In The New York Times’ Gray Matter section on July 3rd, I had the chance to elaborate on earlier work shared on the Heterodox Academy site where I first showed that faculty members in our colleges and universities have shifted ideologically to the left since the 1990s. Figure 1 presents another view of that trend but this time presents the ratio of liberal faculty to conservative faculty. While liberal faculty have always outnumbered conservative faculty, the figure makes it clear that since 1995, the relative number of liberals on campuses has been increasing. The liberal-conservative ratio among faculty was roughly 2 to 1 in 1995. By 2004 that figure jumped to almost 3 to 1. While seemingly insignificant, that represents a 50% decline in conservative identifiers on campuses. After 2004, the ratio changed even more dramatically and by 2010, was close to 5 to 1 nationally. This shows that political diversity declined rapidly in our nation’s centers for learning and social change.
Comparing the faculty change on college campuses as I did in my earlier post with national trends only makes academia’s hard turn to the left more clear. In Figure 1, I also present data from Gallup on the ideology of the general public. Because the ideology items coded and queried by Gallup are identical to those asked of the faculty it is similar to the data to in my earlier post. Once again, I present the ratio of liberals to conservatives and what is immediately clear in the figure is that the 25 year trend is for the American polity essentially flat. There has been virtually no change in the ratio whatsoever among the general public, compared to a 150% increase on the faculty side. Moreover, given that the average ratio over the 25 years is about .52, the trend makes the point that liberal identifiers are lower than conservative identifiers and that the relative number has not shifted much as the slope of the line has remained flat over the past decades. The minor variations over the 25 year period for Americans despite the changing definitions of ideology and partisan sorting among the electorate as a whole suggest that the transformation is very real on the faculty side.
Figure 1. The Ideology of the American Professoriate and Polity Compared: 1989 – 2014
Source: Higher Education Research Institute, UCLA; Gallup and CBS News/NYTimes
Thinking about the national trend in Figure 1 compared to faculty, one could argue that in aggregate, I did not account for increasing liberalism present in those Americans who hold a graduate degree. According to the General Social Survey from 2014, of those who have a graduate degree of some sort, 30% identify as liberal or extremely liberal and 17% are conservative or extremely conservative. This is a liberal: conservative ratio of 1.8 – a far cry from 5 among college professors. Additionally, in contrast to graduate degree holders, those Americans who have a high school degree or less, 13% identify as liberal or extremely liberal and 20% identify as are conservative or extremely conservative – a figure more in line with the national average. So, even if definitions of partisanship and ideology have shifted over time, Figure 1 makes it very clear that in the aggregate those who shape the minds of many in college are liberal and the shift is nontrivial. It seems reasonable to conclude that it is academics who shifted, as there is no equivalent movement among the masses whatsoever over the past 25 years.
In addition to the national trend presented here, I also found and reported in The New York Times that the leftward shift was not uniform. I looked at factors such as the differences between tenured faculty compared to untenured, income and salaries, school type and selectivity, departmental affiliation and age – all factors which have been known to impact one’s political ideology – and the data was muddy. So, I began digging more deeply.
Surprisingly, the factor that had the greatest impact on the ideological leanings of college professors was where they lived and worked- namely, their region of the country. Figure 2 presents liberal: conservative ratios for the 8 regions in the data that were included in the original survey data. While 7 of the regions move up and down over the decades, 7 regions have clearly shifted to the left and one – the Rocky Mountain region – has shifted slightly to the right. The huge outlier is the stark leftward move of those college and university faculty in New England – those who work in Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont – is truly pronounced.
Figure 2. Regional Ideological Variations of Americas Professors: 1989 – 2014
Source: Higher Education Research Institute, UCLA
Figure 3 presents the same data in a slightly different way. Here, I have plotted New England as its own region and collapsed the remaining seven regions into one category and included the national trend line from Figure 1 as well. This enables us to see just how great the change has been in New England over the past two decades. As stated in the New York Times, what is so remarkable here is that those faculty who reside in New England are far more liberal than anywhere else in the nation, even controlling for discipline and institution type (i.e. Christian college vs. large public university).
Figure 3. New England and American Ideologies Compared: 1989 – 2014
Finally, the question of why New England and its professors are so far left compared to the rest mentioned “inertia” in The New York Times and that requires some elaboration.
