In the 18th century, the University of Basel faced a nepotism-driven crisis. Of its 80 professorships, about 50 were controlled by just 15 families. The university’s enrollment and reputation were in decline. In response, they implemented a new method for choosing appointments: a structured lottery system. There was a rigorous, standardized procedure to arrive at the final three candidates. Then, one of the three was chosen randomly.
Not everyone was happy with the system. One scholar, for example, was a finalist 10 times without being chosen, while others lucked into positions on their first try — including one at just 23 years old. But there were also marked benefits. Most obviously, the nepotistic chain was largely broken. There were also reports of decreased envy and jealousy, and greater satisfaction with the final decisions, even among those who did not win the job. And among those who did win, the knowledge that they had been chosen by lottery increased their humility and modesty.
Universities today face a well-known crisis of hiring: In short, there are far too many good applicants for the number of good positions available. And, in some ways similar to Basel’s situation, jobs tend to go to children of faculty members (albeit at different universities). As a 2021 pre-print study from Allison C. Morgan — then a Ph.D. student at the University of Colorado working with the data scientist Aaron Clauset — and co-authors shows, faculty members are up to 25 times more likely to have a parent with a Ph.D. than the general population — and that rate nearly doubles at prestigious universities. Other barriers to access loom large, especially in terms of which graduate program one attends. A 2015 study by Clauset and two co-authors, for example, found that in select disciplines, about three-quarters of tenure-track faculty members came from just a quarter of universities.
Some may argue that this is a simple meritocracy of demonstrated talent: People trained in any field from a young age (whether Tiger Woods or a future professor) will tend to develop their skills better, and the top graduate schools tend to produce the best faculty. If this is indeed the case, then our current situation is unjust, but it’s not necessarily hurting the production of scholarship. The Morgan et al. study of faculty members’ parents, however, finds that this is not a matter of training for academic research from a young age, but rather of having social and cultural capital. It is about knowledge, connections, and access to academic ways of being that are not obvious to outsiders. This helps constantly reproduce a lack of socioeconomic diversity among university faculty, which in turn may affect what kind of research they pursue.
The 2015 report’s authors (Aaron Clauset, Samuel Arbesman, and Daniel B. Larremore), meanwhile, detect a similarly deleterious effect on academic culture based on the effects of concentrated institutional power: “ideas originating in the high-prestige core, regardless of their merit, spread more easily throughout the discipline.” In 2017 Andrew Piper and Chad Wellmon further confirmed that the inequality in Ph.D.-granting institutions fed directly into inequality in dissemination. In a study of four leading journals in the humanities, they found that people with Ph.D.s from just the top 25 percent of institutions had published an astounding 89 percent of the articles.
Prestigious ideas crowding out other good ideas is an old phenomenon, and it is not confined to the humanities. Indeed, the foundational study of this was an essay in the history of science by Robert K. Merton, who termed it “the Matthew effect.” Like faith according to the Gospel of Matthew, recognition redounds on those who receive it early and rarely falls on those who, for whatever reason, previously missed out. Merton reported that even many of those deemed best notice the haphazardness of their continued success: “Nobel laureates … repeatedly observe that eminent scientists get disproportionately great credit for their contributions … while relatively unknown scientists tend to get disproportionately little credit for comparable contributions.”
Merton himself somewhat downplayed the findings, suggesting that what mattered was the ideas and not who contributed them. While there is perhaps some truth to that, it is at the very least callous to those who are never employed, and it can’t come to terms with later research documenting that women and people of color can especially suffer from lack of recognition. Pace Merton, these are problems of both justice and knowledge. What if, for example, potential breakthroughs in climate-saving technologies or women’s health or how to create more-equitable societies are simply not being advanced by already well-known authors?
Diversifying the faculty can help alleviate some of these problems for particular people, but it cannot eliminate the general problem of there being more talent than attention and positions. As prestigious academics are invited to give more talks and contribute to more publications, an inequality of attention and labor skews our academic resources. Some professors are always on a plane to a high-profile conference, dashing off a new essay, while others with equally exciting ideas publish into the void, if at all. Not only is this a poor distribution of resources, but it affects the quality of scholarship by risking a monoculture: A few ideas are perpetually enunciated, while other promising thoughts struggle to reach an audience.
This further erodes university life by creating troubling power dynamics. The point is not to do the best work; rather, it’s to attract attention and approval (and, eventually, job offers) from those who were deemed to do the best work in a previous generation. The powerful cliques that accrue are not always malignant, but nor do they foster a rich and diverse culture of university life. (And sometimes, like when such groups protect alleged sexual harassers, they are malignant.) The tendency of elites in any field is to be more invested in their own power and influence than quality, justice, or transformation. That is not something from which professors are immune.
Perhaps a lottery is too radical a fix, and we can solve our problems simply by reminding ourselves, when on search committees, to look beyond prestige or beyond reproducing our own academic interests and identities? Alas, research into both interview procedures and the social economics of hiring suggest otherwise. In their 2021 book Noise, Daniel Kahneman, Olivier Sibony, and Cass R. Sunstein reviewed the hundred-year-old, interdisciplinary literature on whether interviews lead to well-chosen hires. But before they could even report the minimal success they found, they were first forced to confront a “nontrivial” problem: There is no standard and agreed-upon definition of what a “well chosen” hire means. Not only are there many variables to consider, but the role that the person was meant to “fit” into in the first place is likely to shift.
Kahneman and co. lightly push this problem aside, relying on whether the people who make the hires are happy with their new employee’s performance. What they find is underwhelming: Only 56 to 61 percent of hires succeed, along these lines. Of course, those doing the hiring could themselves be among a previous generation’s unsuccessful hires, and thus possess dubious judgment — it’s hard to see how their self-reporting solves our “nontrivial” problem.
