Underpolling Arabs

The gap in public opinion knowledge probably exacerbates the distance between Jews and Arabs in Israel.

Underpolled 521 (do not publish again) (photo credit: Flash 90)
Underpolled 521 (do not publish again)
(photo credit: Flash 90)
ISRAELIS LOVE TO SAY WHAT they think. And they also seem to love hearing what their fellow Israelis think, judging by the wide range of polls covering many different public issues. But when Israeli Jews say they are fascinated by their fellow Israelis, they usually mean they are most interested in how other Jews think.
When it comes to public affairs, the voice of Arabs is insufficiently heard in surveys and the reporting in the media is often misleading or mistaken. Readers of The Jerusalem Report deserve some explanations as to why this is the case.
Many surveys are described as having either a “representative sample,” or a “Jewish sample.” The term “representative sample” means the survey included Arab respondents. Ironically – although some is better than none – “representative” surveys are usually inadequate representations of the Arab population. Arabs make up about 16.5% of the adult population and 20% of the whole population of Israel. Since many media surveys have a total sample of 500 respondents, quick math will tell us that representative surveys probably include about 83 Arab respondents.
But professional pollsters consider 100 respondents to be the minimum for a viable sample from which conclusions may be drawn. With such small samples, the margin of error is so high as to make results difficult to trust. So a “representative sample” based on 500 people hardly provides enough information to draw genuine conclusions regarding the Arab population with much authority.
In June 2010, the Hebrew daily, “Haaretz,” reported on the survey it commissioned from the Dialogue polling group led by Camille Fuchs, its regular pollster, who is a professor of statistics at Tel Aviv University. The survey tested attitudes regarding the IDF’s interception of the Turkish flotilla seeking to break the Gaza closure – in which Israel’s commandos killed nine violent activists. Like most of the Dialogue surveys the sample was 500 respondents in total, probably including 83 Arabs. Although the article did not cite the sample breakdown, “Haaretz” boldly proclaimed that after the flotilla events, support for Prime Minister Benjamin Netanyahu rose 12 points and over 80% of respondents felt that Knesset Member Hanin Zoabi from the Arab party Balad, who had participated in the flotilla, should be punished in some way. It’s hard to imagine that these findings hold true for the Arabs – but readers wouldn’t know because the article never mentions them.
In other cases, newspapers report on surveys about major public issues in which Arabs were not sampled at all. In December 2010, following the “Rabbi’s Letter” calling on Jews not to rent apartments to Arabs (see “Halakha or Hatred,” January 17), the Hebrew news site Ynet reported on a survey it had commissioned on the topic. Only Jews were sampled – perhaps based on the logic that Arab attitudes would be obvious.
In other cases, the Arab population may be appropriately sampled, but the survey is misrepresented in the reporting. In December 2011, the Hebrew daily “Maariv’s” website, NRG, reported on a survey conducted by the Dahaf polling institute, led by Mina Tzemah, for the Saban Forum for Middle East Policy. “Maariv” reported that the survey had a total representative sample of 500 (again, readers would have to guess that the survey included about 83 Arabs, or 16.5%). The survey dealt with attitudes towards the conflict, but also Jewish-Arab relations.
The NRG article confidently displayed findings such as: “Fifty-five percent of Arabs are not prepared to recognize Israel as a Jewish state under any conditions. Fiftyone percent of Arab respondents are against the loyalty oath.” But, in fact, these and other findings reported in the article relate to two separate surveys: one involved a Jewish-only sample of 500 and the other involved a sample of 600 “Israeli Arab/Palestinians.” Readers who understand the problem of small representative samples would never know from the article that this data actually was drawn from a large, separate Arab sample.
In contrast to the data above regarding rejection of a Jewish state, a different poll shows contrary findings. A study of youth conducted by the research company Ma’agar Mohot in February 2010 for the Citizens’ Empowerment Center in Israel (536 youngsters, 26% of them Arab – or 139 people) revealed that 75% agree with the right of Israel to be a Jewish and democratic state.
Other surveys provide just minimal information on Arabs: The monthly Peace Index survey of attitudes towards the conflict that is run by the Guttman Center of the Israel Democracy Institute, a Jerusalembased independent think tank, reaches 600 respondents, and presumably roughly 100 Arabs, but only reports on them selectively.