Many New England institutions of higher education have long been models for the county as sacred places – often with explicit religious connections when they were established – where students and faculty study the liberal arts and humanities without much concern for practical applications. Harvard College students, for instance, famously head to MIT to take courses in finance and accounting. Yale’s student body complained about the abstract and academic nature of its computer science department which did not offer basic courses in programming and development. Yale now imports a popular programming course, ironically, from Harvard.
The Harvard and Yale examples differ from the Rocky Mountain schools and are in direct opposition to many land/sea/space-grant colleges which were charted to explicitly engage in practical and useful pursuits. The Colorado School of Mines famously developed a teaching and research curriculum “geared towards responsible stewardship of the earth and its resources… [such that the institution has] broad expertise in resource exploration, extraction, production, and utilization.” The University of Montana states that is capitalizes on its strengths to offer “programs and services responsive to the needs of Montanans” – a stark contrast to Amherst, for instance, which asserts that, “Since 1821, we’ve been helping our students find their own voices, discover their own truths and forge their own paths in the world.” Amherst, a selective liberal arts college located in rural Massachusetts in New England, explicitly makes that case that the school is centered on inward self-discovery while the University of Montana is just the opposite and focuses on the larger community. Whiles these cases are just a few examples, it does appear that the very rationale behind New England schools and their goals are different from those in other parts of the United States. While I make no value judgement here, the divergent outlooks could absolutely inform ideological thinking and hiring practices which could lead to New England moving so far to the left.
We also have to consider the principle of homophily. As someone who has been involved in many faculty searches as a committee member and as a vetter for other institutions, professors like to hire people that are close to their own ideas and training as these colleagues could last a lifetime. Minimizing friction and dissent makes life much easier. While many fields deal with non-political content, politics is pervasive outside the classroom and political questions are hard to ignore even in fields like engineering and bio-physics as funding has become politicized.
Faculty on college campuses have been shifting left for two decades, with those in New England shifting much faster. I cannot explain why. What I can say, however, is this: if you want to be exposed to a wide range of viewpoints and avoid uniformly left-wing faculties, you should probably focus your search on schools outside of New England.
Key Takeaways:
- There has been virtually no change in the liberal to conservative ratio whatsoever among the general public, compared to a 150% increase on the faculty side.
- The factor that had the greatest impact on the ideological leanings of college professors was where they lived and worked- namely, their region of the country.
- New England leads the way toward the left in academia.
“The steep change in 2001 implies a mass change of minds in NE–it’s so rapid. So what explanation can account for that?”
Can no one think anything that happened in 2001 that could have sparked a paradigm shift? An individual who experiences traumatic loss can be prone to dramatic changes in ideologies. In New England, in 2001, there were many, many individuals who experienced traumatic losses.
Could this not have created circumstances that might make an individual who strongly disagrees with a single element of a conservative political party (such as going to war with Iraq) reconsider their overall position?
I immediately think of two men I know, one an admiral in the US Navy and the other an award winning poly sci teacher at a New England prep school. Both unwavering in their conservatism up until the early 00s, and both of whom became dramatically liberal in the space of a few years.
There are a lot of reasons why this trend might be particularly strong in New England, some obvious, such as the percentage of people personally affected by the 9/11 attacks, or its proximity to NYC, and some more complex, like the secularism of the region. And, of course, this is likely a single thread of a much trickier web.
But New England is a region, like others have noted, with a culture unto itself, and once a paradigm shift takes hold, it develops a life and a will unto its own.
Be honest: ‘diversity’ makes free speech impossible
Diversity means everyone MUST think alike about ‘diversity’. Diversity just means White Genocide.
Freedom now from this enforced, coercive, parasitic, anti-white, genocidal diversity. Its a crime, not a ‘policy option’
I wonder if New England colleges are more likely to place undergraduates in media roles, evidence that would support the claim on liberal media bias.
I don’t know how one states it politely, but one-third of graduate degrees are in education. That has the lowest entrance standards, has the least-rigorous curriculum, and moves people into government jobs. This latter doesn’t make them all liberals – conservative teachers are often as willing to put up with crap to improve their careers as anyone else – but I’ve got to think it’s got to tip badly leftist, and not quite so much graduate school as other fields. So the picture of people with graduate degrees being more liberal may be technically accurate, but misleading.
