Did Antisemitism in Public Opinion Rise in the Wake of the Israel–Hamas War?

Abstract

Israel’s military response in Gaza to Hamas’s terrorist attack and hostage taking of 7 October 2023 has led to fears of growing antisemitism. Indications of heightened antisemitism include massive spikes in antisemitic incidents and hate crimes around the world and the US, demonstrations and campus unrest, and antisemitic memes on the internet and social media platforms. Questions remain, however, whether public opinion has become increasingly hostile to Jews. The ADL Global 100 reports nearly a doubling in antisemitic sentiment from 2014 to 2024. This paper explores trends in antisemitism using country-level ADL Global 100 data. Results show some countries exhibiting large increases in antisemitism, but not all. For the 2023–2024 comparisons, European nations display relatively stable antisemitic distributions, but Russia shows a large increase. The study also uses American National Election study (ANES) data, both pooled from 1964–2024 and the 2020–2024 panel. The ANES data show a slight drop in warmth to Jews using the feeling thermometer. Demographics do not account for the slight drop, but analysis of the panel data suggests that attitudes toward Israel may account for the decline in warmth.

1. Introduction

The Hamas terrorist attack and hostage taking of 7 October 2023 was a watershed event of potentially massive proportions. Over 1200 Israelis were killed and 251 were taken hostage. Confidence in the Israel Defense Force (IDF) plummeted because of the attack, from 70–75% before1 to nearly one-half of Israelis losing confidence by early 2024.2 Israel responded with airstrikes against Gaza from 8 to 26 October 2023, followed by an invasion on 27 October. Israeli forces have remained on the ground in Gaza ever since, with an estimated 8500 civilians killed in Gaza by the end of October 2023, mounting to tens of thousands by March 2025.3

This paper focuses on the potential antisemitic consequences of the Israel–Gaza war. The large scale of civilian casualties has led to massive public outcry around the world. Some attribute the rise in antisemitic incidents to the war in Gaza. The Anti-Defamation League (ADL), which has tracked antisemitic incidents since 1979, reports a dramatic rise in recent years. Its 2024 audit found that incidents increased from 942 in 2015 to 9354 in 2024, with the most significant surge occurring between 2022 and 2023 (from 3698 to 8873). A particularly sharp spike was recorded between September and October 2023, with incidents rising from 513 to 1813 in a single month, suggesting that the war in Gaza may have sparked the surge. Similarly, FBI data show that anti-Jewish hate crimes rose from 1122 in 2022 to 1832 in 2023—a 63% increase. Anti-Jewish hate crimes have consistently been the most common religion-based hate crimes in the US, comprising 60% of such incidents from 2019 to 2023 and nearly 68% in 2023 alone. The Global Terrorism Index also reports a rise in antisemitic incidents, saying that Europe has witnessed a “wave of antisemitism.”4 In the UK, the Community Security Trust (CST) reported that of 4103 antisemitic incidents there in 2023, 2699 (66%) occurred on or after 7 October 2023.5

Campus protests and unrest broke out in the US and around the world, especially in Europe. The Crowd Counting Consortium at the Ash Center, Harvard University estimates that by late April 2024, there had been 3700 protest days at over 500 US campuses, including 130 with pro-Palestinian encampments.6 Some argue these protesters were antisemitic because of the chants “From the river to the sea” and “Globalize the intifada.” The chants may be antisemitic because they demonize Israel, delegitimize Israel’s right to exist, and by targeting non-Israeli Jews, especially students, use the longtime antisemitic stereotype of Jews’ dual loyalty to the US (or other country) and Israel. Others argue that the civilian carnage in Gaza is deplorable. Thus, it is legitimate to criticize Israel’s policies, and criticism is not necessarily antisemitic.7 In the US, the 2024 American National Election study (ANES) found little support for those protesting the war in Gaza, with 24% approving over 36% neither approving nor disapproving, while about 39% disapproved.8

Another potential indicator of rising antisemitism is internet search activity. We must be cautious in using internet search terms for tracking sentiment, because search terms may be ambiguous in meaning. Additionally, formally search terms measure activity, not sentiment (Hölzl et al. 2025).

Figure 1 plots the Glimpse estimate from Google Trends weekly data from 19 July 2020 through 20 July 2025 for the number of searches for the phrases “From the river to the sea” and “Kill Jews.” These two search terms were employed because of the ambiguity and controversy concerning “From the river to the sea,” while “Kill Jews” is less ambiguous in its animosity toward Jews (Hölzl et al. 2025).

Figure 1. Absolute Google search volume, 2020–2025. Note: Search volume for “Kill Jews” and “From the river to the sea,” weekly from 19 July 2020, using Glimpse to extract volume.

The figure shows a sharp spike shortly after 7 October 2023 Hamas attack for both terms, with “Kill Jews” spiking a week or so before “From the river to the sea,” the latter apparently surging once the Israeli Defense Forces began military action in Gaza during the week of 5 November 2023.9

Jewish Americans increasingly feel unsafe. According to the American Jewish Committee (AJC), the percentage of American Jews who reported feeling less secure than the previous year rose from 42% in 2019 to 73% in 2024. The 2025 Jewish Landscape Report, a survey of Jews in 10 nations, found 76% seeing rising antisemitism.10 Many non-Jews concur that antisemitism has grown worse since Gaza. A November 2023 Associated Press poll found that 78% of Americans expressed some concern that the war between Israel and Hamas had led to increasing prejudice against Jews in the US.11 A Marquette University poll conducted during March 2025 found over 60% of Americans saying that antisemitism was a very big or moderately big problem.12

The assumption of this study was that a nation cannot truly be antisemitic unless public opinion writ broadly also holds antisemitic beliefs. Politicians, especially in democracies, are unlikely to be able to enact and implement antisemitic policies without public consent and that such consent will come more easily if the public is antisemitic. Citing rising incidents does not tell us much, if anything, about sentiment. Enstad (2023), for instance, found a negative correlation at the country level between antisemitic incidents and public opinion.

