voters had statistically significant colder feelings toward feminists, gays and lesbians, and transgender people:


Can we reconcile this hostility with Trump voters caring less about same sex marriage (assuming the discrepancy doesn’t just come from comparing to different sorts of surveys)? Perhaps this lack of warm feelings doesn’t translate into policy concerns. Social issues like this might just be less of a concern.

73% of Trump voters are strong supporters of gun rights, versus 66% of Cruz/Carson voters and 42% of Rubio/Kasich voters. I’m not sure what accounts for this, but people who think (incorrectly) that crime is rising are a bit more likely to oppose gun control, so perhaps part of the reason is that Trump voters are more likely to think that crime is rising, which would fit their sense that life is getting worse.

I couldn’t find any information on how Trump supporters in particular feel about “political correctness” (though they want a leader who speaks his mind), but I did find this amusing poll result:

In an October poll by Fairleigh Dickinson University, 68 percent agreed with the proposition that “a big problem this country has is being politically correct.”

It was a sentiment felt strongly across the political spectrum, by 62 percent of Democrats, 68 percent of independents and 81 percent of Republicans. Among whites, 72 percent said they felt that way, but so did 61 percent of nonwhites….

The PC backlash does not necessarily mean that people support the kinds of things that Trump is saying, or the way he says them.

When the Fairleigh Dickinson pollsters added his name to the same question — prefacing it with “Donald Trump said recently . . . ” — the numbers dropped sharply. Only 53 percent said they agree that political correctness is a major problem.

Foreign policy

Trump voters aren’t isolationists, but they are skeptical of engagement with the rest of the world. The Wall Street Journal found that 37% of Trump supporters said that the US should focus on problems at home rather than abroad, the highest among Republicans. Still, not a majority. The best data out there on foreign policy attitudes comes from this Pew study, which I’ll be drawing on. Again, Trump supporters take the hardest lines against the outside world.

54% say that the US does to much to solve the world’s problems, compared to 40% of other Republicans. Even so, a majority of Trump supporters (57%) also say that problems would be worse without American involvement, so some of the opposition must come from a sense of “Not our problem” than of feeling that involvement is counterproductive.

We’ve already seen that Trump supporters are more likely to support a complete halt to Muslims entering the US. Refugees are a major concern:

For those who back Trump in the primary campaign, the large number of refugees leaving Iraq and Syria is especially worrisome. More than eight-in-ten GOP voters who support Trump (85%) say the refugees are a major threat to the U.S., compared with 74% of those who prefer Cruz and 59% who prefer Kasich. Among Trump supporters, only the threat from ISIS (93% major threat) ranks higher than refugees, among the eight issues included. Among Democratic voters, just 40% of Clinton supporters and 34% of Sanders supporters view the refugee migration as a major threat.

Although Trump supporters say we’re doing too much to solve the world’s problems, that does not extend to further reducing America’s military footprint in the Middle East. In fact, they are as or more supportive of intervention than anyone else:


However, Trump supporters are by far the most skeptical of NATO membership. For supporters of every other candidate, Democrat and Republican, 80+% support NATO membership, with only 10+% opposed. Trump voters are the only ones who differ. 64% support membership, but 30% do not. This at least suggests a greater inclination towards a foreign policy based around unilateral use of force. If even NATO is a bit suspect, I can only imagine that the support for cooperation with other nations is even weaker.

Given that Trump supporters favor ground forces against ISIS two-to-one, they can’t rightly be called isolationists. This stance might appear to contradict their belief that the US does too much to fix the world’s problems, but as 93% of Trump supporters also see ISIS as a major threat, no doubt further military action in Mesopotamia is supported for the sake of America rather than the world. I think that there’s another reason though. The opposition to helping solve the world’s problems is mainly opposition to non-military aid. We’ve already discussed opposition to foreign aid, but Trump supporters are also the most opposed to increasing economic ties to the third world:


Somewhat inexplicably, there’s a lot of opposition from all corners to American businesses investing in developing countries. The exact wording, “Would you support or oppose increasing U.S. companies’ investment in developing countries”, seems innocuous. Maybe this is veiled opposition to offshoring? I don’t know. Lots of voters also think that American involvement in the global economy is harmful, but investing abroad would seem to increase national economic power.

Above you can see that Trump supporters are the most opposed to trade too, and free trade agreements in particular. Look at how many feel personally harmed by free trade agreements! That is a dramatic difference with everyone else.

