Every few years a debate re-emerges on the internet as to whether university faculty have truly shifted to the left, and if so, whether it matters. The debate has just flared up because of a graph that I made after some discussions about ideology in the academy with my friend Jon Haidt, who wanted to document the trend here at Heterodox Academy. The graph (Figure 1) got picked up and debated at a number of websites, including The American Interest, The American Conservative, Mother Jones, and Bloomberg.

But Paul Krugman dismissed the graph and the problem on his blog at the New York Times:

Conservatives are outraged at what they see as a sharp leftward movement in the academy [he then showed figure 1] But what’s really happening here? Did professors move left, or did the meaning of conservatism in America change in a way that drove scholars away?

Krugman notes that the identities of the two political parties were changing in the 1990s, and by some metrics the Republican party moved further to the right than the Democratic party did to the left. He concludes:

Overall, the evidence looks a lot more consistent with a story that has academics rejecting a conservative party that has moved sharply right than it does with a story in which academics have moved left.

I agree that the changing nature of the parties could have caused some academics to simply change their political self-descriptions without actually changing their views on substantive matters. And if that process explained all of the changes in Figure 1, then it would indeed absolve the academy: “Hey, it’s not we who changed, it’s the Republicans!”

But is the process Krugman refers to so powerful that it can explain the full trend, or even most of it? I don’t think so, and here’s why.

First let me explain how I made the graph. The Higher Education Research Institute (HERI) at UCLA has been conducting triennial surveys of undergraduate teaching faculty for the past 25 years. The HERI samples are huge — tens of thousands of professors — and this is a robust and well-executed survey. The data is comparable and responsibly collected over a long period of time. The survey includes a question asking respondents to describe themselves using a 5-point ideology scale that offers these options: “Far left,” “Liberal,” “Moderate,” “Conservative,” and “Far right.” I merged far left with liberal, and I merged far right with conservative, so that we can see the big picture most clearly. Figure 1 plots the percentage of respondents who fell into those three groups, in each of the 3-year data collection periods. The data is weighted to represent the national population of full-time faculty with teaching responsibilities for undergraduate students.

Figure 1 reveals a striking ideological change among faculty over time. While the data confirms that university and college faculty have long leaned left, a notable shift began in the middle of the 1990s as the Greatest Generation was leaving the stage and the last Baby Boomers were taking up teaching positions. Between 1995 and 2010, members of the academy went from leaning left to being almost entirely on the left. Moderates declined by nearly a quarter and conservatives decreased by nearly a third.

It certainly looks like a very large change that happened in just 15 years. Could it have happened for the reason that Krugman proposed? We can begin to answer that question by looking at broader national trends. We can compare trends in the academy to shifts in partisan and ideological definitions that were occurring to the electorate as a whole. To do that I examined the GSS dataset.

Figure 2 presents data from the General Social Survey. I plotted self-reported ideology over time for the population as a whole. The figured reveals remarkable stability from 1974 through 2014 with moderates making up a plurality of Americans. If we look at the 25 year period that the HERI data encompasses, the overall average shows that 38% of Americans are moderates, 35% are conservative, and 27% are liberal. There is little change over the 25 year period despite the changing definitions of ideology and partisan sorting that Krugman referred to.

If we compare figures 1 and 2, we see that the professoriate was changing while the electorate as a whole was not. Professors were more liberal than the country in 1990, but only by about 11 percentage points. By 2013, the gap had tripled; it is now more than 30 points. It seems reasonable to conclude that it is academics who shifted, as there is no equivalent movement among the masses whatsoever.

The people who shape the minds of America’s students have long leaned left, on average. But students who entered college before 1990 could count on the fact that their professors did not all vote the same way or hold the same views on the controversial issues of the day. Students who arrived after 2005 could make no such assumption. For example, Figure 3 is from Jon Haidt’s recent post plotting new data on the policy views of social psychologists, on nine culture war issues (such as abortion and gay marriage).

Only one social psychologist, out of a sample of 327, had views that were right of center. This graph is incompatible with Krugman’s hypothesis that professors didn’t change their views, they just changed their labels.

I’d like to end this post on a personal note. I joined the faculty of Sarah Lawrence College in 2010. When I told my friends from graduate school about my new job, they warned me to keep my head down and avoid discussing political topics with other faculty. This suggestion course, was hard to follow as I was explicitly hired to teach American politics at Sarah Lawrence College, a college which happens to be one of our nation’s most well-known progressive and politically active institutions. I will never forget my second day of teaching at the College where I was called a right-wing wingnut. That accusation was made without any hesitation when I pointed out that an empirical trend suggested a very different policy outcome than what was being discussed around a lunch table. I teach and have long-followed the Daniel Patrick Moynihan mantra of “Everyone is entitled to their own opinions, but they are not entitled to their own facts.” It is important to follow this both intellectually and pedagogically and I continue to push this idea regularly in the face of dissent.

This is what happens when viewpoint diversity disappears and orthodoxy reigns. When faculty in the social sciences can no longer have open discussions among themselves about political issues because dissent from the progressive stance is treated as treason, then what kind of political extremism and intolerance will we breed among our students? The answer, unfortunately, became clear in the demands that students issued on campus after campus last Fall, and in the cruelty and aggression some of them showed to anyone they deemed to be an opponent.

Orthodoxy is fundamentally incompatible with the mission of a university. Increasing viewpoint diversity should be at the top of every college president’s priority list.