The Brookings Institution’s Saban Center is a fixture of Washington’s Middle East policy establishment. Prime ministers and secretaries of state have spoken at the center’s yearly forum, and when a respected Brookings-affiliated researcher finds that a 46 percent plurality of Americans back U.S. action on the Israeli-Palestinian conflict at the UN, the think tank’s imprimatur strongly suggests that the finding reflect some broader reality in public opinion. But University of Maryland professor Shibley Telhami’s latest poll of American attitudes on the Israeli-Palestinian conflict, published earlier this month, is worth a closer look, both in spite of, and because of, its lofty provenance. Along with its discovery of a broad base of support for prospective American peacemaking efforts at the UN, the study contains a number of startling findings.
According to the survey, 55 percent of Democrats agree with the statement that Israel is a burden to the United States—but 70 percent also agree with the statement that the country is an “important ally.” Some 46 percent of respondents—including 60 percent of Democrats—agree with the U.S. imposing “economic sanctions or more serious action” over Israel’s West Bank settlement policies. Only 34 percent of respondents, and just a slim 51 percent majority of Republicans, support the U.S. using its Security Council veto to block a resolution on Palestinian statehood.
The survey gives the impression that Americans are more supportive of their government pressuring Israel on the peace process than is generally assumed. After all, according to a February Gallup poll, 71 percent of Americans hold a favorable view of Israel, and 62 percent sympathize with the country compared to 15 percent for the Palestinians. Telhami’s findings paint a slightly different although not technically conflicting picture, as they hint that the American public would tolerate or even welcome dramatic shifts in the U.S.’s longstanding posture. At the very least, Telhami’s results reveal a certain lack of clarity in terms of how Americans work through the issue in their own minds. In his view it isn’t all that surprising that an apparent majority of Democrats view Israel as both an ally and a burden: “It’s not uncommon to have respondents agree with opposing arguments, and it does indicate ambivalence,” he wrote by email.
The findings would complicate the conventional wisdom that Americans are reflexively supportive of pro-Israel moves from their government and reflexively against anti-Israel ones—as the New York Jewish Week put it, the study shows that Democrats have “turned sharply on settlements.” But there are aspects of the survey and its rollout that urge caution and even skepticism.
The survey questionnaire was not published directly on the Brookings website, although it is included on the page for the Critical Issues Poll administered by the University of Maryland, the academic institute that actually conducted the study (Brookings, where Telhami is a non-resident senior fellow, hosted the report’s launch event). The questions on U.S. policy towards Israel were asked in the context of a much longer survey on a host of political issues, and came after scores of often-lengthy questions related to U.S. policy in other parts of the Middle East—something that could conceivably have skewed or informed the outlook of respondents who may have had little prior knowledge of the Israeli-Palestinian conflict. The poll also included a number of agree/disagree questions—although it’s now recognized that agree/disagree poll questions can produce an “acquiescence bias” in respondents that inflates the actual level of support for a given proposition.
The questionnaire also did not include a “no opinion” or “don’t know” option, something that might have dramatically inflated the percentage of respondents even expressing a view on certain topics. The researchers reported that 34 percent of Americans supported the U.S. voting in favor of a UN resolution on Palestinian statehood, 31 percent wanted a U.S. veto, and 32 percent backed U.S. abstention—results implying that 97 percent of Americans hold an opinion on a fairly obscure and technical foreign policy issue. For comparison, an October ABC news poll found that 40% of respondents couldn’t name the major party vice presidential candidates. (Telhami explained by email that “Respondents were able to skip questions in the survey and in cases where they did, their answers were coded as “refused” on the questionnaire. For the questions listed in this specific questionnaire, ‘don’t know’ was not an option. Respondents simply had the answers choices listed out that they could choose from or they could choose to skip the question, except in the case where a question was open-ended, in which case they were free to write in anything that they chose or they could skip the question”).
Some potentially helpful information is also missing from Brookings and the University of Maryland’s published materials. The report published demographic crosstabs for party identification, but does not include a polling breakdown by gender, race, religion, or education level. (It is not unusual for pollsters to refrain from releasing all of their demographic sub-data at once, and Telhami said he may publish additional cross tabs at a later time.) The study did not explain how the margin of error was calculated, and while the results are weighted for party identification the report itself did not indicate the source of the statistical baseline used for the weighting (by email, Telhami said the baseline was Gallup’s average of its surveys on party identification over the past six months).
Ken Goldstein, a professor of politics at the University of San Francisco and expert on polling research and analysis was critical of the study. “This is not a level of transparency that I would expect from an academic or media poll and it’s unambiguously best practices to offer a ‘don’t know’ option when asking about complicated policy issues,” said Goldstein.
Another polling expert, Scott Keeter, a senior survey adviser at Pew research, wrote by email that the study had met all major transparency standards, since the report included field dates, sample sizes, information about methodology, and a full questionnaire. For instance, prevailing industry standards “do not require organizations to present crosstabs as a standard disclosure element, in part because it’s not clear what among the myriad possible crosstabs would be relevant.” Still, there were other, minor issues that Keeter identified, like a lack of information on mode of data collection and margin of error computation. Keeter said that Pew only rarely offers an explicit “don’t know” option in its surveys, but added that the lack of such an option could still have had an effect on polling results. “Given the subject matter (foreign policy), the rates of ‘don’t know’ in the Brookings/Maryland survey are pretty low, indicating that DK or Refused option was not explicitly offered to respondents. If you offer it on questions that people don’t think about a lot, you tend to get pretty high percentages of people opting for it.”
So what does the poll actually reflect? Respondents answered questions pertaining to head-spinning policy issues without an explicit “don’t know” option and at the tail-end of a lengthy questionnaire that included inherently suspect agree/disagree question formulations. Americans are probably as conflicted on the U.S.’s proper response to the Israeli-Palestinian impasse as they are on any number of other vexing public policy issues. The poll conveys some of the confusion that the finer points of the conflict no doubt inspires, but it also might have captured the effects of its own idiosyncrasies along the way.
More significantly, it conveyed results that wouldn’t have seemed particularly plausible if they hadn’t been published under the Brookings logo. If American public opinion actually does correspond with the study’s findings, a U.S. administration would be given broad cover to pursue a historic shift in how America approached its most critical ally in the Middle East, while the Democratic party would have the support of a solid majority of its members in advocating even more drastic measures, like sanctions against Israel. The realities of U.S. politics—in which American policy towards the Middle East conflict is marked by relative continuity and support for the country remains solidly bipartisan—tell a much different story, both about the possible state of public opinion and the astounding legitimizing power of Washington’s premiere policy institutions.