Other than using force against ISIS (and presumably other terrorist groups), they seem to desire withdrawal from the world on all fronts. I think it’s fair to say that a typical Trump voter’s preferred foreign policy is one of economic autarky, minimal engagement with multilateral institutions, and major willingness to use force. This isn’t isolationism exactly, and it isn’t nation building. I would call it nationalism that believes we live in a zero-sum world.


Trump voters are conservative, but only by the standards of America as a whole, not by the standards of the Republican party. He draws much support from people who aren’t actually opposed to lots of government programs, or to abortion or same sex marriage. There’s no evidence that Trump supporters want cuts in spending or tax cuts for the wealthy. Social conservative issues don’t seem to be a motivator. There’s no one “conservative” foreign policy, but their opposition to trade and international engagement is at odds with Republican practice, though not their support of using military force.

Next: Trump supporters: authoritarians?

(Source for header image)

[Overview] Trump Voters VII – Authoritarianism

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa


Now that we’ve seen what Trump voters believe in, and seen that they aren’t much motivated by conservatism, I’ll close by looking at some overarching theories that attempt to tie everything together. Judging by Facebook shares, one of the more popular theories out there is that Trump voters are authoritarians in some sense, that they crave strongman leadership. The first major article on this was a January piece by Matthew MacWilliams in Politico, who found that having an authoritarian personality predicted Trump support:

My finding is the result of a national poll I conducted in the last five days of December under the auspices of the University of Massachusetts, Amherst, sampling 1,800 registered voters across the country and the political spectrum. Running a standard statistical analysis, I found that education, income, gender, age, ideology and religiosity had no significant bearing on a Republican voter’s preferred candidate. Only two of the variables I looked at were statistically significant: authoritarianism, followed by fear of terrorism, though the former was far more significant than the latter.

A second study from Vox found similar results:

Trump has 42 percent support among Republicans but, according to our survey, a full 52 percent support among very high authoritarians.

Authoritarianism was the best single predictor of support for Trump, although having a high school education also came close. And as Hetherington noted after reviewing our results, the relationship between authoritarianism and Trump support remained robust, even after controlling for education level and gender.

Trump support was much lower among Republicans who scored low on authoritarianism: only 38 percent.

Vox also found that non-authoritarians who nevertheless felt afraid of foreign threats were also more likely to vote for Trump. I’m not sure why the relationship reverses for authoritarians:


Striking findings, especially considering all of the ways that we’ve seen Trump supporters stand out from other Republicans. Are all of those other factors incidental? Is authoritarianism the only thing that really matters? First, here’s what researchers mean by authoritarianism and how they measure it. MacWilliams says:

Authoritarians obey. They rally to and follow strong leaders. And they respond aggressively to outsiders, especially when they feel threatened. From pledging to “make America great again” by building a wall on the border to promising to close mosques and ban Muslims from visiting the United States, Trump is playing directly to authoritarian inclinations.

This personality trait comes to the fore when people feel “threatened by social changes such as evolving social norms or increasing diversity, or any other change that they believe will profoundly alter the social order they want to protect.” Authoritarianism only becomes obvious when people feel frightened, otherwise it’s “latent”.

How is this measured? Obviously you can’t straightforwardly ask people “Are you an authoritarian?” and expect to get an honest answer. Political scientists have a standard set of indirect questions they ask to measure authoritarianism, based on attitudes towards child rearing:

  • Please tell me which one you think is more important for a child to have: independence or respect for elders?
  • Please tell me which one you think is more important for a child to have: obedience or self-reliance?
  • Please tell me which one you think is more important for a child to have: to be considerate or to be well-behaved?
  • Please tell me which one you think is more important for a child to have: curiosity or good manners?

Giving the second answer marks you as more authoritarian. The theory behind this being that people who value order and hierarchy in one domain will value it in others. These questions have been standard for over two decades now. lot of people are authoritarian according to this measure:

Our results found that 44 percent of white respondents nationwide scored as “high” or “very high” authoritarians, with 19 percent as “very high.” That’s actually not unusual, and lines up with previous national surveys that found that the authoritarian disposition is far from rare….

Today, according to our survey, authoritarians skew heavily Republican. More than 65 percent of people who scored highest on the authoritarianism questions were GOP voters. More than 55 percent of surveyed Republicans scored as “high” or “very high” authoritarians.

