A researcher takes issue with the idea that one can assess the strength of rightwing extremism by polling attitudes of its members.
In the Dark Ages of the 1970s and 1980s we knew who our fascist enemies were – we judged groups like the National Front and the National Party in terms of their deeds and made sure that they ‘did not pass’. Today we are being asked to know them by their words garnered through opinion polls. It might be ok to carry out a poll before an election to assess the voting strength of the extreme right but, now, in 2011, think tanks are using polls as a basis for analysing a social problem and suggesting socio-political remedies.
Both the Searchlight Educational Trust (working with Populus in April) and Demos (using a self-completion questionnaire on the English Defence League’s Facebook page) end up ‘discovering’ a section of English Defence League supporters to be won back to decency. Of course one should never put individuals beyond redemption, but since everything is wrong about the methodology of such studies, the results should not be taken too seriously.
Attitude surveys, especially on emotive issues such as ‘race’, were already being discounted as unreliable in the late 1960s, not least by the Institute of Race Relations after it carried out its significant Nuffield-funded Survey. People do not tell the truth about their attitudes; they are swayed by the way questions are formulated, the expectations of the question setter, the alternative choices, the order of questions etc. Opinion polls reduce what can be complex and nuanced views into neat and convenient categories; they may reflect the immediacy of current events rather than long-standing changes in public consciousness; they push issues to the forefront of a respondent’s mind when they might not otherwise have been; they may reflect more the perspectives of the pollsters rather than the respondent as the ways of understanding a particular issue; questions can be loaded and they are vulnerable to weighting (ie, placing emphasis on one set of answers over another). And when such a survey deals with immigration or Islam, at a time when the government and the media have been the attack dogs on both fronts, results will be a foregone conclusion.
And, in any event, what do attitudes matter? It is the acting out of those attitudes in discriminatory and bigoted behaviour that affects other people.
But there are other methodological issues, stemming seemingly, from the researchers’ predilections. The first study (endorsed in the Guardian by David Miliband), divided respondents into different tribes, according to a politics of identity, finding two tribes, ‘Cultural Integrationists’ and ‘Identity Ambivalents’ whose members need ‘saving’ by Tories and Labour respectively. But these are not scientific categories, but conjured up by the authors themselves to explain what they already found. Jamie Bartlett (Demos) goes one better and finds that a section of the EDL membership is democratic, not that openly racist and defines its opposition to Islam in an acceptable way. But if one asks questions about democratic rights of an organisation ostensibly set up to further such rights – as against repressive authoritarian Islam – what answers would one expect? And to say that people’s racial attitudes are not extreme because they mirror those of the mainstream – when that mainstream has already been tainted by the far Right – is tautologous to say the least.
We don’t need polls to tell us that there is a danger: just look at the number of attacks being carried out against Muslim communities.
Read an IRR News story by A. Sivanandan: ‘Fallacies and policies: the “Fear and HOPE” report’
Read an IRR News story by Jon Burnett: ‘Searchlight: polling the “new politics of identity”‘
Read an IRR News story by Ryan Erfani-Ghettani: ‘Taking Facebook at face value?’