Should we measure and scale anti-Semitism? Do we really need to know whether Harvard is the most anti-Semitic college in the United States or whether Egypt ranks number one in the Middle East—or is it Jordan? Anyone can see that those places are steeped in Jew-hatred, regardless of who “wins” the competition. The Netherlands records one of the lowest levels of anti-Semitism globally—according to the ADL—yet there was violent anti-Semitic unrest in Amsterdam in November 2024. Some commentators even referred to it as a “pogrom.” Meanwhile, South Korea scores five times higher than the Netherlands, but we don’t see outbursts of violence against Jews in Seoul. It was at Cornell that Associate Professor Russell Rickford yelled in celebration across campus that October 7 was “exhilarating” and “energizing,” while one of its students threatened that he was going to “slit the throat” of Jewish men on campus, rape its Jewish women and throw them off a cliff, behead Jewish babies, and “shoot up” Cornell’s kosher dining hall. Does it matter that Cornell previously scored lower in campus anti-Semitism according to the AMCHA Initiative than Harvard, at number one?
Perhaps the right question to start with is this: are we even able to do it? There is a general problem when it comes to psychological measures, questionnaires, and scales called the “toothbrush problem”—scientists don’t want to use somebody else’s. This, of course, makes it difficult to compare studies and easy for scholars to proclaim the discovery of knowledge and authority where there is none. Unsurprisingly, Irving Sarnoff’s Jewish anti-Semitism scale was largely ignored by other scholars when it came out in 1951. Likewise, the AzAs (Antizionist Anti-Semitism) scale did not find much resonance since its publication in 2019, despite the authors suggesting it to be of “general use.” Nor did their “Judeophobic Antisemitism” scale.
So, three years later, the same research team of Daniel Allington, David Hirsh, and Louise Katz developed the “Generalized Antisemitism (GeAs) Scale” and claimed to reliably measure anti-Semitism as a single underlying psychological trait with robust validity and reliability across different demographic groups. And this time, marketing kicked in.
The famous phrase attributed to Benjamin Disraeli comes to mind: “There are three kinds of lies: lies, damned lies, and statistics.” The editor-in-chief of the Journal of Contemporary Antisemitism, however, wrote in her introduction to the study, “that the scale is both reliable and valid” and that “the authors present a thorough statistical analysis” while they would “provide a powerful research instrument for the further study of antisemitism.” Allington himself is deputy editor of the journal—by its own merit, “the leading scholarly publication in the field.” Hirsh serves as its editorial consultant.
[RELATED: When Anti-Semitism Is Claimed, Facts Must Follow]
Published open access in another journal, the authors then attested that they had made a “decisive contribution to the scholarship of antisemitism” with their scale. The scale has since been reported on in the Times, was referenced in the House of Commons, and incorporated into the research efforts of the Campaign Against Antisemitism. And at a 2025 conference at the London Centre for the Study of Contemporary Antisemitism—founded by Hirsh—Allington proclaimed about their own work: “this has to be considered one of the most consistent findings of modern social science.”
If so, what justifies the team to share this knowledge? Do they expect bad actors in Iran or Qatar not to use it so they can manipulate Western societies and increase anti-Semitism accordingly? Or do they believe that sharing the results does not matter since bad actors know how to manipulate anyway—which would point to the common-sense nature of their finding?
The truth is, the GeAs scale does not have to be considered anything because it can’t be used effectively: It simply does not reliably measure anti-Semitism as a single underlying psychological trait with robust validity and reliability across different demographic groups. Afterall, this would oversimplify complex social and cultural attitudes. The scale was conducted exclusively in the United Kingdom. This, obviously, does not capture different cultural and religious factors that shape anti-Semitism across the globe. In fact, it would be foolish to think about the phenomenon as taking on a single form independent of it happening in places like Poland or Iran, in the mind of a Muslim socialist from Uganda or a republican podcaster in his cabin in Maine, all based on UK data. There are similarities, sure, between the liars’ smile and the hysterical cackle. However, that doesn’t mean social scientists can detect a single anti-Semitic gene or virus. Social scientists should care that the 2025 democratic candidate for mayor of NYC wants to “globalize the intifada” and thinks that violence is an “artificial construction,” not if he scores higher or lower on anti-Semitism than the podcaster who spreads lies after lies about Israel for clicks.