These people aren’t all salivating for a dictator, but (so the theory goes) they are likelier to desire authoritarian leadership when they feel threatened. (So it makes sense that non-authoritarians who are also worried about foreign threats are also more likely to support Trump.)

Now, I think it’s important here to distinguish between authoritarianism the supposed personality trait they’re trying to measure and authoritarianism the attitude towards child rearing that is hopefully correlated with the personality trait. So what else is this measure correlated with, beyond support for Trump? A lot of things, many of them similar to characteristics of Trump voters we’ve already discussed:


This all sounds, to be honest, suspicious. How much can a few questions about child rearing tell us? How much use is a test that finds 44% of people to be authoritarian? There are, I think, good reasons for skepticism. For instance, this test of authoritarianism may only work in whites:

Using a scale of child rearing preferences, scholars find that African Americans are far more authoritarian than Whites. We argue that this racial gap in authoritarianism is largely a measurement artifact. The child rearing scale now used to measure authoritarianism is cross-racially invalid because it draws heavily on a metaphor about hierarchy. Akin to someone who favors enforcing conformity in a child, the authoritarian is thought to be inclined toward enforcing conformity in social subordinates. In both cases, one’s perspective is drawn from a position of relative power. We believe this metaphor is effective among members of a majority racial group because individual dominance at home meshes with group dominance in society. For members of a racial minority, we believe this metaphor breaks down. Using multi-group confirmatory factor analysis, we establish that Blacks and Whites construe the child rearing items differently. Consequently, authoritarianism correlates highly with the things it should for Whites, but rarely so for Blacks.

I believe that in both polls only whites were queried, so this isn’t a direct mark against them. The problem is that the test’s validity is culturally contingent, and the test could be invalid for reasons other than race:

After all, the new index measures approval for some old-fashioned ideas about raising children. These ideas were once widespread, but have become more characteristic of the working class, especially in the South. To be “authoritarian”, in other other words, means little more than endorsing the folk wisdom of a class and place that many academics find alien.

You might also ask how “real” these values are. People often respond to surveys by giving what they regard as the appropriate answer, rather than the one that most accurately reflects their behavior. This “social desirability bias” helps explain why people report voting and going to church more often than they actually do them.

So people who express authoritarian attitudes may not be more authoritarian in practice than anyone else (they may even endorse authoritarian values precisely because their lives are disordered).

Might these tests be picking up cultural variation instead of psychological differences? Maybe at least in part. In fact, researchers writing at the Washington Post polled the same questions and found that, while Trump voters were more authoritarian than the average American, they weren’t even the most authoritarian-scoring group, more religious Cruz voters were:

In fact, they score slightly lower on these scales than Cruz’s voters. Why? Partly, this is because scales measuring child-rearing correlate very highly with fundamentalist Christian beliefs. By these measures, most Republicans look like “authoritarians” because so many are conservative Christians who advocate strict child-rearing practices. This is also why Bernie Sanders’s supporters are so much less authoritarian than Hillary Clinton’s — “Berners” are much less religious than other Democrats….

Granted, we don’t have a lot of other measures of authoritarianism, such as an attraction to strong leaders or intolerance of ambiguity. It may be that Trump’s supporters are more swayed by these traits than other Republicans.


According to Quinnipiac:

Agreement with the statement, “What we need is a leader who is willing to say or do anything to solve America’s problems” is 53 percent among all voters, 68 percent among all Republicans and 39 percent among Democrats. Trump backers agreement is highest with 84 percent….

There is a lower level of agreement, 64 percent, with the statement, “The old way of doing things no longer works and we need radical change.” Agreement is 71 percent among all Republicans and 58 percent among Democrats. Agreement is highest among Trump supporters with 83 percent.

Among all American voters, 56 percent agree with the statement, “Leaders don’t worry about what other people say; they follow their own path.” Agreement is 65 percent among all Republicans, 46 percent among Democrats and 74 percent of Trump voters agree, the highest of any candidate.

Additionally, 96% of Trump voters believe that “America needs a powerful political leader that will save us from the problems we face”, versus 80% of Republicans and 74% of Democrats. Of course, that might be partly due to Trump voters being the most dissatisfied.

Forty-three percent of registered voters blame both sides, while 29 percent of voters think it’s the protesters who are mostly to blame for these incidents and 23 percent mostly blame Donald Trump’s supporters.