This paper asks whether public opinion around the world and in the US has grown more antisemitic due to the events in the Middle East since 7 October. The “new antisemitism” theory is used to causally connect attitudes toward Israel to antisemitism (Abady 2019).

Public opinion seems to have turned against Israel since 7 October. Pew reports a decline in favorability toward Israel by 11% across 24 nations between March 2022 and March 2025, with a median 2025 unfavorable score of 62 versus 29 favorable.13 US public opinion has also shifted, from 54–31 sympathy for Israel versus the Palestinians in early 2024 to 46–33 by early 2025, According to a Gallup Poll, American public opinion has not been highly supportive of Israel’s military action, from 50–45 (approve vs. disapprove) in November 2023 to 42–48 by September 2024.14

A primary claim of the new antisemitism theory is that criticism of Israel will lead to antisemitic beliefs. The rise in criticism and unfavorable opinion toward Israel thus should result in heightened levels of antisemitism in public opinion. Data from the ADL Global 100 Index suggests so, with 26% of the world’s population reporting antisemitic opinions in 2014, but 46% in 2024.15 Still, many events and processes besides the Gaza war occurred in the decade between 2014 and 2024 that could have led to greater antisemitism worldwide.

This article uses several data sources, the ADL Global 100 polls and the ANES data from 1964–2024, to address the question of whether there is a marked increase in antisemitic opinion after 7 October. Unlike the global population-weighted average ADL figures of the previous paragraph, this analysis focuses on country-level changes. The large rise in the ADL antisemitism estimate could be due to populous nations like China and India. Second, the ADL also surveyed citizens in 2023, just before the war began, which we can compare to 2024. This shorter time frame may be better for isolating the effects of the Gaza war than the longer 2014–2024 decade. But this analysis suffers because the ADL surveyed in only 11 nations, all in Europe, which is unrepresentative of nations worldwide.

This paper also uses the 1964–2024 ANES feeling thermometer, building on Cohen’s (2018) analysis of predictors of attitudes toward Jews. Additionally, it utilizes the ANES panel of 2020–2024 to investigate changes in attitudes of the same individuals over time. Fortunately, the 2020–2024 panel included several questions about Israel, which allows a more direct test of the new antisemitism thesis.

This paper is structured as follows. The next section reviews the new antisemitism thesis and how it relates to the events since 7 October. Then, it presents a country-level analysis of the ADL Global 100 comparing 2014 with 2024, as well as a more detailed analysis of the subset of nations the ADL Global 100 surveyed in 2023 and 2024. Section 4 analyzes the pooled 1964–2024 ANES, followed by analysis of the 2020–2024 ANES panel.

2. From Criticism of Israel to Antisemitism: The New Antisemitism Perspective

Fein (1987, p. 67) defines antisemitism as “a persisting latent structure of hostile beliefs toward Jews as a collectivity manifested in individuals as attitudes … in culture as myth, ideology, folklore, and imagery, and in actions—social or legal discrimination, political mobilization against the Jews, and collective or state violence—which results in and/or is designed to distance, displace, or destroy Jews as Jews.” This study focuses on one manifestation, attitudes towards Jews as recorded in public opinion surveys (Enstad 2023).

Public Opinion Research on the “New Antisemitism”

The “new antisemitism” perspective makes several claims. First, the perspective argues that some criticisms of Israel are inherently antisemitic. These include denying Israel’s right to exist, equating Israel with Nazism, and holding Israel to standards not applied to other nations (Sharansky 2004; Chesler 2015; Wistrich 2016). Second, the “new antisemitism” perspective maintains that criticism of Israel becomes antisemitic when no distinction is made between the Israeli government’s policies and Jews around the world, when non-Israeli Jews are blamed or held culpable for Israeli’s policies, especially regarding Palestinians in the West Bank and Gaza. One consequence of the “new antisemitism” is the increase in antisemitic incidents and hate crimes (Alexander and Adams 2023; Waxman et al. 2021; Feinberg 2020).

There are critics of the “new antisemitism” perspective. The primary criticism is that the perspective combines “antisemitic criticism of Israel and other ways of criticizing Israel” (Kempf 2012, p. 518), that is, people may criticize Israel without being antisemitic (Bindman 2019; Klug 2004, 2013). Further complicating matters is that people may be antisemitic, yet supportive of Israel, which often is the case among some Christian evangelicals and religious conservatives (Inbari et al. 2021).

Finally, there is no agreed-upon set of questions that measure criticism of Israel, especially a standard to distinguish legitimate criticism of Israel from antisemitic-inspired criticisms. Kaufman et al. (2020) argue that antisemitism scales need updating, especially to accommodate the “new antisemitism.” Allington et al. (2022) criticize existing measures, offer a revised scale that updates measures of classic antisemitism, and incorporate items measuring the “new antisemitism,” especially criticism of Israel (Allington and Hirsh 2019).

There is some modest support for the new antisemitism perspective. Numerous studies find a correlation between criticism of Israel and anti-Jewish attitudes (Allington and Hirsh 2019; Allington et al. 2022; Baum and Nakazawa 2007; Beattie 2017; Brym 2024; Cohen et al. 2009; Hersh and Royden 2023a; Kaplan and Small 2006; Kempf 2012; Jaspal 2016; Shenhav-Goldberg and Kopstein 2020; Staetsky 2020). Brym (2019) cautions that some studies use small and unrepresentative samples, and others note that critics of Israel do not always hold anti-Jewish attitudes (Beattie 2017; Staetsky 2020; Shenhav-Goldberg and Kopstein 2020).

