What this covers

Musa al-Gharbi, a sociologist at Columbia University, joins Sean Carroll to examine why academia has developed a pronounced leftward skew and what intellectual consequences follow. The conversation centers on a self-reinforcing mechanism: documented biases in peer review, hiring, and institutional review boards push conservatives toward think tanks rather than university careers, while those who remain often conceal their political views to survive professionally. Al-Gharbi argues this is not simply that conservatives are worse scholars—the ratio was much closer until the mid-1990s—but rather a compounding effect of individual rational choices that aggregate into systematic exclusion.

The discussion moves outward to epistemic damage. Adversarial Knowledge Production depends on participants holding diverse foundational axioms; when decision-makers share the same blind spots, institutions designed to correct bias amplify it instead. Al-Gharbi identifies a pattern wherein researchers treat their own views as an unexamined baseline, so that those who voted differently become subjects of pathology rather than subjects of inquiry. He points to a concrete example: academics who study marginalized communities often consult unrepresentative elites like Ta-Nehisi Coates rather than polling those populations directly, then design research that assumes conservative or religious views held by millions of Americans are themselves the problem to be solved. Al-Gharbi frames viewpoint diversity as an Instrumental vs Absolute Good—valuable for knowledge production and democratic engagement, not pursued for its own sake—and proposes that absent self-correction, external actors will intervene clumsily, threatening academic autonomy.

Sharpest takeaway

Al-Gharbi argues that the liberal skew of academia results from a self-reinforcing loop of discrimination and self-selection, and that viewpoint diversity matters as an instrumental good because homogeneous knowledge communities exacerbate rather than correct biases and lose touch with the people they study and serve.

  • The left/right skew in academia emerged sharply in the mid-1990s and is maintained by documented biases in peer review, hiring, and review boards combined with conservative self-selection into think tanks
  • Bias-correcting institutions (peer review, committees) only work when participants hold diverse axioms; shared blind spots get amplified instead
  • When researchers treat their own views as the baseline, they pathologize dissenters and ignore the actual preferences of the marginalized groups they claim to champion

The argument · threads6 threads · 27 claims
0.78

Institutional bias and anticipatory exit create self-reinforcing feedback that systematically excludes conservative scholars from academia.

6 pointscentrality 5/5
  • The academic skew is primarily a combination of discrimination and self-selection that reinforce each other: documented biases in peer review, review boards, and hiring along political and religious lines lead conservatives to anticipate that an academic career is out of reach, so they plan from the outset to enter think tanks instead, creating a gnarly self-reinforcing feedback loop.

    “it's primarily a combination of discrimination and self-selection, and those two things are related to each other”

  • Conservatives who stay in the academy increasingly work on non-controversial, non-political topics and keep their political leanings close to the vest, both to avoid having their work spiked in peer review and to avoid becoming social pariahs, which means academia cannot even reap the epistemic benefits of the ideological diversity that does exist because it is concealed.

    “mostly, a lot of the conservatives who remain in the Academy today try to work on non-political issues or try to keep their political leanings kind of close to the vest, and so one consequence of that, ironically, though, is that we don't even... We're not even able to reap the epistemic benefits, I guess, of the ideological diversity that is in the Academy, because what of it there is is kind of concealed”

  • When records are comparable, individuals reasonably prefer hiring an ideological ally over someone they disagree with; this is a normal impulse, but when virtually everyone in a left-skewed field makes the same calculation, certain perspectives get systematically locked out of the academy even though no single decision was illegitimate.

    “people tend to gravitate towards the person who's more like them, the person who's going to be an ally... but the problem is when everyone makes that same kind of calculation, and virtually all institutions of higher learning have a sort of decisive left skew... then you see... certain kinds of perspectives are sort of systematically locked out of the Academy”

  • Institutions like peer review and review boards are premised on adversarial clash between people with diverse frameworks to push toward more objective conclusions; but when all decision-makers share the same axioms and blind spots, these very processes amplify biases instead of correcting them.

    “in a world where the people making decisions all share the same axioms, the same blind spots, the same assumptions about the world, then you can have... Rather than correcting biases by means of things like peer review or committee decisions, you can actually exacerbate biases, as result of these processes that were supposed to correct the biases”

  • The current liberal-conservative ratio in academia is not because conservatives are intrinsically bad scholars; until the mid-1990s the ratio was much closer, and the sharp shift around that time corresponds to changing political makeup of university-hosting geographies and the rise of think tanks as alternative homes for conservative intellectuals.

    “until the mid-90s or so, the ratio of liberals to conservatives in the Academy was much different. So Sam Abrams, a political scientist at Sarah Lawrence College, has some good data on this.”

  • When the professional managerial class studies racialized inequality they reflexively blame millionaires and billionaires, but the 1% exercise their will through institutions—universities, government bureaucracies—run by the professional managerial class itself; failing to turn the critical lens on their own institutions produces a major blind spot in understanding how these phenomena are actually maintained.