The professors want to live unchallenged. If they were forced to live and work in a place where they might have to interact with those who don’t agree with them, they,would be traumatized.
I wonder how the professors at Historically Black Colleges rate on this scale?
This kind of bizarre comment doesn’t help your case.
Art Deco, your suggestion seems like it would be the most logical answer to the broader question of why the professoriate might be trending left. I’m not sure it addresses the New England piece, though, unless New England schools are particularly vengeful compared to the rest of the country. That’s part of why I think it’s a sampling problem. My back-of-the envelope calculation is that New England schools comprise about 6% of all US colleges & universities, so it would be easy for the data to be skewed.
The other problem with that story is that under your scenario, you would expect to see lower percentages of conservatives among associate profs, right? That is, people tenured sometime in the 1990s or 2000s. Not only is the percentage of conservatives among full, assistant, and associates virtually the same over this time period (nationally, anyway), but it moves up and down at almost the exact same rate between 2004 and 2014. That suggests to me, again, that it is a matter of case selection, the weighting process, something like that.
Vengeful? No, they just need to be conceited, complaisant, and get no pushback. After 2002, they’d have gotten none if my hypothesis is right.
New England’s not Wyoming. The larger society is different and you have a large mass of private colleges and private research university who have only a truncated vocational component (and, keep in mind some vocational programs are as bad as the arts and science faculty – law, teacher training, social work, &c). Harvard’s got a law school and a teachers’ college.
There’s a non-zero share of faculty who fancy that dissidents are dissidents because they’re not ‘well-informed’, or that dissident scholarship is some sort of propaganda game, or that dissidents are peddlers of something akin to astrology or phrenology, or that of course no one wants them around because they’re hostile to the intellect (something conjoined to rather fanciful notions about what the political parties are like or have been like). Get on a comment board with pseudonymous academics, and you hear these sorts of things. You hear them in public fora. I can name a few names.
I’m constructing my hypothesis with the assumption that tenured dissenters were a small corps in 1970 and that no further additions were made. Measures of dissenting faculty would show a corps of untenured faculty of stable dimensions as new cohorts were added and the older cohorts were denied tenure. The sum of tenured and untenured faculty would decline slowly as the small corps of tenured professors retired. You don’t see this because measures from the period running from 1970 to 1989 are not present. I suspect it would be partially masked during the period running from 1989 to 2002 because the novice faculty in that period were drawn from the birth cohorts (1958-70) who were Republicans in their late adolescent years (recall both Ronald Reagan and George Bush the Elder won a majority of voters under 25).
A great deal of backing and filling, I realize.
Here’s a suggestion as to why the graph looks the way it does:
1. Around about 1970, they quit granting tenure to faculty members who had certain properties, properties discovered during their assistant professorship.
2. Through the course of the period running from 1970 to 2000, the dissenting faculty consists of (a) untenured and temporary faculty and (b) professors tenured prior to 1970.
3. Around about 2002, the number of tenured faculty from the older cohorts drops dramatically as they retire.
4. Around about 2002, they quit hiring new faculty who studied at certain venues, lack recommendations from established researchers known to be ‘sound’, etc.
5. The number of dissident faculty drops like a stone, leading to what you see there. What you see between 2007 and 2011 would be the pre-2002 hires being denied tenure.
Art Deco – I think your overall reasoning is sound, but I will add two more points relevant to colleges in New England:
1. The general public in these areas, perhaps excepting New Hampshire, is far more “liberal” (“progressive”) than in other parts of the united States.
2. This is where the bulk of nationally recognized “elite” universities are – so this is where the stakes have been highest and this is where progressives have concentrated all their intellectual firepower, in order to control the commanding heights of higher education in the united States.
2.a. Elite institutions tend to have large endowments, which means they can better afford to be insulated from reality and from consequences for their errors.
2.b. Elite institutions are more likely to have close connections to long-established funding sources which themselves are insulated from reality and from consequences for their errors.
This is where the bulk of nationally recognized “elite” universities are
Just a quibble: it’s where Harvard and Yale are, and these are collecting pools of people with a certain sort of ambition. By way of example, look at our competitive presidential candidates over the last generation. There’s dozens who attended either college or professional school at one of these universities. Now look at other private research universities (including the rest of the Ivy League) and how many such alumni do you see.