Additionally, debate exists over causality between criticism of Israel and negativity toward Jews. Few studies test for causality (Kaplan and Small 2006; Baum and Nakazawa 2007; Beattie 2017; Shenhav-Goldberg and Kopstein 2020; Staetsky 2020). Those that do find that criticism of Israel affects antisemitism (Cohen et al. 2009, 2011; Hersh and Royden 2023a). However, some analysts question the direction of causality (Lelkes et al. 2016; Binstok et al. 2024), arguing the relationship may be reciprocal or spurious (Brym 2019).

The massive growth in criticism of Israel after the events of 7 October provides a best-case context for testing hypotheses from the new antisemitism perspective.

3. ADL Global 100 Data: Trends in Antisemitism Around the Globe

The Antidefamation League (ADL) has been tracking antisemitic opinion in the US and elsewhere since the mid-1960s. In 2014, the ADL surveyed over 50,000 individuals in 102 nations, the largest survey of attitudes toward Jews until then, calling the project the ADL Global 100.16 It repeated that with a global survey in 2024. Below, these two surveys are used to track changes in antisemitism around the globe from 2014 to 2024, something rarely done (Cohen 2018; Enstad 2023). Also, where most surveys ask respondents one or a few questions about their attitudes toward Jews, the ADL poses 11 questions, asking respondents if the statement is “probably true,” “probably false,” or they have no opinion. Table 1 displays the 11 statements.

Table 1. ADL Global 100 questions on attitudes toward Jews.

Number	Statement	Short Name
1	Jews are more loyal to Israel than to [country name]	Loyal
2	Jews are responsible for most of the world’s wars	Responsible for Wars
3	Jews don’t care what happens to anyone but their own kind	Don’t Care
4	Jews have too much control over global affairs	Global Power
5	Jews have too much control over the global media	Global Media
6	Jews have too much control over the United States government	Control US
7	Jews have too much power in international financial markets	Control International
8	Jews have too much power in the business world	Control Business
9	Jews still talk too much about what happened to them in the Holocaust	Holocaust
10	Jews think they are better than other people	Think Better
11	People hate Jews because of the way Jews behave	Behavior

Source: ADL Global 100, https://global100.adl.org/about, accessed on 7 June 2025.

From these questions, the ADL calculates an antisemitism score for each respondent, defining a person as antisemitic if the respondent answers “probably true” to six or more of the statements. Based on this definition, the ADL estimated that 26% of the world’s population was antisemitic in 2014 and 46% in 2024.17 A national antisemitism score can be calculated similarly for each nation.

3.1. Critiquing the ADL Antisemitism Scale

The ADL antisemitism scores can be critiqued. First, all the statements point in the same direction (“probably true” is the antisemitic response), which can lead to acquiescence bias and reduced respondents’ cognitive engagement, among other issues (Yan et al. 2018). Second, using six “probably true” statements to classify a person as antisemitic is arbitrary. It assumes that each item weighs the same in deciding if someone is antisemitic. Additionally, the binary classification of antisemite/non-antisemite discards variance, that is, it views someone who answers “probably true” 12 times is equally antisemitic as the respondent who answers “probably true” six times. Staetsky (2021) argues that antisemitism across individuals (or other units) should be regarded as a scale that ranges from greater to lesser antisemitism, what he calls the “elastic” view.

Furthermore, the 11 statements have not been psychometrically validated (Allington et al. 2022; Duckitt and Sibley 2007; Fiske et al. 2018). Greene and Kingsbury (2017) argue for a neutral response option. In their study “between a third and half of the subjects switch to a neutral answer, mostly from more benevolent answers.” (p. 112). Thus, the Global ADL 100 may underestimate antisemitism. Finally, there are cross-cultural and language concerns. All the statements are based on Western-centric antisemitism constructs, which may not be relevant in non-Western societies. The ADL Global 100 used over 80 languages, raising concerns about semantic equivalence (Vandenberg and Lance 2000).

Despite these concerns, reasons exist for analyzing the Global ADL 100 data, including its geographic scope, the use of identical items across countries, the potential for longitudinal analysis, and the fact that the 2014 data have been analyzed at both the country and individual level, producing findings that make substantive sense (Tausch 2014; Berggren and Nilsson 2021; Nyhan et al. 2024). These data may be imprecise at measuring antisemitism levels but are still useful for analytic purposes if imprecision is kept in mind.

3.2. Global Antisemitism from 2014 to 2024

Before turning to country-level trends in antisemitism, inspection of the 2014 data reveals three countries with unrealistically low antisemitism scores: Laos at 0%, Philippines at 3%, and Vietnam at 6%. Such scores are comparable to those for northern Europe, but other countries geographically closer to these three have much higher antisemitism scores, like Indonesia (48%), Japan (23%), Malaysia (61%), Singapore (16%), and South Korea (53%). Even Australia and New Zealand have higher scores, both at 14%, and hence the analysis deletes these three nations.

Figure 2 plots changes in national antisemitism, with “Decrease” indicating the difference in national antisemitism from 2014 to 2024 dropping 6% or more, “Stable” is when the change is ±5%, and “Increase” is when the change increases 6% or more. Note that data do not exist for large swaths of Africa, some of Asia, or Southeast Asia.

