On Super Tuesday, Donald Trump continued his surprise series of wins, gaining seven out of the 11 states in play that day. In response, many are asking: Why have establishment candidates had such a hard time in this year’s Republican nomination contest?
Some suggest that Republicans who bother to vote in the primaries are farther to the right than the Republicans as a whole – and are unimpressed with the somewhat more moderate establishment candidates. Others suggest that less-educated citizens who don’t usually vote are turning out just to vote for Trump.
These may be part of the answer. But we offer a different suggestion to explain Trump’s meteoric rise: The GOP establishment is out of touch with its base. Trump actually represents Republicans better if you examine beliefs issue by issue.
To investigate, we have teamed up with Pollfish, a start-up survey platform, to track American public opinion on 11 key campaign issues weekly during the election season. Pollfish’s survey platform delivers online surveys to almost 200 million potential respondents. We specifically target smartphone owners in the United States. Our sampling technique is best described as randomly targeting people of a sample of adults in the U.S. with smartphones. Using highly technical statistical techniques to adjust for non-response (a legitimate concern in all surveys) coupled with demographic breakdowns of the voting population derived from Big Data on all registered voters in the U.S., we monitor public opinion on these issues broken down by the general population and by party, and compare it to leading political figures in the 2016 primary season.
Let’s look here at opinion on three highly divisive and partisan issues:
- Abortion. We track agreement with a conservatively anchored position: “Do you support abortion in cases of rape/incest?.”
- Gun rights. We track agreement with a neutrally anchored position” “How do you feel about federal laws to make it more difficult for people to buy a gun?”
- Income inequality. We track agreement with a liberally anchored position: “How do you feel about government measures to reduce differences in income levels?”
On all three issues, most Democratic politicians (including both Hillary Clinton and Bernie Sanders) agree very strongly or strongly. The “establishment” Republican candidates (including both Sens. Ted Cruz and Marco Rubio) disagree very strongly or strongly.
But where do Republican voters stand? Let’s compare them to the general population to see.
If you look at the figure below, you can see that only 20 percent of the general population supports a complete abortion ban, even in cases of rape and incest – and only 25 percent of Republican voters do. Cruz and Rubio advocate this position. Trump breaks with the GOP establishment – and agrees with the nearly 60 percent of Republicans who support the right to abortion in these circumstances.
Less than 35 percent of the general population would oppose any additional gun regulation or control. More than 50 percent support at least some increased gun regulation.
On both abortion and guns, less than 15 percent of the general population holds the neutral position; 75 percent hold strong or very strong views. On guns, the Republican voters agree with their elites; all Democratic candidates favor increased gun control while all of Republican candidates favor no additional gun control.
We find that 35 percent of the general population strongly support government action to reduce differences in income inequality; 20 percent strongly oppose; and 45 percent hold positions that are either neutral or soft. However, there is a partisan divide. Most Democratic voters do strongly support such measures. Sanders strongly advocates this position; we have defined Clinton as neutral on the issue, both in her rhetoric and her policy proposals, although this may change in coming months.
Republican voters are less clearly aligned – and are certainly not as cleanly aligned with the Republican elite. Trump, Cruz, and Rubio’s tax plans would massively redistribute after-tax income to the highest earners, although that’s not what their rhetoric says. Despite proposing the largest tax cuts, Trump calls CEO pay “a joke and disgraceful,” but offers no policy solutions, and has specifically advocated for the end of the hedge fund carried interest loophole.
What’s the takeaway?
We asked a total of eight questions, all framed to measure whether people hold strong or moderate positions on hot-button issues. We consistently found that people were more likely to hold strong opinions on one side or the other than they were to be neutral or mildly opinionated. In other words, public opinion was very polarized. This indicates that elite polarization (i.e., the distance between the policy positions of Republican and Democratic lawmakers) has certainly rubbed off on the general population, at least to a small extent.
But here’s something interesting. Across these three representative issues, 60 percent of the Democratic electorate supports Democratic candidates’ policies strongly (or more so). But only 35 percent of the Republican electorate supports their candidates’ positions.
Does this misrepresentation matter to the voters? It’s hard to know. In 2012, 92 percent of self-identified Democrats voted for Barack Obama, while 93 percent of self-identified Republicans voted for Mitt Romney. Only 6 and 7 percent, respectively, crossed over to vote for the “other” party’s candidate.
But the Democratic candidates are further from the Republican base than Trump is. And Trump’s breaks with Republican orthodoxy may actually be closer to the beliefs of the Republican base. He breaks left to match Republican voters’ beliefs on abortion – and he breaks even further right to match those voters in his extreme positions on on opposing immigration and any pathway to citizenship for undocumented immigrants.
In other words, Trump’s apparent hodgepodge of moderate and conservative positions may actually make him a more representative Republican presidential candidate than the establishment candidates we’ve seen in the past eight years.
It’s certainly possible that Trump’s candidacy will pull the Republican Party back towards the views of Republican voters.
Notes on methodology: We sampled about 1,000 American smartphone users once a week for five weeks between Jan. 11 and Feb. 12, 2016. The total respondents is 4,209. We then model each question using a Bayesian hierarchical regression with age, race, gender, education, race, and party identification. We then weight the obtained probability estimates based on total counts of likely voters derived from the TargetSmart voter file.
Tobias Konitzer is a Ph.D. candidate in communication at Stanford University.
David Rothschild is an economist at Microsoft Research. Find him on Twitter @DavMicRot and PredictWise.com