February 26, 2019
The Jewish Chronicle recently carried a report luridly headlined “Survey Reveals Clear Link Between BDS and Jew-hate”.
In fact, as authors Jamie Stern-Weiner & Alan Maddison make clear, it did no such thing. The trigger response of interpreting antisemitism research in the most sinister possible light is shown here to have been without foundation.
By Jamie Stern-Weiner and Alan Maddison, for JVL
24 February 2019
Anti-Jewish animus is often perceived to be lurking behind seemingly legitimate criticism of the state of Israel. This belief has received intellectual articulation under the conceptual rubric of the ‘new antisemitism’, which purports that hostility to Israel is merely classical antisemitism in new guise. In a recent front-page article headlined ‘Survey Reveals Clear Link Between BDS and Jew-hate’, the Jewish Chronicle proffered empirical support for this enduring suspicion. Alleging that ‘a majority of people in the UK who support boycotts or regard Israel as an apartheid state hold anti-Jewish views’, the Chronicle reported:
No fewer than 58 per cent of those who consider Israel an apartheid state strongly agreed or tended to agree with five antisemitic ideas presented to them, while 52 per cent of boycotters identified with six or more.
A Jewish Telegraph front-page relayed the same findings (headlined: ‘Israel Hate “Is Antisemitism”’) in near-identical words: ‘a majority of people in the UK who support boycotts or regard Israel as an apartheid state also hold anti-Jewish views’.
Misrepresenting the data
Both articles were based on a recent paper by David Graham and Jonathan Boyd, published by the Institute for Jewish Policy Research (JPR) and Community Security Trust (CST), which explored the relationship between antisemitism and hostility to Israel.
But both articles misrepresented the paper’s findings.
Graham and Boyd based their study on a nationally representative survey of 4,005 non-Jewish respondents, gathered between October 2016 and February 2017. They focused their analysis on reactions to two statements:
● ‘Israel is an apartheid state’
● ‘People should boycott Israeli goods and products’
21 percent of respondents strongly or partly agreed that ‘Israel is an apartheid state’ while 9 percent agreed that ‘People should boycott Israeli goods and products’.
The authors then plotted the percentage of respondents in agreement with the ‘apartheid’/‘boycott’ statements against groups of respondents defined by the number of ‘anti-Jewish sentiments’—a list of statements about Jews deemed prejudicial by the pollsters—with which the respondents agreed:
Figure 9 (above) demonstrates a clear positive relationship between the number of ‘anti-Jewish sentiments’ held and the percentage of those in agreement with the ‘apartheid’ and ‘boycott’ statements. But it says nothing about the prevalence of the former among the latter.
Specifically, Figure 9 does not show—per the Chronicle and Telegraph—that ‘58 per cent of those who consider Israel an apartheid state strongly agreed or tended to agree with five antisemitic ideas’. Rather, it shows that of those respondents who harboured five ‘anti-Jewish sentiments’, 58 percent considered Israel an apartheid state. The Chronicle and Telegraph read the chart backwards.
If the Chronicle and Telegraph misinterpreted the report’s findings, however, they cannot bear sole responsibility. For the study’s authors themselves appear determined to draw unwarranted and alarmist conclusions from their data.
Graham and Boyd observe that most British Jews believe supporters of anti-Israel boycotts (and also, they speculate, those who view Israel as an apartheid state) to be antisemites:
[A] 2012 survey . . . found that two out of three British Jews (67 percent) would consider a non-Jewish person who endorsed a boycott of Israeli goods and products to be either ‘probably’ or ‘definitely’ antisemitic.
They conclude, on the basis of their survey data, that this suspicion is ‘scientifically reasonable’:
It is . . . scientifically reasonable to conclude that when such claims are made about Israel by non-Jews, there is a relatively high likelihood that they are being made by someone who is also predisposed towards anti-Jewish feeling, thereby indicating antisemitic feeling, motive or intent. [emphasis added]
Boyd further comments (as quoted by the Jewish Telegraph):
Recently-published data from the European Union Agency for Fundamental Rights demonstrate that most Jews experience these claims as antisemitic, and this analysis indicates that they are often right to make that call. [emphasis added]
Finally, in an article for the Jewish Chronicle, Boyd writes that the ‘fundamental question’ addressed by the JPR-CST study was, ‘Are boycotters of Israel antisemitic?’ His answer: ‘British Jews are right to be cautious’.
In fact, as we show below, the data point to the opposite conclusion: when ‘apartheid’/‘boycott’ claims are made there is a high likelihood that they are being made by someone who does not harbour anti-Jewish feeling; when Jews experience these claims as antisemitic, they are only rarely right to do so.
In short: only a minority of those agreeing with the apartheid/boycott statements are likely to be motivated by anti-Jewish sentiment.
Asking the right question
The JPR-CST’s Figure 9 (above) shows a correlation between ‘anti-Israel’ and ‘anti-Jewish’ sentiments.
But it does not address Boyd’s ‘fundamental question’: What proportion of those who believe Israel is an apartheid state or who support anti-Israel boycotts also harbour anti-Jewish sentiments?