Figure 2. Change in antisemitism, 2014–2024, ADL Global 100. Note: Decrease is 6% or less antisemitic, Stable is ±5% antisemitic, and Increase is 6% or more antisemitic. Grayed-out areas indicate no data.

Many more countries show increases in antisemitism than declines, especially in Latin America, Asia, sub-Saharan Africa, Oceania, and parts of the Middle East. Declining antisemitism levels are noticeable among Western European nations, Canada, and surprisingly, some of northern Africa and the Middle Eastern nations. Large numbers of European nations plus the US and a handful of Latin American nations display stable antisemitic percentages. Overall, more nations show increases in antisemitism from 2014 to 2024, 48 of the 96 (50%) nations, while 34 (35%) have stable opinions and only 14 (15%) show lessening levels of antisemitism.

3.3. Comparing Antisemitism, 2023–2024

There are several limitations to comparing antisemitic opinion across the decade. One, over such a long period, many events and processes may account for any differences. The time gap is too long to isolate the effects of the hostage taking and Israel–Hamas war. Two, there is no apparent regional pattern for changes in aggregate antisemitic opinion. Three, the aggregate data cannot reveal which types of people became more (or less) antisemitic in 2024 compared to 2014.

The ADL Global 100 surveys are not limited to just the massive global studies of 2014 and 2024. Between those years, the ADL Global 100 surveyed subsets of nations. Most instructive for our purposes is that 11 countries were surveyed in 2023 and 2024. The 2023 surveys were conducted before the hostage taking on 7 October. This shorter time frame of a year or so, while it cannot definitively isolate the effects of 7 October as the only causal factor, still provides useful information on the possible effects of those events on antisemitic opinion. Table 2 lists the countries, their 2023 and 2024 scores, and the difference between the 2024 and 2024 scores.

Table 2. ADL Global 100 Antisemitism scores for 2023 and 2024.

County	2023	2024	2023–2024
Germany	12	9	−3
France	15	13	−2
Hungary	37	36	−1
Spain	26	26	0
Netherlands	6	8	2
United Kingdom	10	12	2
Poland	35	40	5
Ukraine	29	36	7
Belgium	22	30	8
Russia	26	62	36

Source: ADL Global 100, https://www.adl.org, accessed on 7 June 2025.

First, all nine nations are European, but include both Western and Eastern Europe. These ten do not constitute a representative sample of countries worldwide. Still, there is variance in antisemitism levels in 2023, with six countries having scores above 20%, a relatively antisemitic level. Only three countries in 2023 had scores of 12% or less, which is generally regarded as not highly antisemitic. Three nations, Germany, France, and Hungary, show small negative changes, becoming slightly less antisemitic. Four countries show small increases in antisemitism of 5% or less (Spain, Netherlands, UK, Poland). Two, Ukraine and Belgium, show moderate increases in antisemitism of 7–8%, while one, Russia, shows a massive 36% increase in antisemitism.

The overall correlation of the 2023 and 2024 scores is Pearson’s r = 0.76 (p = 0.01) but excluding Russia the r is 0.96 (p < 0.001), indicating great stability. A regression analysis reinforces the stability perspective for the nine nations, with a constant very close to zero (−0.10) and a regression coefficient of 1.09 (SE = 0.13), suggesting almost a perfect one-to-one correspondence between scores for 2023 and 2024. At least in these nine nations, there is no suggestion that the events of 2023–2024 led to increasing antisemitism in public opinion.

Russia, in contrast, displays a massive increase in antisemitism. Here we can only speculate as to the reason(s). The Putin regime has been highly antagonistic to Israel, viewing it as a client state of the US. The Russian president embarked on an increasingly antisemitic campaign, in part because Ukraine’s leader, Volodymyr Zelenskyy, is Jewish, and Russia’s war was not progressing as Putin hoped, with stiff resistance from Ukraine. Russia’s public antisemitic campaign appears to have escalated with 7 October, as Russia publicly sided with Hamas.18 State-controlled Russian media broadcasting news and government propaganda increasingly leaned against Israel.19

The ADL estimate of the rise in antisemitism is based on population-weighted averages, which give greater weight to very populous countries like China and India, both of which show increases in antisemitism from 2014 to 2024 (China from 20% to 58%; India from 20% to 39%). A country-level comparison shows a more modest rise from 38% to 44% based on national averages of the same nations surveyed both times, with about one-half of nations exhibiting a rise of 6% or more. The major issue with comparing 2014 and 2024 is the long time gap, which makes it impossible to isolate the causal factor(s) that produced changes in levels of antisemitism. Another country-level comparison looked at nations that the ADL Global 100 surveyed in 2023 and 2024. These nations come exclusively from Europe, and all but Russia show modest changes in antisemitism.

Still, all these analyses are aggregated. We would also like to know if certain types of individuals became more antisemitic because of the events in the Middle East. The next section turns to ANES data, which allows such individual-level analysis.

4. Trends in Antisemitism, 1964–2024 (ANES)

This section replicates Cohen’s (2018) analysis of ANES individual-level data on a Jewish feeling thermometer question from 1964 to 2016, extending the analysis through 2024. A primary limitation of Cohen’s study is the emphasis on demographics, where personality, situational, and other variables may also affect such ratings, but limitations of questions ANES asks its respondents constrains and limits analysis using the full pool of surveys from 1964 through 2024. Further, there is only one question on attitudes toward Jews, the feeling thermometer, which may be a blunt measure (Allington et al. 2022).