    “millionaires and billionaires can't just sort of... make things happen as if by magic. They work through institutions, institutions run by the professional managerial class... it would actually be impossible for a lot of these conditions to be maintained without the active participation of people in the professional managerial class”

0.71

Lack of internal ideological diversity causes researchers to use their own position as unexamined baseline, treating disagreement as pathology rather than legitimate difference.

8 pointscentrality 4/5
  • Viewpoint diversity should be understood as an instrumental good valuable for increasing knowledge and engaging society, not as an absolute good pursued for its own sake; this implies some viewpoints legitimately fall outside its scope, though deciding which must be worked out by intellectual communities responsive to their needs.

    “viewpoint diversity I think should be understood best as an instrumental good rather than an absolute good, so it's valuable for the purpose of increasing our knowledge and understanding about the world... but it's not something that you pursue just for its own sake”

  • Economics enjoys higher prestige among policymakers and the social sciences than sociology or anthropology partly because it is relatively less left-skewed, so even left-leaning economists can engage amicably with the political right, whereas Republican policymakers assume a sociologist will only tell them they are wrong, stupid, and evil.

    “this is part of the reason why economics has this kind of prestige it has among the social sciences compared to other fields like anthropology and sociology, is because it's relatively less skewed”

  • Because many social researchers cannot imagine reasonably voting for Trump, much research on his supporters starts from the premise that something is wrong with them, designing studies that only offer pathological explanations (racism, sexism, authoritarianism, ignorance) while applying no symmetrical analysis to why people voted for Hillary Clinton—revealing that researchers use their own position as the unexamined baseline against which deviance is treated as a deficit.

    “there are whole studies that are literally designed like what best explains why someone would vote for Trump is that they're more racist or sexist or authoritarian or ignorant, and so all the options are bad... and there's just a lack... if someone designed a study that was like, why would someone vote for Hillary Clinton? Is it because they're communists or they hate America... We would immediately go, well, that's a prejudicial study design”

  • Since living the value of engaging opposing views is hard, the strategy is to lower the effort required by crowdsourcing tools and resources—such as syllabi from professors who have successfully incorporated conservative or religious perspectives—so practitioners don't have to reinvent the wheel, which otherwise disincentivizes people who agree in principle but don't know where to start.

    “we can crowdsource what works to lower the bar, to lower the amount of effort it takes for people to put it into practice... that in itself often disincentivizes people who agree on principle but don't know where to start, so they just kind of do what they've always done”

  • Some campus speakers like Milo Yiannopoulos have nothing substantive to contribute and exist purely to provoke reactions; when students give the desired reaction, it makes the students look like reactionaries, dragging everyone into a dynamic where both the provocateurs and the over-reacting opponents are acting in bad faith.

    “people like Milo Yiannopoulos, who doesn't have anything particular to say, his whole shtick was literally just trying to provoke people... when they got the reaction they were looking for, it made them seem like their opponents were reactionaries”

  • If academics fail to address the diversity problem themselves and with urgency, impatient external stakeholders—Fox News firing campaigns, state legislatures defunding sociology, federal executive orders on free speech—will increasingly intervene in ham-fisted ways that threaten academic freedom and autonomy, evidenced by the radical erosion of tenure with about 70% of new academic appointments now non-tenure-track.

    “if these problems continue to be ignored by us, they're not being ignored outside and eventually, probably sooner rather than later, these outside actors are going to be able... To threaten academic freedom, threaten autonomy of professors”

  • The widespread expert-class overconfidence that Hillary Clinton would win in 2016 was ironically a product of the very groupthink that Heterodox Academy was founded to counter, since dissenting predictions like al-Gharbi's were not the predominant view in the organization or among experts.

    “part of the reason people have that over-confidence that Hillary Clinton was going to win, ironically, it was because there was a certain kind of group think that had set in along the expert class. It's the very kind of problem... That Heterodox Academy was supposed to be responding to.”

  • Among those who work in higher education, the ratio is roughly 10:1 left to right in social research fields and about 5:1 left to right in the academy overall, because highly educated people tend to be more partisan and less centrist than the general population.

    “among those who work in higher ed, the ratio is something like 10:1 left to right in social research fields, and about 5:1 left to right in the Academy overall, in terms of political diversity”

0.63

Conservative intellectual tradition offers epistemic value by guarding against naive interventionism and helping prevent harmful unintended consequences.

2 pointscentrality 5/5
  • Conservative perspectives are more valuable in the academy than recognized because their orientation toward tradition guards against harm from naive interventionism—things that persist often persist for a reason, and changing them without understanding how they arose can make things far worse, even sliding into totalitarianism and famine, as history repeatedly shows.

    “It's precisely this orientation towards tradition and concern about things like harm caused by naive interventionism, the fact that the things that persist sometimes persist for a reason, and that if we don't understand sort of how they came about and why they persist, then when we change things that we can often make things worse”

  • Social orders need both the conservative tradition-orientation and the progressive critique-orientation in dialogue: without critique you get stagnation and orders that fail to respond to contemporary needs and collapse, while social orders persist partly because they solve real problems and work for many people, yet always fail some who are marginalized and whose perspective is the source of legitimate critique.