To answer this, we need to know the size of each group along Figure 9’s horizontal axis: that is, how many survey respondents endorsed 0, 1, 2, etc. of the study’s list of purported ‘anti-Jewish sentiments’.
This information is not provided by Graham and Boyd, but is available in a previous publication by Daniel Staetsky based on the same survey data.
By incorporating this data on group sizes into the proportional data displayed in Figure 9, we can calculate what proportion of respondents agreeing with the ‘apartheid’/‘boycott’ statements also harboured one or more ‘anti-Jewish sentiments’.
The result of this procedure is shown below.
Unlike Figure 9, this illustration provides a breakdown of the 21 percent and 9 percent of all respondents who agreed with the ‘apartheid’ and ‘boycott’ statements, respectively, according to the number of ‘anti-Jewish sentiments’ held.
What can we conclude from this?
● More than half of those who agreed that ‘Israel is an apartheid state’ and nearly half of those who endorsed a boycott of Israeli goods endorsed no ‘anti-Jewish sentiments’ whatsoever.
● Only 9 percent (‘apartheid’) and 17 percent (‘boycott’) endorsed 5 or more ‘anti-Jewish sentiments’—the threshold designated by Staetsky as scoring ‘high’ on the antisemitism index.
● 71 percent (‘apartheid’) and 59 percent (‘boycott’) endorsed just one ‘anti-Jewish’ sentiment or none.
● More than 81 percent (‘apartheid’) and 69 percent (‘boycott’) endorsed two ‘anti-Jewish sentiments’ or fewer.
The Chronicle and Telegraph reported that 58 percent of those agreeing that Israel is an apartheid state and 52 percent of those endorsing a boycott of Israeli products harboured five and six ‘anti-Jewish sentiments’ respectively. Our estimates, as shown above, are 5 percent and 10 percent respectively.
When assessing these findings, it ought to be emphasised that endorsement of one or more ‘anti-Jewish sentiments’ need not reflect animus toward Jews. Survey respondents who agreed with just one or two ‘anti-Jewish sentiments’ thereby rejected all the others; Staetsky writes, ‘[i]ndividuals holding just one attitude considered to be antisemitic by Jews . . . cannot be labelled as antisemites in any conventional political, or indeed moral, sense’. A large majority (79 percent) of survey respondents who agreed with at least one ‘antisemitic statement’ also endorsed at least one ‘positive statement’ about Jews, while the proportion endorsing at least one ‘antisemitic statement’ was greater among those with a ‘favourable’ than a ‘neutral’ opinion of Jews. Whereas approximately 30 percent of the British public endorses at least one ‘antisemitic statement’, fewer than 10 percent report a negative opinion of Jews.
Presenting the study to Jewish Chronicle readers, Boyd notes that its findings do ‘not mean that everyone endorsing these ideas [‘apartheid’ and ‘boycotting’] is necessarily an antisemite’. Indeed not — but isn’t this a curious way to report data which suggest that the majority of those endorsing ‘apartheid’ framing and/or anti-Israel boycotts are not antisemites? Boyd continues: ‘the data also indicate that some people who hold these views about Israel exhibit no particular hostility towards Jews at all’. To be sure—but isn’t ‘some’, as used here, a rather reserved synonym for ‘most’?
Graham and Boyd allege that there is a ‘relatively high likelihood’ that a person who characterises Israel as an apartheid state or endorses anti-Israel boycotts is harbouring ‘antisemitic feeling, motive or intent’. Boyd asserts that, if Jews perceive ‘apartheid’/‘boycott’ calls as ‘antisemitic’, they are ‘often’ right.
But their data suggest that, in a large majority of cases, those who agree with ‘apartheid’/‘boycott’ claims reject all or almost all of the ‘anti-Jewish’ statements put to them and are unlikely to be motivated by anti-Jewish animus.
Boyd has elsewhere criticised the ‘Pavlovian’ instinct to interpret antisemitism research in the most sinister possible light. Noting that by the standard metric antisemitism levels in Europe have seen ‘no discernible change’ while favourability toward Jews is rising, Boyd cautioned:
[T]o tackle antisemitism, we need to understand it fully and make sure we don’t mistake fears for facts. We do ourselves and our cause a disservice when we fail to read results like this carefully.
By focusing on the correlation between ‘anti-Israel’ and ‘anti-Jewish’ sentiments rather than the proportion of the former likely to be motivated by the latter; by insinuating without foundation that a large proportion of anti-Israel activists are likely to be antisemites; and by over-stating the extent to which ‘anti-Jewish sentiments’ imply anti-Jewish animus—in short, by making ‘a beeline for the bad news’ and failing ‘to read results . . . carefully’, Boyd and his colleagues have indeed done their readers a disservice.
A 2015 JPR report lamented the ‘hyperbole, bias and conjecture that litter public discourse’ on antisemitism. It is surely cause for regret that the JPR and CST have lent their imprimatur to a report that exacerbates this confusion as it validates the overblown fears of British Jews rather than subjecting them to the cold—but, in this case, reassuring—light of reason.
 After one of this article’s authors contacted the JPR about the error, the Jewish Chronicle report was corrected.
 And not only because—according to the Chronicle—the JPR signed off on its error.