Nevertheless, Noise’s authors try to suggest ways to fix the limitations of hiring, based primarily on a case study of Google’s hiring practices. They find that implementing Google’s structured interviews improves the outcomes to 65 to 69 percent. While not nothing, this doesn’t help us define what, ultimately, makes a good hire. After all, humans are not like cargo. They do not have a clear list of contents, and what they do changes over time. Fit, especially for an academic job, is a snapshot of a generally young scholar who will likely change their research trajectory, and perhaps even personality, in entirely unforeseeable ways.
Even at the moment of selection, fit is incredibly hard to deduce from what we can reasonably learn in the job-interview process. Some people are more comfortable and self-assured simply because of their accrued cultural capital. Other people, regardless of their situation in life, simply don’t interview well (on either side of the table). Some introverts can lose steam over the course of a multiday process. Some have anxiety from the start. But some who get stressed in interviews are calm in other situations, and some of the best academics and kindest colleagues are anxiety-prone introverts.
Moreover, the academic-job interview process poorly represents what it’s actually like to be a colleague, teacher, or researcher. Never after they get a job will academics sit through a multiday, multi-interview process, and yet this is what they are judged on. And even the aspects of the job-interview process that are repeated, like teaching a sample class, are so artificial as to be of dubious value. Even excellent and passionate teachers will sometimes panic and perform poorly in a teaching demonstration. And, as the authors of Noise note, when this happens, our bias toward vivid impressions overwhelms our ability to consider general data and experience.
The difficulty of evaluating the success of a hire is also exacerbated, particularly in academe, because the Matthew effect tends to bring more attention to people simply by virtue of their having a prestigious Ph.D. or previous job. And because academic labor is so dependent on time and funding, especially in the sciences, getting the job can enable the very possibility of completing research. Finally, again, many applicants are qualified for any given academic job. That a person succeeds after hiring in no way demonstrates that they were objectively better than any other candidate.
As the social economist Fred Hirsch helped us understand in Social Limits to Growth (1976), no amount of economic expansion can ever fix the problem of the inherent limits of what he called “positional goods” — that is, things like attention, award, and recognition. There can only be one winner, no matter how many qualified applicants there are. As Hirsch put it, “what is possible for the single individual is not possible for all individuals — and would not be possible even if they all possessed equal talent.”
Fairness in jobs is about both material and positional goods. As Hirsch argued (and as is fairly obvious if you think about the material lives of adjuncts), these two are quite closely related. So long as a belief continues that some people deserve jobs and that others (even though they may be equally qualified) don’t, we will be stuck in this system of perpetual reward for some and deprivation for others.
There are at least three fundamental problems with our current hiring system. First, it creates concentrated power and academic monocultures. Second, convincing evidence about its success is lacking (and complete evidence is impossible). And third, it has no claim to being just or fair.
To genuinely shift the ways in which we think of about academic hiring and success, it may be time to revisit the University of Basel’s lottery approach. Research on college admissions and grant funding over the past few decades has shown that Basel’s success is not simply a fluke. While not undisputed, much of this research has found that organized lotteries such as the one once used in Basel can help undo some of the negative effects of our current system, including by weighting for diversity.
What the research generally suggests is that, when it comes to admissions and grant funding, pure lotteries of just anyone who applies tend to create many distortions. But there can be an organized system that culls down to however many excellent applicants there are (three is perhaps itself too arbitrary a number) and then selects by lottery among those remaining. The results from these related fields suggest that modified lotteries lead to less bias, stress, and hubris, and greater equity, mutual appreciation, and innovation.
Of course, there is something strange about using a lottery to choose who gets a job, especially given that the chosen hire may become one’s lifelong colleague. And some advocates of lotteries have suggested that lotteries should only apply when there is no clear best candidate. Job-hire lotteries could certainly experiment with different formats, such as allowing the lottery to be overridden by a unanimous committee vote. And they would of course need to determine fair criteria for arriving at the finalists who enter the lottery. Other issues would undoubtedly arise and require tinkering, but the experiments would be worth it. To try the lottery method is not to deny the talents of the lucky, but to affirm the talents of the luckless.
At the very least, lotteries for academic hires would create a future generation in which we would all explicitly know that the most prominent professors owe a significant degree of their success to luck, and that other, less-renowned people may be as talented if less lucky. Perhaps this would lead scholars to look beyond the most prestigious handful of departments in their field for interesting ideas — perhaps the power of prestigious departments would be spread out or even, eventually, dissolve. Networks of reproduction would lose some of their power as hand-picked candidates become rarer. And with less attention focused on individuals, collaborations, currently so rare in the humanities and some social sciences, might flourish. Finally, and perhaps most significantly, recognition of the excess of talent, and the randomness of hiring, might give greater cause to demands for redistribution and equitable pay across the profession.
This final possibility is the key point. A lottery on its own will not solve all our academic hiring woes. It will make explicit the unfairness that is now implicitly part of our system, but it will not alter the fundamental excess of good applicants. If we are serious about justice in our work, we must also be serious about justice in the conditions of our work. A lottery does not suffice, because a lottery does not directly deal with the problematic underlying conditions.
Solving these underlying conditions is, as we know all too well, easier said than done. And in the meantime, there is no justification for the stress and hubris and resentment engendered by our current system. Whatever limits lotteries may have, they also have important advantages of increasing humility and modesty and lessening stress and anxiety. In the absence of a just system for ensuring decency for all academics, an organized lottery of the qualified may be our fairest option.
