The False Inevitability of Donald Trump
George Mason Scalia Law Professor, and staunch Trump supporter, Frank Buckley, Jr., published a Wall Street Journal editorial explaining Trump's victory and claiming Democrats face insurmountable challenges. See here. Buckley's real message was to conservatives: either get on the Trump train or become irrelevant. Thankfully, Buckley's analysis is profoundly flawed.
Buckley's op-ed, "How Trump Won, in Two Dimensions," a happy story for him and like-minded Trump supports, rests on his interpretation of a single graphic from a more comprehensive study by political scientist, Lee Drutman. The study, "Political Divisions in 2016 and Beyond: Tensions Between and Within the Two Parties" (June 2017), available here, relied on data from the Democracy Fund Voter Study Group. The data are based on 12 sets of topic-specific questions to 2016 voters. The graphic, which forms the centerpiece of Buckley's analysis, the second of 16 detailed graphics in Drutman's study, is reproduced here:
The graphic plots voters based upon two dimensions, economic liberalism (left side) or conservatism (right side), and social liberalism (lower half) or conservatism (upper half). This creates the following four quadrants, starting with the upper left and moving clockwise:
Populists (upper left): economically liberal and socially conservative, comprising 28.9% of the electorate;
Traditional conservatives (upper right): economically conservative and socially conservative, comprising 22.7% of the electorate:
Libertarians (lower right): economically conservative and socially liberal, comprising 3.8% of the electorate; and
Traditional liberals (lower left): economically liberal and socially liberal, comprising 44.6% of the electorate.
A quick bit of computation, simple addition really, is helpful in assessing the underlying graphic and Buckley's interpretation of it. We will sum the numbers both vertically and horizontally, thus identifying the total percentages of economic liberals or conservatives, and social liberals or conservatives. The results: those identifying as social conservatives total 51.6%, and those identifying as social liberals total 48.4%. Summing vertically, those identifying as economically liberals total 73.5%, and those identifying as economical conservatives total 26.5%.
Buckley's analysis hinges critically on the (vertical) liberal economic dimention total, 73.5%. Buckley states: "By Mr. Drutman’s count, 73% of all voters were left of center on economics. Most of the remaining Trump supporters were quite moderate on economic questions." This leads to Buckley's central claim: "The crucial differences between the two parties came down to social concerns, including pride in America, immigration, and especially moral issues such as abortion and gay marriage." Now the clincher: Today's Democratic party leaders hold core commitments on these issues. And because they are unwilling to tack right (speaking in conventional political terms), or using the graphic to tack upward, to attract populists (economic liberals and social conservatives), Trump's rise and ongoing success is inevitable.
The real, albeit unstated, point of the article, which, after all, was written by Buckley and published in the Wall Street Journal, was not to give wise counsel to Democratic leaders. No, Buckley was, instead, offering strong-arm counsel to Republicans: It's time to stop any Pence scheming, or any Cruz or Kasich nonsense. By Buckley's lights, the only one who gets it right is Sean Hannity: Even if you can't afford Trump Tower, conservatives, you had better board the Trump Train. It's that, or become irrelevant.
To see how this all falls apart, we have to quickly return to Drutman's study. Drutman discusses 12 issue dimensions, but he doesn't really mean dimensions in the sense that I have repeatedly explained that term in my scholarship and on this blog. See here and here. There is an important difference between policy issues and dimensions. Parties form coalitions of voters who care about large numbers of issues, and we can routinely align countless issues along a single, conventional, left-right axis to capture them. I explained this in some length in my post titled "The Dimensionality of Trumpism." See here. There I maintained that, contrary to the conventional left-right axis, both the Trump and Sanders campaigns potentially implicated an "inside-outside" or "establishment-challenger" axis. I further posited that, inevitably, after an outsider such as Trump wins, he becomes an insider, and at most, he winds up affecting the eventual locations of the endpoints that previously defined that left-right axis. Although I will not claim that it proves my thesis, I will say that the Drutman study, and this one graphic, is more consistent with my thesis than with Buckley's.