The Israeli part of the Joint Israeli- Palestinian polls by Jacob Shamir (at the Truman Institute of the Hebrew University) surveys over 900 respondents; in the last poll from December, the Arab sample should have been about 152 respondents and the data for Arabs is listed separately, for easy analysis.
Sammy Smooha, one of the country’s leading sociologists, has conducted extensive research among the Arab and Jewish communities for decades. But his academic studies are few and far between; average Israelis are more likely to be influenced by regular media reporting.
Unfortunately, this column, “By the Numbers,” is not able to sample Arabs at all.
The explanation is simple: Our surveys are conducted through “omnibus” polls – in which a number of clients can request just one or two questions in a weekly survey conducted by a research firm (New Wave Research conducts the surveys for The Jerusalem Report.) Typically, this “omnibus” polling does not include Arabs.
RESEARCH COMPANIES SUCH as New Wave explain that there is simply not enough demand from clients for such “omnibus” polling among Arabs. When it comes to political issues, and especially issues related to the Israeli- Palestinian conflict, it seems that either clients think Arab attitudes can be taken for granted or else it’s only the Jewish opinion that matters to them on such sensitive national issues.
Compounding the problem of low interest is the genuine difficulty of polling among the Arab population of Israel.
First, there are technical problems.
Hisham Jubran, founder and CEO of the Haifa-based qualitative research company, Afkar Research, tells The Report that he estimates that among 40 unrecognized Arab towns, some 60-70 thousand residents do not have land line phones – but surveys in Israel are still not conducted through cellular phones and face-to-face polling is costly and slow.
Additional issues are also at work here.
Many Arabs are suspicious of answering direct questions about public affairs. Jubran says Arabs are concerned when strangers call their homes. They are suspicious, he says, and ask themselves, “‘Who are you to call me at home? How do you know my number?’ Every Arab has some suspicion, even if it’s low intensity. There’s very little trust.”
Jubran offers two main reasons for low trust. First, Arabs may be less aware of public opinion surveys; the concept is not as familiar to them as it is to Jews – perhaps precisely because Arabs are polled less than Jews in Israel. The other possibility, he says, is “they know about public opinion, but they don’t see that it has any influence. [They ask,] “Why answer? What will it give us? It’s just another survey. Especially when it’s related to the state, parties, government – it’s just like elections – there will be no influence.” The result, he says, is either low response rates, or skewed answers.
Many Jews also feel less than socially efficacious. But as the majority population, they do not experience the same sort of insecurity and suspicion of the majority-group state that Arabs might.
Language is also an issue. It’s fairly obvious that most people would feel more comfortable answering a survey – especially about sensitive topics – in their native language.
But the Hebrew media almost never state whether the regular polls they commission are translated into Arabic or if respondents were interviewed in Hebrew. My guess is that for fewer than 100 respondents, surveys are not translated. With respondents already suspicious of strangers asking their political views, a Hebrew-language interviewer is not likely to put them at ease.
Furthermore, at best, most polling firms just hire Arab interviewers, translate questionnaires written in Hebrew and draw samples from Bezeq, the national phone company.
There are hardly any commercial Arab research firms. After all, who has the time or the money to do it right? Whatever the reasons, sadly, the gap in public opinion knowledge probably exacerbates the distance between Jews and Arabs in Israel. Further, in-depth Arab polling might reveal troubling findings.
In just one example, Professor Smooha’s excellent “Index of Arab-Jewish Relations in Israel 2003-2009” showed that 48% of Arabs are dissatisfied with their lives in Israel, 40% do not trust the court system, and 41% support boycotting elections, along with other indications of alienation and discontent.
These are serious problems, and all Israelis should want to know more.
According to Jubran, to overcome the lack of representation, or underrepresentation, of Arabs in national representative samples, pollsters must create samples based on much more intimate knowledge of the population spread and developing interviewing techniques that allow for introductions, explanations and conversation to break the ice – to gain the interviewees’ trust.
The problem won’t be easily solved until those commissioning surveys begin to view Arabs as inseparable from Israeli society, and find the resources to poll them, and poll them properly. In the meantime, editors and reporters could help by reporting more precisely on the sampling methodology, and they could explain the implications. That way, readers would know when to take findings with a grain of salt – or two.
Ultimately, Israelis will have to realize that better polling means better mutual understanding, which we all need.