7 days ago
Polling About War Crimes Comes With Extra Responsibilities
Against the backdrop of Israel's escalating ground offensive in Gaza and this weekend's massacre at a humanitarian aid distribution site there, several recent polls show Israeli support for the war in Gaza on the decline. Approximately 70 percent of Israelis now prefer to end the war if it means bringing home the hostages still being held in the territory.
Distinguished Israeli leaders are also speaking out more forcefully against what they view as crimes against humanity committed by Israeli forces in Gaza. In a recent op-ed, former Prime Minister Ehud Olmert wrote Israel is waging a 'war of extermination: the indiscriminate, unrestrained, cruel, and criminal killing of civilians,' a view he reiterated on CNN on Sunday. Over 1,200 Israeli academics have penned a letter calling for an end to the suffering of the Palestinians. And as many as 40 percent of Israeli conscripts are refusing service; one stated, 'I would prefer [jail] over killing children.'
But even as this peace—or anti-war crimes—movement gains ground, a much more disturbing poll of Israeli citizens published on May 22 alleges that nearly half of Israelis support the extermination of all Palestinians in Gaza. This poll had asked Israelis to signal their support for or opposition to Israeli policies in Gaza as well as dehumanizing rhetoric, particularly biblical references to exterminating enemies, popularized by rightwing religious figures and political elites since Hamas' attacks of Oct. 7, 2023. The survey focused on the rhetoric of ultra-orthodox Rabbi Yitzhak Ginsburg, but Prime Minister Benjamin Netanyahu has used similar references to justify Israel's siege warfare against Gaza, which has killed an estimated 64,000 people—a third of them under 18 years old—according to Reuters, with many more at risk of starvation.
The poll was conducted using rigorous weighted sampling methods by Israeli polling firm Geocartography Knowledge Group at the request of two U.S.-based historians—Tamir Sorek and Shai Hiskani of Penn State University and the University of Maryland, respectively. None of the standard meta-data—including margins of error, demographic breakdowns and comprehensive question wording—appear to have been made public in English, but the scholars' stated aim, according to an op-ed in Haaretz, was to 'ask impolite questions' to discover 'just how appealing the apocalyptic language offered by Ginsburg is to Israeli-Jews from all walks of life.'
To get more in-depth news and expert analysis on global affairs from WPR, sign up for our free Daily Review newsletter.
The poll asked respondents whether they 'support the claim that the [Israeli military], when conquering an enemy city, should act in a manner similar to the way the Israelites acted when they conquered Jericho under the leadership of Joshua, that is, kill all its inhabitants?' Just under half of respondents—47 percent—responded in the affirmative. And 82 percent supported forced displacement of Palestinians from Gaza, which may have sounded to respondents like a more humane option by contrast to extermination.
Given these scholars' previous work on conflict and resistance, the publication of their op-ed in a left-wing, secular newspaper like Haaretz, and a separate piece outlining their concerns over what they refer to as the 'mainstreaming' of genocidal rhetoric, it is reasonable to believe their intent was to generate opposition to war crimes in Gaza, if not the war itself. But unfortunately, based on what social scientists know about the politics of conflict research, it is more likely this survey could have an exacerbating rather than constraining effect on Israel's campaign in Gaza.
The first thing to know is that survey respondents who are invited to express support for war crimes may be likelier to believe such acts aren't crimes at all. This effect is illustrated by research conducted by myself and colleagues Alexander Montgomery and Alexandria Nylen a few years ago. We analyzed what happens to U.S. public attitudes on the laws of war when respondents are asked to express a preference for or against a policy that qualifies as a war crime. The fictional scenario in question involved a gross violation of humanitarian law—firebombing an Iranian city—presumably to end a war, a question we replicated from an earlier study.
One difference between the original study and ours is that we also asked our respondents their attitudes on the lawfulness of targeting civilians in the abstract. But we asked a randomly selected half of the respondents this other question before they were asked their views on the Iran scenario, and the other half after. This enabled us to determine not only whether respondents' baseline views on targeting civilians affected their attitudes toward the firebombing scenario, but also to measure how much the experience of first answering the firebombing question affected their views on the legality of targeting of civilians.
Both effects were evident. Knowledge of international law, or even just being asked to think about international law, reduced support for the deadly strike. But conversely, being asked to express a policy opinion about targeting actual civilians reduced respondents' belief that doing so was even banned by international law at all—even for those who in fact opposed the strike. So researchers can inadvertently affect citizens' viewpoints on what conduct is legally permissible in the way they ask survey questions. Since studies show international law has a powerful moderating effect on citizen attitudes, when researchers undermine those understandings, they may inadvertently 'prime' respondents to go along with war crimes.
Second, this effect can actually be exacerbated when such survey results are disseminated in the media through press releases or op-eds that get picked up and uncritically circulated by other sources. For example, the bombing survey we were replicating had been publicized with headlines like 'Americans Wouldn't Mind if We Just Nuked Somebody.' To test the impact of this dynamic, we created a treatment group in our survey, informing them about the earlier survey result that Americans were 'okay' with targeting civilians. That knowledge increased their own willingness to firebomb the Iranian city. Ours was a fictional scenario, but this effect may be worse in ongoing conflicts, such as Israel's war in Gaza, where the media ecosystem is especially polarized.