Fifty percent of Republican primary voters, and eight in 10 Trump supporters, approve of how Trump is handling the violence. By contrast, among voters overall two in three disapprove of how Donald Trump is handling these incidents.

Independent voters are more similar to Democrats. 81% of Democrats and 68% of independents disapprove of his handling of the violence. So there’s not much evidence that the incidents had a major negative impact among those who weren’t already opposed to Trump. Some of this is just a barometer for whether you approved of him already. However, hearing Trump’s actual words on the violence makes him much less popular:

Take the following statement Trump made in Iowa on the day of the caucuses: “If you see somebody getting ready to throw a tomato, knock the crap out of ’em, would you? Seriously. Okay? Just knock the hell — I promise you, I will pay for the legal fees.”

Just 16 percent of voters viewed Trump more favorably after reading that statement. But a whopping 61 percent said it made them view him less favorably, including 45 percent of Republicans.

That pattern held across a range of similar statements, including:

  • “See, in the good old days this didn’t use to happen, because they used to treat them very rough. We’ve become very weak.”
  • “It was really amazing to watch,” in reference to seeing his supporters “taking out” a protester.


FiveThirtyEight took a interesting approach to test how tolerant supporters of a given candidate are towards groups they dislike. First they asked supporters of each candidate which group they liked least:

lewis-tolerance-3 (Source)

They then asked respondents if members of their least-liked group “should be banned from running for the U.S. Congress; should not be allowed to teach in public schools; should be outlawed; should be allowed to make a speech in this city; should have their phones tapped by the government; [or] should be allowed to hold public rallies here”. They found that “Trump supporters would grant about 40 percent of the rights asked about to the groups they dislike.” However, supporters of other candidates were statistically indistinguishable from those of Trump, including the Democrats, except for those of Kasich, who were a bit more tolerant. According to this at least, Trump voters are as likely to wish to curtail rights as anyone else.

There isn’t much polling on how Trump supporters view torture, but “82 percent of Republicans said torture is “often” or “sometimes” justified, compared with 53 percent of Democrats.”

Quinnipiac University released a poll testing the impact of associating a policy with Trump. One of the topics was torture. Some people were asked if they agreed with a plan “To broaden laws regarding the use of torture in interrogations of suspected terrorists.” 71% of Trump supporters somewhat agreed or strongly agreed, compared to 24% of non-Trump supporters (including Democrats). Another group was asked “Trump would like to broaden laws regarding the use of torture in interrogations of suspected terrorists. Is this something with which you strongly agree, somewhat agree, somewhat disagree, or strongly disagree?” When the policy was linked to Trump, 76% of Trump supporters agreed, versus 19% of non-Trump supporters.


The evidence that Trump voters have authoritarian personalities is weak, in my estimation. I’m not a statistician and I haven’t gone deep into the numbers, so please take my analysis with a grain of salt. But there are too many competing hypotheses and possibilities out there to endorse any of them with much certainty. But even limiting the analysis to more typical polling questions, one can find that Trump voters are more likely to support torture, more likely to approve of how Trump has handled violence at his rallies, and more likely to desire a strong leader. Whether these things do or do not indicate authoritarianism I leave to the reader.

Next: Are Trump voters best thought of as populists? And conclusions.

(Source for header image)

[Overview] Trump Voters VIII – Populism

[Part I; Part II; Part III; Part IV; Part V; Part VI; Part VII; Part VIII; Single page]

by David Severa

Social psychologist Jonathan Haidt, along with Emily Ekins, researched how the supporters of different candidates fit in with his moral foundations theory, the idea that people use a number of distinct mental axes to think about ethics, and that people vary by how important each axis is. Haidt has identified six foundations: care/harm, fairness/cheating, liberty/oppression, loyalty/betrayal, authority/subversion, and sanctity/degradation. So, when judging an action, one person may care whether or not authority is respected but not much care whether the action is particularly fair, while another person may do the reverse.

They polled voters to measure their moral foundations. (Only one of the authority questions is about child rearing, and they combined authority with some other foundations, so that measure may not be directly comparable to the other research on authoritarianism. Proportionality is equivalent to the fairness/cheating axis – whether you think “The world would be a better place if we let unsuccessful people fail and suffer the consequences” for instance.)