    “absent the sort of progressive impulse to critique and critiquing the prevailing order and looking for improvement, then you get stagnation and, similarly, those are social orders that also tend not to persist because they aren't responsive to contemporary needs... you really need these orientations to be in dialogue with one another”

0.49

Ideological segregation into separate institutional spaces prevents the genuine engagement and clash necessary for knowledge production and self-correction.

2 pointscentrality 4/5
  • The impulse to critique is usually other-oriented, aimed at ideological adversaries, and a lack of reflexivity—turning the critical lens on oneself and one's own allies and causes—is often why well-intentioned reform efforts go awry; this is precisely why diversity is essential, because a knowledge community of like-minded critics cannot self-correct without others to turn the critical lens back on them.

    “a lack of reflexivity of turning the critical lens towards oneself, towards the actors that one agrees with... is often the reason why these sort of well-intentioned things go awry... This is why diversity is important... because you need the lens to be turned against the people who are making those critiques”

  • The market 'solution' of conservatives moving to private think tanks fails because it produces two separate echo chambers rather than genuine diversity; the benefits of diversity arise only when opposing viewpoints actually engage, clash, and collaborate, so segregation yields conservatives talking among themselves and a parallel left echo chamber in the academy.

    “The benefit of diversity happens when you have these different viewpoints actually engaging with each other and having to answer to each other and clashing with each other and exposing one another's vices, but instead what you see is two separate echo chambers”

0.43

Researchers studying disadvantaged groups rely on unrepresentative elite interlocutors rather than their actual constituents, distorting understanding of real needs.

4 pointscentrality 3/5
  • African-Americans, Hispanics, immigrants, and lower-socioeconomic-status people skew more socially conservative and more religious on average than whites and upper-class people, so creating an academic environment hostile to socially conservative or religious views disproportionately alienates the very disadvantaged groups that progressives claim to champion.

    “blacks and Hispanics are more socially conservative on average than whites, so are many immigrant communities, they're also more religious on average than whites... if you create an environment that's hostile towards, for instance, socially conservative or religious views, that will disproportionately affect people of color”

  • Academics tend to consult unrepresentative elites like Ta-Nehisi Coates to learn what black or Hispanic people want, rather than polling those populations, but such elites' views are demonstrably out of step socioeconomically and ideologically with the typical member of the group, so relying on them leads researchers astray about the group's actual interests—evidenced by majority Hispanic rejection of the term Latinx and minority responses to alleged dog whistles and microaggressions.

    “people are going to consult Ta-Nehisi Coates about, well, what do black people want... rather than, like, let's poll African-Americans. And that's a problem because Coates is... his views on the world are demonstrably out of touch, his socioeconomic position is demonstrably out of step”

  • People overwhelmingly endorse free speech and viewpoint diversity in the abstract, but when presented with concrete cases that challenge their core beliefs they balk; this gap between stated attitudes and actual behavior—what Khan and Jerolmack call the attitudinal fallacy—means you must examine behaviors, not survey answers, to know what people actually believe.

    “in order to understand what people actually believe, it's not enough to ask them questions, but to look at behaviors. So in the abstract, when you ask people, do you support free speech, do you support diversity... people overwhelmingly say yes, but then when they're faced with difficult cases in their local context”

  • Since perceiving oneself as a target of racism correlates with negative emotional, mental, and even physical health effects, teaching people of color to be offended by microaggressions they do not currently find offensive would likely correlate with worse outcomes for them—meaning a well-intended intervention could harm the people it aims to help.

    “when people perceive themselves to have been the target of racism or discrimination, that that corresponds to a number of negative emotional effects, mental health effects, even just physical health effects. So sensitizing people more to these things that they don't currently find offensive... would probably correlate with worse and mental health outcomes”

0.35

Academics must engage widely held and policy-relevant views and communicate research accessibly to the public that funds taxpayer institutions.

2 pointscentrality 3/5
  • Views held by millions of Americans or by policymakers must be engaged with even if you disagree, because refusing to engage undermines your ability to reach those people and may even undermine your ability to understand the phenomenon; niche views held by few can more safely be excluded.

    “if it's a widely held view, that a lot of people in society hold, that a lot of policy makers hold, then it's a kind of view that you're going to have to engage with, and that if you don't engage with, it's actually probably going to be counterproductive”

  • Academics have done a poor job explaining why their work matters and connecting it to the needs of ordinary people, which is problematic especially at taxpayer-funded institutions where the public reasonably expects some benefit from their investment; intellectuals should therefore engage the public more, communicate research accessibly, and connect scholarship to practical concerns.

    “academics have been kind of bad at explaining things like why our work matters, connecting the work we do to the needs and priorities of ordinary people, and this matters, especially, I think for scholars at institutions that are paid for largely through taxpayer funds”