Finally, publicizing popular support for war crimes risks affecting conflict dynamics directly. A perception of public support, however accurate, can make it easier for politicians to order abuses and dilutes the political space for resisting atrocities. The perception that support for war crimes is widespread among a population also increases support for crimes against that population. This risks putting all civilians in danger of revenge attacks for crimes perpetrated by their nation's army, even if many of them actually oppose those crimes, as 53 percent of Israelis did in the poll in question.
Understanding the political psychology of public support or opposition to war crimes is vitally important, and we know what we do about the subject thanks to good survey research. Nevertheless, because of the framing effects that survey designs themselves create, pollsters and social scientists are increasingly cognizant of their role in measuring public opinion—but also their responsibility to avoid potentially influencing it in ways that harm. That responsibility is even greater when asking respondents about ongoing atrocities rather than fictional scenarios.
Some best practices are emerging in the social sciences and the opinion polling industry to gather information on conflict psychology while minimizing potential harm in high-risk environments. Among these are transparency—pollsters should always make their methodologies and question wording easily available to audiences; neutrality—pollsters should adopt a nonpartisan stance applying equally to both conflict parties; and protection of human subjects' data—pollsters should ensure informed consent and confidentiality.
These precautions protect vulnerable populations and the integrity of the research enterprise. But conflict researchers can go beyond this to ensure their work minimizes social harm.
First, when it comes to researching war crimes, conflict researchers and pollsters can use the preamble to the actual survey questions to highlight current international legal standards. They can then invite respondents to evaluate policymakers' actions and rhetoric against those standards, rather than simply inviting them to express support for illegal acts. A study structured in this way would likely yield a different result.
Of course, the purpose of the study conducted by Sorek and Hiskani was specifically to analyze the influence of apocalyptical biblical references on Israeli attitudes toward the war in Gaza, rather than legal standards alone. So how might researchers get at questions like this without giving the impression that such policies are legitimate?
One possible solution would be to compare two respondent groups, each receiving a neutral question about proposed policies in Gaza and facts about current international law, but with only one group also hearing mention of the biblical reference to Jericho in the preamble, rather than baked into the question itself. This approach captures the impact of rhetoric on answers about policy without privileging that rhetoric or inviting citizens to directly endorse it themselves. And it allows a comparison of the two groups' answers in order to isolate the effect of the framing.
However, social scientists already know a lot about how genocidal rhetoric affects attitudes, and less about how to counter it. So, one could also imagine studying the effect of biblical references that support tolerance, restraint and protecting the innocent rather than those that include genocidal rhetoric—or even comparing the influence of both. And pollsters can build on what we already know: Even just adding an informational disclaimer at the end of such a survey, citing actual international legal standards, while doing so in any press releases as well, is a way to inoculate respondents and audiences against the pernicious effects of thinking about the unthinkable.
Another solution is asking respondents to think aloud about how to solve specific policy problems, without priming them with specific policy options at all. This creates a more neutral window into what resonates in a particular political moment without the risk of over-determining the result. For example, in a separate Iranian firebombing study, we asked one group of respondents not a close-ended question about whether they supported or opposed the strike, but rather an open-ended question: 'If you were the Commander-in-Chief what would you do in this scenario?' The number wishing to bomb the city dropped dramatically when not forced into one of two boxes. Many respondents thought outside the box altogether, a sorely needed ability in polarized conflicts.
The point is not to criticize this particular survey but to highlight choices pollsters can make in conflict zones. Like other forms of reporting on war and peace, surveys carry risks but can also have a constructive informational effect on conflict-affected populations: amplifying alternative frames, opening space for resistance and yielding results that, when published, can support efforts to resolve rather than escalate violence by stressing humanizing rather than dehumanizing narratives across collective groups.
The same critical lens could be applied to other opinion polls on Israel's war in Gaza. Polls asking Israelis whether they care more about eradicating Gaza or bringing the hostages home imply these are the only options—and mutually exclusive ones at that. And polls asking Israelis whether they support the war rarely ask, among those who say yes, whether they nonetheless oppose war crimes. Measured correctly, the opposition might be much stronger than observers think, if not as strong as they wish.
Without that ethical nuance, such polls are unlikely to fully capture public opinion and may instead risk perpetuating a polarizing media climate that increases the possibility of war crimes and decreases the chances for peace. Rather than dwell on the resonance of the more alarming narratives measured by this latest poll, it may be worth thinking about how to amplify and measure the impact of the Israeli and Palestinian voices now fighting so hard to offer alternatives.
Charli Carpenter is a professor of political science and legal studies at University of Massachusetts-Amherst, specializing in human security and international law. She tweets at @charlicarpenter.
The post Polling About War Crimes Comes With Extra Responsibilities appeared first on World Politics Review.