We need a baseline for any claim that voters are left or right of center, including with respect to economic or social issues. It makes no sense to claim that 73.5 percent of voters are more liberal than the average, whether average is defined in terms of mean or median. Average is average, the midpoint, or 50%, whichever measure one uses. (I do want to be clear that if you ask people to self identify as liberal or conservative on any given issue, then as Drutman's study rightly observes, you can wind up with more than half appearing left of center. My point is that when you aggregate the data, this reveals a mistaken general understanding of the center's actual location on that dimension.) This implies that the center vertical dividing line separating economic liberals and economic conservatives must move substantially to the left. Given the density of economic centrists who are also social conservatives, this results in moving a large number of Trump voters from the populist quadrant into the traditional conservative quadrant. Although I don't have the underlying data, from a visual survey, it is clear that after that move, the division of populists between Clinton and Trump is far more even.
The graphic's social dimension, dividing the conservatives as 51.6% and liberals as 48.4% is closer to 50/50. Adjusting the graphic implies moving the horizontal dividing line slightly upward, pulling into the traditional liberal quadrant more blue than red dots, but still moving some red dots, and also moving from the traditional conservative box to libertarian box some scattered red dots. Those combined moves have two dramatic consequences, neither of which bodes well for Buckley.
First, moving the economic line left has two important consequences. Clinton's appeal to the populist box is not necessarily inspiring, but it is greatly improved. Conversely, what Buckley presents as Trump's great appeal to populists becomes far less inspiring as so many of those voters, clustered closely to the traditional conservative base, are now rightly placed within that base. This implies that populists are, in fact, nervous about Trump's economic policies, including quite possibly his thus far unsuccessful ACA repeal, which likely disproportionately burdens such voters. Trump's pull among libertarians is also marginally improved, but as Buckley concedes, the numbers there are far too small to matter.
There is a second, deeper, point, which also affects the methodology of the larger Drutman study. Trump received 62,984,825 popular votes, and Clinton received 65,853,516 popular votes. See here. In percentage terms, thats 46.4 % for Trump and 48.5 % for Clinton. Let's assume that we do the shifting of the mean (or median) divisions described above. Drutman isolated the social and economic dimensions in the preceding graph because with the percentages depicted based on the questionable dividing lines, there appeared to be a notable enough split to demonstrate a potential dimensionality problem. With the changes that I've suggested to accommodate the real centers on each dimension, that's far less clear. The divisions are now closer to even in the populist category. And so it becomes increasingly plausible that these two dimensions actually collapse back into one: liberal, embracing traditional liberals, some populists (upper left), and a smaller number of libertarians (lower right); and conservative, embracing traditional conservatives, and also some from the upper left and lower right. To the extent the voters in the other quadrants roughly cancel each other out, we no longer implicate a second dimension and can capture this in left-right terms. Were we to reconcile the graphic that way, the median point of the single liberal-conservative dimension would necessarily be on Clinton's side. After all, she received nearly 3 million more popular votes than Trump out of approximately 129 million votes cast.
Twelve dimensions are really not twelve dimensions. They are twelve sets of bundled issues that for the most part can be captured in conventional liberal-conservative terms. Same with the two dimensions depicted in the graphic. I previously posted that there might have been a second dimension in the election, and Drutman's study, at a minimum, appears not to contradict that claim. Drutman states: "The data suggest that the main divide within the Democratic Party electorate is about attitudes toward the establishment and the existing order than it is about specific issue positions (with the exception of trade policy)." (See here p.20). People don't vote strictly based on trade policy, at least in significant numbers. And this assertion is about Democrats! Surely we know that Trump won the Republican primary by challenging establishment politics, defeating a field overwhelmed by establishment candidates, even as his supports are now learning each day that knowledge, integrity, character, and competence turn out to actually matter.
In the end, Buckley is mistaken in his interpretation of the graph. The stakes are high, and the Wall Street Journal publication risks giving Buckley's analysis more prominance than it deserves. This president is reprehensible, and it is getting worse daily. We have gone from offensive personal conduct to such irresponsible behavior that he threatens potentially devastating consequences. I'm just going to come out and say it: I agree with Frank Bruni that Trump has proven himself beyond redemption. See here. My advice won't be subtextual; it will be quite direct. It is now time for Republicans to do the right thing, meaning to initiate lawful steps to have this man removed from office. It is not time to listen to dubious stories, or subtle threats, that unless conservatives do whatever this despicable man wishes, they will no longer have a political future.
I welcome your comments. More so, I hope you will consider sharing this post with other. I've not said that before, but it is time to stop playing around.
[Author's note: I would like to thank Lee Drutman for his helpful response to an email inquiry on methodology.]