Cohen presented several sets of findings. One, there is a general warming toward Jews from the mid-1960s into the 2010s: mean warmth rose from 62 in 1964 to 70 in 2016. He investigated eight variables: black, education, age, female, party identification, Catholic, a black thermometer, and a white thermometer. Overall, with a control for the time trend, he found that blacks are cooler toward Jews, while those with higher education, Catholic, and older are associated with greater warmth. The black and white feeling thermometers also strongly relate to the Jewish feeling thermometer, suggesting that attitudes toward Jews may also be related to attitudes toward other groups (Hersh and Royden 2023b). Overall, the demographic variables do not account for much variance, with an R² = 0.02, but this rises to 0.42 when the black and white feeling thermometers are added.

Cohen also notes some trends in group assessments of Jews, as the effects of education and being Catholic weaken, while the association with age strengthens. This last finding is of special interest, as the thermometer ratings between younger and older people has widened over time. There has been considerable news coverage of anti-Israel demonstrations and attitudes among the young because of the Israel–Hamas war, which some fear may spill over to foster greater antisemitism among the young too. The age–antisemitism relationship has received attention in the literature (Hersh and Royden 2023a, 2023b).

Figure 3 plots the average predicted feeling thermometer rating for each year. As in Cohen, there is a dip in warmth in 1976, but still there is a general upward trend in warmth from the low 60s earlier in the series to the upper 60s in the latter part of the series. However, there is a slight decline in warmth from 2020 to 2024, from the all-time high of 71.5 to 68.6. The difference between 2020 and 2024 is statistically significant, even if it is substantively modest. Table 3 presents the results of regressing dummies for 2020 and 2024 on the feeling thermometer. With a constant (or overall average) of 64.2, the predicted value for 2020 is 71.1 (64.2 + 6.9) compared to 68.6 (64.2 + 4.4) for 2024, a difference of 2.5 degrees, which an F test indicates is significantly different (F = 17.8, p = 0.000).

Figure 3. Predicted Jewish feeling thermometer, ANES 1964–2024.

Table 3. Regression of dummies for 2020 and 2024 on the Jewish thermometer rating, ANES 1964–2024.

2024	4.40 ***
	(0.50)
2020	6.93 ***
	(0.39)
Constant	64.22 ***
	(0.15)
Observations	32,462
R-squared	0.02

Standard errors are in parentheses. *** p < 0.01.

Can we account for the drop in warmth from 2020 to 2024? To do this, the analysis interacts a dummy variable for 2024 with the variables used in Cohen (black, party identification, age, education, female, Catholic, black thermometer, white thermometer) plus a time trend variable (year) and dummy for 1976 (to pick up the drop in warmth that year). Results of the regression are presented in Table 4.

Table 4. Interaction of dummies for 2020 and 2024 with independent variables, Jewish thermometer ratings, pooled ANES 1964–2024.

	1964–2016	2020	2024	F Test 2020–2024
Dummy 2020	10.42 ***
	(2.51)
Dummy 2024	15.13 ***
	(3.21)
Black	−5.59 ***	−5.52 ***	−3.16 *	1.05
	(0.53)	(1.57)	(1.84)	(0.31)
Party ID	0.10 *	0.12	0.57 **	1.94
	(0.06)	(0.17)	(0.29)	(0.16)
Education	2.16 ***	0.51	−1.64 ***	22.29
	(0.15)	(0.43)	(0.26)	(0.000)
Age	0.08 ***	0.06 ***	0.08 ***	0.46
	(0.01)	(0.02)	(0.03)	(0.50)
Female	0.63 **	0.26	2.68 ***	4.26
	(0.26)	(0.73)	(0.99)	(0.04)
Catholic	2.33 ***	−2.03 **	−2.17 *	0.01
	(0.29)	(0.81)	(1.21)	(0.91)
Black therm.	0.40 ***	−0.01	−0.21 ***	37.15
	(0.01)	(0.03)	(0.02)	(0.000)
White therm.	0.28 ***	−0.17 ***	0.02	21.12
	(0.01)	(0.03)	(0.03)	(0.000)
Year trend	0.07 ***
	(0.01)
Dummy 1976	−4.68 ***
	(0.42)
Constant	−134.19 ***
	(16.97)
Observations	30,118
R-squared	0.32

Standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.10.

For a variable to account for the dip in the thermometer from 2020 to 2024, the sign on the 2024 coefficient must be negative and the coefficients for 2020 and 2024 should be statistically significantly different. The column labeled 2024 in Table 4 reports only three negative coefficients, −3.16 for being black, −1.64 for education, and −0.21 for the black feeling thermometer. The F tests for coefficient differences only attain statistical significance for education and the black feeling thermometer. Still, it can be difficult to interpret interaction effects from tabular regression results alone, especially if the relationship changes with a variable’s value. Thus, the analysis now presents marginal effect graphs for these eight variables, arrayed in two figures to help visualize results.

Figure 4 and Figure 5 plot the effects of the eight independent variables, separating the effects for 1964–2016, 2020, and 2024. Separating 2020 from 2024 will help us identify if any of the independent variables can account for the small drop in the feeling thermometer between those years.

Figure 4. Interactions of 2020 and 2024 with independent variables, part 1. Note: From results in Table 4.

Figure 5. Interactions of 2020 and 2024 with independent variables, part 2. Note: From results in Table 4.

The primary comparison of interest is the relative position of the slopes for 2020 and 2024. For the independent variable in 2024 to account for some of the drop in warmth from 2020 to 2024, the 2024 slope should fall below that of 2020, or at least a segment of the 2024 slope should. Results in Figure 4 and Figure 5 in all cases show the 2024 slope sitting above that for 2020: for all these independent variables, 2024 predicted thermometer ratings are warmer than for 2020. None of the variables helps us explain the reason for the dip in warmth toward Jews from 2020 to 2024.