Duke (Richard Nixon, Ron Paul)
Princeton (Ted Cruz, Steve Forbes)
Columbia (BO, Pat Buchanan)
Stanford (Frank Church, Henry Jackson)
Emory (Newt Gingrich)
Dartmouth (Nelson Rockefeller)
Penn (Donald Trump)
University of Chicago (Bernie Sanders)
Georgetown (Bilge Clinton, Pat Buchanan)
Cornell (nada)
Brown (nada)
MIT (nada
Johns Hopkins (nada)
Carnegie-Mellon (nada)
CalTech (nada)
Northwestern (nada)
Claremont Colleges (nada)
Rice (nada)
It’s no different half-a-ratchet down: RPI (nada), University of Rochester (nada), Tufts (nada), Brandeis (nada).
Now look at our most selective private colleges. AFAIK, only Pat Robertson, Michael Dukakis, Ron Paul, and Helligula attended any such institution. You’ve got about 150 private colleges with some cachet, and only Washington and Lee, Swarthmore, Gettysburg, and Wellesley make even one appearance.
Half a generation ago, Norman Cantor maintained that for all their elaborate procedures, Harvard had a history department ‘no better than 10 others.’ Recently Harvard has been making efforts to bulk up their engineering school. Suggest Harvard and Yale are over-rated above and beyond the gross-v-net scam that is elite education in this country.
I agree on this point; a random sample of New England schools would look little different from the rest of the country with regard to that sort of “elite-ness.” Getting back to the HERI survey, though, you should note that neither Harvard nor Yale, nor any other Ivies, are included in the sample for 2011 or 2014. The New England schools include a few large state schools (UConn, URI), some selective private colleges (Williams, Hampshire), and a handful of less selective places. So if there is something going on with New England schools (and I’m still not convinced there is), it’s not necessarily, or not exclusively, the elite universities.
Like the last commenter, I think there is a clear trend here but I am skeptical of the effort to measure the change in New England. To put some numbers on this, the data here would suggest that (assuming about 30% of respondents identify as moderate) there were about 9 self-identified conservatives per 100 in NE in 2004, but only about 2 or 3 in 2014. So approximately 3/4 of New England conservatives retired/quit/moved/were denied tenure/converted to being liberals over the past decade. The HERI data show that this is not a matter of retirement, and there is no evidence that adults (especially highly educated ones) are prone to changing their political views to such a degree. If 3/4 of New England conservatives quit, moved, or were denied tenure, that would be an extraordinary change.
I wonder if this is a sampling problem. I don’t know much about how the HERI study works, but I understand it is a cluster sample where people opt in, and then it’s weighted in some way. Perhaps something has gone wrong there.
I’d note, as well, that it is tricky to compare this study to studies of the general public. Mass surveys like the GSS typically allow respondents to answer “don’t know” or “haven’t thought much about it”, and large percentages or respondents choose those answers. The HERI questionnaire does not allow that.
So not to say there isn’t a trend here, but it seems difficult to me to quantify.
While I can buy the idea that the character of higher education varies in the country, with New England colleges differing from the rest, that’s a partial explanation. It might explain why in 1989-2001 NE professors were more liberal than the rest of the country. It offers no explanation at all for the change in NE professors beginning in 2001.
Changes in ideology could occur through individuals changing their minds, or a more liberal professor replacing a more conservative one. The steep change in 2001 implies a mass change of minds in NE–it’s so rapid. So what explanation can account for that?
I think, values on “Lib./Con. Ratio” axis, for exapmle, in Figure should start with “1”, not “0”.
Value “0,5” means that for every liberal there are two conservatives. Am I wrong?
Good data. Two points:
(1) There seems to be an acceleration of Lib:Con Ratio growth starting around 2004 or so. Any guesses about what might have changed around that time?
(2) It would be great to see a breakdown by discipline. My understanding is that lack of viewpoint diversity is relatively more severe in the humanities and social sciences (perhaps ex-economics) than in the physical sciences, engineering, and medicine. Separating these might be very helpful in understanding the trends and causes. For example, some people argue that conservative representation in academia is low because conservatives are anti-science. If that were the case, though, we would expect conservatives to be more relatively under-represented in the physical sciences and engineering than in the humanities and social sciences rather than the other way around. As another example, maybe there is a change in the relative number of faculty in different disciplines that could partially explain these trends.
Why do you insist on calling the hard leftists, liberal? As a classical liberal I am offended.