The GeAs scale does not even produce general knowledge on the UK itself since the sample suffered from underrepresentation of ethnic minorities and male respondents—particularly younger men. What is more to it is that the GeAs scale was designed with the International Holocaust Remembrance Alliance (IHRA) definition of anti-Semitism in mind. But then the authors argue that their findings support the IHRA definition. That is circular reasoning. Since the scale was designed with that definition in mind, it is no surprise that it aligns with it. As much as I agree with the IHRA definition, academia would do better if scientists stopped presenting design choices as validation.
This is a common problem with the work of Allington and Hirsh. In their article “Statistical Associations between Antisemitism and Higher Education”—also published in their journal—the authors overlook methodological flaws, including reliance on small subgroup sample sizes (e.g., N = 65) and an overemphasis on non-significant findings. The result of their study then comes as a surprise, given what is currently happening on many college campuses: People with college degrees, so the team, tend to be less anti-Semitic than those without degrees. The authors also offer the guess—based on exploratory analysis—that psychology majors may be among the least anti-Semitic. This seems unlikely and not to be the view of many Jews themselves. Last year, the Association of Jewish Psychologists (AJP) found, for instance, that members “expressed significant concerns about rising anti-Semitism within the profession, feeling marginalized and silenced, and the possibility of diminution of Jews in the field.” I have not heard the same about mechanics or firefighters.
What Allington and his team did with the GeAs scale was to use statistical tools and then rely on the assumption that statistical patterns reflect stable, universal truths—when in reality, such patters depend on measurement choices, context, and sampling. Their scale consequently misinterprets probability by claiming to reveal inherent human traits. Just because the model might fit their UK data, it does not represent a universal construct. Ultimately, this shows a lack of understanding that statistical models do not describe reality; they approximate it. And while the study mistakes statistical coherence for an objective psychological truth, it ignores the complexity of anti-Semitism across the world.
[RELATED: The Anti-Semites and Their Betrayal of Conservatism]
Treating the phenomenon as such means that it becomes a detectable condition that is inherent in individuals rather than something that is shaped by social, historical, religious, cultural, or political influences. While the research team overstepped the limits of what social science can achieve, they elevated themselves as scholars by claiming to have discovered a universal diagnostic tool for anti-Semitism. This medicalization of the phenomenon inevitably allows the authors to position themselves as unquestionable authorities in the field and to gain control over how anti-Semitism is defined and studied anywhere it occurs. Creating a widely adopted measurement tool is always a powerful way to gain authority and influence in a field. It’s one of those rare moments in science. And it needs to be justified. Due to the numerous flaws of the GeAs scale, I do not see any justification for it. After all, it is not a scientific breakthrough. In fact, it is not even a scientific finding.
Our academic funding system, unfortunately, tends to create these caricatures of science and false authority. The idea to develop a scale for anti-Semitism seems, at first glance, like something perfectly scientific and result-driven. And as such, it writes up well into a grant proposal: “we identify what anti-Semitism is; and we identify that one thing that is causing it; and by isolating it we can illuminate anti-Semitism, or at least it’s a world-changing discovery we are about to make. In fact, not just any discovery, but the most consistent in all modern social science.” And it’s all measured and put into numbers and graphs that the applicants are absolutely certain about. And most importantly, it can be well organized into something clear, simple, and straightforward. But the problem is that anti-Semitism studies is not virology or epidemiology. It is not straightforward and simple. It involves history and culture, religion and ideology, economics, law, and cognition. But that’s a more complicated story. But then, who gets the funding for what exact work one is proposing to do in order to reach what result? It’s that perverse incentive of funding in favour of simple, reductivist, easily counted research that looks exactly like what scientists do in the lab. And that makes it unlikely that the GeAs is going to be the last failed attempt to scale anti-Semitism. Or perhaps the last attempt by scientists to use a flawed measurement tool to gain control over a societal issue.
I understand that putting a complex phenomenon like anti-Semitism into a single scale and pointing to that one psychological trait that is causing it must feel cathartic to scientists or anyone else, for that matter. The GeAs, however, is a form of epistemic overreach, where social scientists claim to have created a definitive measurement tool for something that is inherently variable. It turns a contested social issue into something that looks like an objective, scientific fact, as well as a definitive psychological reality, which it is not. Once introduced into the literature and marketed, the GeAs scale distorts future research by reinforcing its own flawed premises. One should not use this “toothbrush” because it’s not one’s own, but because it’s broken.
Visit our Minding the Science column for in-depth analysis on topics ranging from wokeism in STEM, scientific ethics, and research funding to climate science, scientific organizations, and much more.
Image by AndriiKoval on Adobe Stock; Asset ID#: 295683723