Using the pooled ANES data for 1964–2024 shows that none of the demographical variables used in Cohen can help explain the dip in warmth toward Jews from 2020 to 2024. Rather, all the variables suggest equal or greater warmth toward Jews for each demographic group and category in 2024 than 2020. To account for the dip in warmth from 2020 to 2024 probably requires inspection of other variables. The analysis thus far has only focused on demographics employed across all the ANES surveys from 1964 through 2024. The next section turns to the 2020–2024 panel, when ANES asked questions about Israel, allowing a test of the new antisemitism perspective, which argues that attitudes toward Israel will affect antisemitic sentiment.

5. Change at the Individual Level: ANES 2020–2024 Panel

The above analysis using the pooled ANES found no variable that could account for the dip in warmth toward Jews from 2020 to 2024. To conduct the analysis using the pooled ANES required limiting the variables to those asked across polls that contained the dependent variable, the feeling thermometer toward Jews. Those variables were primarily demographic, but research on antisemitic opinion holding shows that factors beyond demographics, such as personality and attitudes, may be consequential.20

Criticism of Israel, according to the new antisemitism perspective, may lead to increased antisemitism (Abady 2019; Chesler 2015; Sharansky 2004; Wistrich 2016). The pooled ANES did not contain any long-running question regarding attitudes toward Israel, and thus those data could not be used to test the new antisemitism perspective. The sharp rise in criticism of Israel in reaction to the Israel–Hamas war provides a critical context for testing the new antisemitism perspective. Although numerous studies have found a correlation between attitudes critical of Israel and antisemitism, those relationships are empirically weak (Allington and Hirsh 2019; Allington et al. 2022; Baum and Nakazawa 2007; Beattie 2017; Brym 2024; Cohen et al. 2009; Hersh and Royden 2023a; Kaplan and Small 2006; Kempf 2012; Jaspal 2016; Shenhav-Goldberg and Kopstein 2020; Staetsky 2020).Those studies were also conducted on Western societies where most citizens did not display critical attitudes toward Israel (BenLevi et al. 2019). The changing context of public criticism of Israel during the Israel–Hamas war period presents a strong test case for the new antisemitism perspective.

Two design elements of the 2020 and 2024 ANES allow for a limited test of this hypothesis. First, ANES conducted a panel by reinterviewing the same respondents in 2020 and 2024, which enables us to assess individual-level attitude change. The pooled ANES only allowed an investigation of changes in attitudes of types of individuals. Second, the 2024 ANES contained three questions on Israel, support for US military assistance to Israel, and relative support for the Palestinians versus Israel asked in the pre-election 2024 wave and a question on whether the respondent thought Israel posed a threat to US security asked during the post-2024 election wave.

5.1. Measuring Attitudes Toward Israel in the 2024 ANES

The wordings of the three questions on attitudes about Israel in the 2024 ANES read (variable number in boldface type):

V241403x: “Do you favor, oppose, or neither favor nor oppose the United States giving military assistance to Israel?” with a follow-up, “Do you (favor/oppose) that [a great deal, a moderate amount, or a little/a little, a moderate amount, or a great deal]?” These are the resulting categories (codes in parentheses): favor a great deal (1), favor moderately (2), favor a little (3), neither favor nor oppose (4), oppose a little (5), oppose moderately (6), oppose a great deal (6).

V241409x “In the Middle East conflict, do you side more with [the Israelis/the Palestinians], both equally, or neither?,” with the follow up, “How much more do you side with the [Israelis/Palestinians]? A little, a moderate amount, or a lot? These are the final categories (codes in parentheses): side a lot with Israelis (1), side a moderate amount with Israelis (2), side a little with Israelis (3), side with both equally or side with neither (4), side a little with Palestinians (5), side a moderate amount with Palestinians (6), side a lot with Palestinians (7).

V242370: “Next, we’ll ask you the same question for several countries: How much is Israel a threat to the United States? [Not at all, a little, a moderate amount, a lot, or a great deal/a great deal, a lot, a moderate amount, a little, or not at all]?”, with the following coded categories (codes in parentheses): not at all (1), a little (2), a moderate amount (3), a lot (4), a great deal (5).

Table 5 presents the distribution of opinion on these three questions. All three questions show more support for than opposition to Israel. On giving military aid, approximately 38% support compared to 28% who oppose. Importantly, about one-third neither support nor oppose. About 31% are sympathetic to Israel compared to only about 11% who sympathize with the Palestinians. Most consequently, 20% support both equally and another 38% say they side with neither. While more Americans sympathize with Israel over the Palestinians, the bulk do not take a side (58%). Finally, nearly 42% do not see Israel as a threat to the US, while only 13% say that Israel poses a lot of or a great deal of threat to the US. We cannot say whether Americans have become less supportive of Israel, because these questions on the ANES only exist on the 2024 study. Overall, these distributions indicate relatively strong support for Israel among Americans.

Table 5. Distribution of opinion on questions about Israel, ANES 2024 panel (weighted and Jewish respondents excluded).