Trump voters don’t particularly stand out relative to the average American. They care less about care/harm (“morality requires caring for and protecting the vulnerable”) and more about proportionality (people getting what they deserve, “highly predictive of a preference for small government and a dislike of activist government and the welfare state”) and a combined measure of authority, loyalty, and sanctity, which the authors finger as the foundations of social conservatism. Their scores on those social conservative foundations are similar to those of other Republicans, and they place less weight on proportionality than some other Republicans, but more on care.

In further comments, Ekins says that:

Because Trump has a larger share of the vote, his supporters’ standardized scores for each moral foundation are more likely to be closer to average. Since 0 indicates the median voter, candidates with larger vote shares will appear closer to average.

They also tested “to see if adding the moral foundations improved prediction above and beyond the demographic factors that are usually used.” They did:

Among Republicans, four moral patterns stand out. First: Voters who still score high on authority/loyalty/sanctity and low on care — even after accounting for all the demographic variables — are significantly more likely to vote for Donald Trump. These are the true authoritarians — they value obedience while scoring low on compassion. They are different from Huckabee supporters, who appeared to score even higher than Trump supporters on authority/loyalty/sanctity, but also scored high on care, as we saw in Figure 1.

Authoritarianism is often assessed in social science research by asking people two or three questions about the relative importance of teaching their children obedience and respect, but our data shows a limitation of this approach: Not everyone who wants their kids to be obedient is an authoritarian. Some past research has been too quick to lump in social conservatives with authoritarians.

A further reason to doubt any simple interpretation of the “authoritarianism” measure.


We have done some preliminary tests finding roughly 2-3 groups [of Trump supporters]. One group scores particularly high on Loyalty-Authority-Sanctity, while another group appears to be less engaged and scores lower on all the foundations. When these different groups are averaged together, the average Trump supporter is closer to the average.

This research on morality and authoritarianism is interesting, but still preliminary I think. It’s perfectly plausible that Trump voters have a different psychological profile on average, but I’d like to see more work disentangling the effects of regional culture, and so on.


Populism isn’t any one thing, so there isn’t any simple test for populism. After all, Trump and Sanders have both been labeled populists and we’ve seen that their supporters have often radically different attitudes and beliefs. Populism is, perhaps, a style of thought that encompasses many different schools of thought. The Washington Post compares populism to authoritarianism:

Populism, on the other hand, is a type of political rhetoric that casts a virtuous “people” against nefarious elites and strident outsiders. Scholars measure populism in a variety of ways, but we focus on three central elements:

  • Belief that a few elites have absconded with the rightful sovereignty of the people;
  • Deep mistrust of any group that claims expertise;
  • Strong nationalist identity

Of course, authoritarians and populists can overlap and share dark tendencies toward nativism, racism and conspiracism. But they do have profoundly different perceptions of authority. Populists see themselves in opposition to elites of all kinds. Authoritarians see themselves as aligned with those in charge.

The authors polled voters to see if populist proved a better descriptor of Trump supporters than authoritarian. (I’ve already discussed this paper a few times, including in the section on authoritarianism.) They found that, while Trump voters were more authoritarian than average, populism mattered more:


“Trump voters are the only ones to score consistently high on all three populist dimensions.” I’d be interested in seeing more research on this question, but based on everything else I’ve discussed, I think that populism is a better lens to understand Trump than authoritarianism. There may well be authoritarian elements, but there’s more than that going on.

“Populist” is a useful, but vague, shorthand that captures a good deal about Trump’s appeal. I haven’t had time to discuss the history of right-wing populism in the United States (let me know if you’d like a series on that!), but I think that there’s a strong case to be made that Trump is best understood as part of that tradition. Remember that support for segregationist George Wallace in 1968 was a strong predictor for a county’s support for Trump in 2016. And they’ve drawn from similar wells of support: “overwhelmingly white, not college educated, racially intolerant, hostile toward immigrants, dismissive of social progress, and facing bleak economic prospects.” Analogies have even been made to Andrew Jackson’s candidacy almost two centuries ago, each being “a wealthy celebrity always ready for a fight, a superpatriot who says he will make America great again. He vows to attack government corruption and defend the common man.” And all three men with particular appeal to Appalachia. (Which isn’t to say that Trump supporters hold all the same views as their antecedents, but that they occupy a similar political position.)

I don’t know of a specific term to capture this tendency, but “Appalachian Populism” wouldn’t be too far off the mark, if perhaps too narrow. But a label is less useful than knowing everything to which the label refers.

