During Wednesday night’s Republican presidential debate, former UN Ambassador Nikki Haley stated that “we really do need to ban TikTok once and for all.” She justified her proposal with an alarming claim: “for every 30 minutes that someone watches TikTok every day, they become 17% more antisemitic.” This can’t possibly be true, can it?
It would seem not. Haley was referring to a recent survey analysed by Anthony Goldbloom and colleagues. In the survey, 1,323 Americans aged 18–29 were asked about their use of social media, and their views of Jews and Israel. The researchers created a measure of antisemitic/anti-Israel views, and then compared individuals who use TikTok at least 30 minutes per day to those who don’t use the platform.
They found that those who use the platform at least 30 minutes per day were 17% more likely to hold antisemitic/anti-Israel views. In other words, they did not find that people become 17% more antisemitic for every 30 minutes they spend on TikTok — which would imply that they become twice as antisemitic after just two and a half hours.
In addition, the researchers found that individuals who use Instagram and X at least 30 minutes per day were only 6% and 2% more likely to hold antisemitic/anti-Israel views than non-users. They took this as evidence that TikTok is particularly conducive to the promotion of such views. The platform, Goldbloom writes, is “creating a dangerous environment for Jews”.
While Haley’s claim was clearly based on a misunderstanding, the findings themselves also raise suspicions. Taken literally, a 17% difference between two groups on a measure of antisemitic/anti-Israel views is extremely small. It means that if 10% of non-users hold such views, the figure for users is 11.7%. What Goldbloom presumably meant to say is “17 percentage points”, which would equate to a figure of 27% for users in this example.
I obtained the raw survey data and analysed it myself. The survey contains six items for antisemitic views, each measured on a seven-point scale from “strongly disagree” to “strongly agree” (e.g. “Jewish people talk about the Holocaust just to further their political agenda”). I created an index from these six items by averaging them (after reverse-coding three). Since Haley referred specifically to antisemitism, I ignored the items for anti-Israel views.
The overall level of antisemitism in the sample was low. Only 11.5% of respondents scored higher than 4 on my index — the midpoint of the scale for measuring each item. And among TikTok users, the figure was only slightly higher at 13.2%. In fact, TikTok users’ average score was 3 — indicating disagreement with antisemitism rather than neutrality or agreement.
Yet when I standardised the index, I found that TikTok users indeed scored significantly higher than non-users (almost half a standard deviation higher). And this difference became only slightly smaller when controlling for age, sex, race, education and religion. I then ran the same multivariate models for Instagram and X, respectively. The difference for X was about half as big as the one for TikTok, while the difference for Instagram was small and non-significant.
Of course, my analysis does not establish causality. It’s possible that factors other than the ones mentioned above explain why TikTok users (and, to a lesser extent, X users) score higher on antisemitism than non-users.
Haley was clearly mistaken. But there may be some truth to the claim that TikTok is making young Americans less likely to disagree with antisemitism. At this point, further research is needed.