Panel 5A. Military Aid to Israel
PRE: SUMMARY: Favor/oppose US giving military assistance to Israel	Freq.	Percentage	Cum.
1. Favor a great deal	421.89	20.94	20.94
2. Favor moderately	257.90	12.80	33.74
3. Favor a little	85.32	4.23	37.97
4. Neither favor nor oppose	677.99	33.65	71.62
5. Oppose a little	47.45	2.35	73.97
6. Oppose moderately	182.23	9.04	83.02
7. Oppose a great deal	342.22	16.98	100.00
Total	2015	100.00
Panel 5B. Sympathize with Israel or the Palestinians
PRE: SUMMARY: Side more with Israelis or Palestinians?	Freq.	Percentage	Cum.
Side a lot with Israelis	354.56	17.66	30.59
Side a moderate amount with Israelis	178.70	8.90	12.93
Side a little with Israelis	80.94	4.03	4.03
Side with both equally	411.41	20.49	51.08
Side with neither	763.05	38.00	100.00
Side a little with Palestinians	22.88	1.14	62.00
Side a moderate amount with Palestinians	77.47	3.86	60.86
Side a lot with Palestinians	118.98	5.93	57.00
Total	2008	100.00
Panel 5C. Israel Poses a Threat to the US
POST: How much is Israel a threat to the United States?	Freq.	Percentage	Cum.
Not at all	830.90	41.71	41.71
A little	431.14	21.64	63.36
A moderate amount	476.83	23.94	87.29
A lot	140.58	7.06	94.35
A great deal	112.56	5.65	100.00
Total	1992	100.00

5.2. Do Attitudes Toward Israel Affect Attitudes Toward Jews?

Fortunately, the ANES panel study asked respondents the feeling thermometer toward Jews in both the 2020 and 2024. Thus, we can compare the stability or change in the Jewish feeling thermometer ratings for the same individuals at two points in time. However, a strong test of the new antisemitism perspective would also have the same questions on attitudes toward Israel asked on both panels, but ANES only asked each of the Israel questions once. Thus, we can only test a weaker form of the hypothesis: that support/opposition for Israel is related to changes in attitudes toward Jews.

These panel data show a slight dip in warmth toward Jews, like the pooled ANES data, with the average feeling thermometer (for non-Jewish respondents who completed the question on both panel waves, weighted) at 71.5 for 2020 and 68.9 for 2024, or a drop of 2.6 degrees. Significantly, the correlation between the two feeling thermometers is moderate, at a Pearson’s r = 0.48 (p < 0.001).

The modest correlation suggests instability in attitudes toward Jews across the two time points, echoing a similar finding in Cohen (2024), who used feeling thermometer data from the Voter Study group (VSG) 2011–2020. Cohen argued that for many individuals, attitudes toward Jews are better characterized as non-attitudes. Jews, according to Cohen, are not salient: non-Jews think about Jews little and they lack much information about Jews. When asked about Jews in surveys, non-Jews use easily accessed information, which might be the last thing they heard recently (if they heard anything recently), such as from trusted sources like the president.

Cohen’s (2024) data come from a period when there might not have been much news coverage about Jews (or Israel), but in 2024, due to the Israel–Hamas war, the campus and other protests, etc., there was likely much more news coverage of Jews and Israel. Hence, one might expect, based on Cohen’s theory, greater stability (a higher correlation) in the ANES 2020–2024 panel, but the correlation found with the ANES panel is modest. Perhaps news coverage of Jews was slight in 2020, but high in 2024. This variation in news coverage levels may have contributed to the modest correlation reported here.

5.3. Multivariate Analysis of Panel Data

Results of several multiple regression estimates are presented on Table 6. The dependent variable is defined as the 2024 thermometer rating minus 2020 thermometer rating. There are three sets of independent variables. The first consists of a set of demographics, like those used above. The specific demographics are age, education, female, black, Hispanic, party identification, Catholic, Protestant, and Muslim. These demographics address whether any type or group of individuals becomes warmer or cooler toward Jews from 2020 to 2024. The second independent variable set is the 2020 Jewish thermometer. This variable is used as a baseline and to correct for regression to the mean in the thermometer ratings. The third independent variable set consists of the questions toward Israel, each entered separately. These are the variables of greatest interest for the purposes here.

Table 6. Impact of attitudes toward Israel on changes in individual thermometer ratings of Jews, 2020–2024 ANES panel (weighted data, Jewish respondents excluded).

	Estimations
	6A	6B	6C	6D	6E
Age	0.10 *	0.19 ***	0.15 ***	0.12 **	0.15 ***
	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)
Education	1.01	1.74 ***	1.90 ***	1.79 ***	1.46 **
	(0.70)	(0.63)	(0.62)	(0.61)	(0.61)
Female	4.23 ***	4.87 ***	5.67 ***	5.23 ***	5.78 ***
	(1.57)	(1.40)	(1.39)	(1.39)	(1.41)
Black	−3.72	−6.41 **	−6.33 **	−4.72	−5.52 *
	(3.71)	(3.15)	(3.21)	(2.98)	(3.01)
Hispanic	−0.97	−10.09	−0.63	−0.28	−0.55
	(2.59)	(2.43)	(2.46)	(2.36)	(2.47)
Party ID	1.06 ***	0.43	−0.06	−0.15	0.13
	(0.36)	(0.31)	(0.36)	(0.32)	(0.31)
Catholic	−0.54	0.34	0.02	0.19	0.08
	(1.83)	(1.64)	(1.61)	(1.61)	(1.64)
Protestant	−1.22	1.49	0.27	0.36	0.78
	(1.86)	(1.66)	(1.65)	(1.59)	(1.70)
Muslim	−23.41 ***	−34.86 ***	−30.97 ***	−32.49 ***	−29.87 ***
	(7.63)	(8.29)	(7.62)	(7.57)	(7.53)
Jewish Therm. 2020		−0.54 ***	−0.54 ***	−0.56 ***	−0.55 ***
		(0.04)	(0.04)	(0.03)	(0.03)
Israel vs. Palestine			2.13 ***
			(0.51)
Israel—Military Aid				−2.10 ***
				(0.38)
Israel is a Threat					−2.94 ***
					(0.7)
Constant	−16.05 ***	17.32 ***	11.28 **	32.41 ***	28.07 ***
	(4.97)	(5.08)	(4.85)	(5.75)	(5.6)
Observations	1839	1839	1832	1834	1827
R-squared	0.04	0.28	0.29	0.31	0.3

Standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.10.