Trump voters in review

After all these words, all these graphs, all these polls, all these numbers, it’s time to bring things to a close and review everything we’ve covered. What can we say about Trump supporters?

  • Certain basic demographic factors matter somewhat. Women, the young, those with college educations, and those with high incomes are all less likely to support Trump, though not always by much, and even so Trump received more of his support from likelier-to-vote high income Republicans. It’s important to remember that Trump was still often winning pluralities of the groups he did worse with.
  • He didn’t bring in many Democrats or independents, but he did energize weakly attached Republicans.
  • Racially, Trump voters closely resemble other Republican primary voters: both are over 90% white. But among white ethnicities, his supporters stand out, drawing especially heavily from those regions in and around Appalachia that claim “American” ancestry, and in regions, like in the Northeast, with ancestry from Catholic countries like Italy.
  • Those Trumpist regions share other characteristics, many of them related to atomization and social and economic decay. These are areas that have lost manufacturing jobs, areas with low work force participation, areas with weak religious ties, areas that are poorly educated, areas with high mortality. The parts of white America that are doing worst, that have been left behind. But, and I can’t emphasize this enough, with exceptions like Utah, Trump did pretty well most everywhere. (That is, pulling in at least a quarter of the votes.)
  • Additionally, although Trump did disproportionately well with poorer voters, they did not vote in large enough numbers to make up the bulk of his support. The average Trump supporter (like the average supporter of every other candidate) had an above average income. Trump voters do report more financial problems, and it’s possible that, being older, their incomes overstate their real socioeconomic position. But still, it seems more plausible to describe the typical Trump supporter as someone doing alright for themselves in a region that is decidedly not.
  • Trump voters have what I’ve chosen to call ethnocentric attitudes, above and beyond those of the typical Republican (who are more ethnocentric than non-Republicans). They are more likely than other Republicans to say that most blacks are violent and lazy. A majority think that Muslims should be subject to increased scrutiny solely because of their religion and that all Muslims should be temporarily banned from entering the United States. They have the most hostile attitudes towards immigration on every measure. This ethnocentrism doesn’t only matter for attitudes towards outgroups. There is also a sense of white ethnic solidarity. Voters who think their identity as whites is extremely important are more likely to vote Trump. Voters who perceive a great deal of discrimination against whites are more likely to vote Trump. These factors simply cannot be wholly reduced to economic concerns. Attitudes towards race matter. Again, not all Trump supporters share ethnocentric attitudes and not all those with strong ethnocentric attitudes voted for Trump. But it is predictive.
  • Republicans are angry, but Trump voters are angrier. Only 1% said that they were content with the federal government. 75% said that life had been better for people like them 50 years ago.
  • Despite this anger at the government, they have more favorable opinions of individual government programs than other Republicans. They are more likely to want to increase spending and raise taxes on the rich. They are the most opposed to free trade. They are not particularly economically conservative, though they are more conservative than Democrats.
  • They are socially conservative, but less so than many other Republicans. They are fairly evenly split on abortion, for instance. Trump voters are probably more socially conservative than the establishment Republicans who backed Bush or Rubio, but less so than those who backed Cruz. Certainly this is not an area where Trump supporters stand out and, unsurprisingly, Trump voters are mostly not inspired by these kinds of social issues.
  • Trump voters are the most opposed to American engagement with the rest of the world with the sole exception of using military force. They are, if anything, more militaristic than anyone else while also being the most hostile to trade, to foreign aid, and even to NATO. I called it “nationalism that believes we live in a zero-sum world”. It somewhat closely matches what Walter Russell Mead called Jacksonian foreign policy.
  • There is evidence that Trump voters are more authoritarian than the average American, but the evidence that they are more authoritarian than other Republicans is more uncertain. I’m not sure how useful the attempts to psychologically pigeonhole Trump supporters are at this point. Still, Trump voters are likeliest to want a strong leader, who will “say or do anything to solve America’s problems”.
  • They are the most likely to support expanding the use of torture.
  • Probably the best way to encapsulate the world and worldview of the typical Trump voter is what I called Appalachian Populism, though it isn’t confined to Appalachia. White ethnocentrism, an openness to some kinds of government largesse, a militaristic foreign policy, and a distrust of experts, all in regions that are socially and economically struggling. That is the core of Donald Trump’s support.

Thanks for reading! Up next, a dialogue on objective morality.

(Source for header image)