Although several of the demographics attain statistical significance in the estimations, collectively they do not account for much of the variance in the change in the thermometer rating when they are the only variables included (Estimation 6A, R² = 0.04). Two have negative signs, indicating that members of that group became cooler toward Jews, blacks and Muslims. The coefficient for blacks is significant in most estimations, with declines from 4–6 degrees. The Muslim coefficient is much larger, from 23 to 35 degrees, but Muslims constitute a very small percentage of the population, so they can only contribute a fraction of the dip in warmth across the panel.

Adding the 2020 thermometer rating to the estimation improves model fit (Estimation 6B, R² = 0.28), and the variable is statistically significant (b = −0.54, p < 0.001), suggesting strong regression to the mean effects, which should be expected given the instability in the feeling thermometer.

Turning to three of the Israel questions, each is statistically significant, indicating that criticism or opposition to Israel leads to cooler attitudes toward Jews, as the new antisemitism hypothesis predicts. On the question regarding siding with Israel versus the Palestinians (Estimation 6C), the regression coefficient suggests that a one-step shift in the seven-step variables corresponds to a two degree shift in warmth. A one-standard-deviation change in this variable (1.5 steps) will lead to a three-degree change in warmth, while the mean value (4.6) corresponds to a −2.5-degree downward shift in warmth.

Second, the coefficient for the military aid question (Estimation 6D) suggests that a one-step change in the variable corresponds with a 2-degree change in warmth. This variable has a standard deviation of two steps on the seven-step variable, and a standard deviation change could produce a 4-degree change in warmth. At its mean value (3.6), this variable predicts a −2.3-degree drop in warmth toward Jews. Third, the variable of whether Israel is a threat to US security suggests that a one-step change in this five-point variable will lead to a 3-degree change in warmth. A one-standard-deviation shift, one step, will also produce a 3-degree change in warmth. And at the mean value of 2.0, warmth toward Jews will decline by 2.4 degrees.

Each of the Israel questions separately can account for the drop in warmth toward Jews, which was estimated above for the panel at 2.6 degrees. Still, it is important to note the limitations of this analysis. The Israel questions were measured at one point in time. A stronger test would be if changes in attitudes toward Israel lead to changes in the feeling thermometer toward Jews. The test here only shows that negative attitudes toward Israel are associated with change in individual thermometer ratings of Jews consistent with the new antisemitism thesis.

6. Conclusions

The new antisemitism perspective argues that criticism toward Israel will foster antisemitic attitudes more generally, no matter where Jews live. This paper addressed that question in the context of the mounting criticism of Israel due to the Israel–Hamas war, in which tens of thousands of Gazans lost their lives. In contrast, there was much less attention internationally given to the Hamas terrorist attack and hostage taking, leading some critics to think the international reaction was insensitive toward Israel and potentially antisemitic.

The analysis employed several types of data, all of which have their limitations, as noted here. First, the analysis turned to the ADL Global 100, making two sets of comparisons of data from 2014 and 2024 of nearly 100 nations. Those comparisons found increasing antisemitic attitudes in about one-half of the nations surveyed, with about one-third showing stable antisemitism levels and about 14% reporting lower levels of antisemitism. But there was no easily observed pattern across nations for why some became more antisemitic. Moreover, the ten-year gap between the two global surveys renders it impossible to say definitively that the Israel–Hamas war led to the changes in antisemitism levels.

A second analysis looked at ADL Global 100 survey results for ten nations polled in 2023 and 2024. The ten nations, however, were exclusively European. Nine of these countries showed little movement in antisemitism levels, but one, Russia, saw a massive increase, probably due to Putin’s anti-Israel policies and associated news coverage in the Russia press.

The analysis also used data from the American National Election Study, first a pooled analysis of polls from 1964–2024 that included a feeling thermometer about Jews. Analysis of these data detected a small 2- to 3-degree drop in warmth toward Jews from 2020 to 2024. This drop, despite being small, is still statistically significant. Further analysis of these pooled data was unable to identify the source of the dip in warmth toward Jews.

Finally, the analysis used a 2020–2024 ANES panel, which included several questions on Israel. Each question on Israel was associated with a decline in warmth toward Jews of 2–3 degrees, but since the Israel-related questions were asked only once, analysis could not answer whether changes in attitudes toward Israel lead to changes in attitudes toward Jews, a stronger test of the new antisemitism thesis. The modest impact of Israel questions on attitudes toward Jews reported here are consistent with past research on the new antisemitism, which also reports small to modest correlations.

It is unclear whether the mounting criticism of Israel due to the Israel–Hamas war has led to increased antisemitism in public opinion. That appears the case in some countries, but not others. Country-specific factors are necessary to determine why antisemitism rose in some places, but not others. The detailed analysis of US data suggests a small rise in antisemitism of 2–3 degrees on the feeling thermometer.

In some places, people can and do distinguish between Israel and Jews, but in other places that distinction does not appear to be so firmly held, perhaps as the Russia case illustrates. However, we need better data on Russia and other nations that showed large increases in antisemitism to outline the conditions that fuse attitudes toward Israel with attitudes toward Jews. The new antisemitism thesis may be context-specific. Fighting antisemitism would benefit from understanding those contextual conditions better.