Dr Matthew Goodwin is a lecturer at the University of Nottingham, whose forthcoming book The New British Fascism: Rise of the BNP , will be published by Routledge on May 4th.  Follow him on Twitter  here.

Last week, I wrote an article outlining why Sayeeda Warsi was right to highlight the growing problem of anti-Muslim sentiment in modern Britain. I pointed to evidence of significant public hostility toward Muslims, anxiety over the role of Islam in society, and the absence of a rigorous debate over what is driving these trends, and what they mean for policy and practice. In response, Paul Goodman raised valuable points. He has previously urged Parliament to confront anti-Muslim hatred and violence, highlighted the lack of data on this issue and made the case for why something needs to be done. Our points of disagreement, therefore, are slight. But there is one point that warrants further attention.

If these trends are to be tackled, then the first thing we need is reliable data, both on public attitudes toward Muslims and the attitudes of Muslims themselves. It is partly for this reason Goodman urges us to look beyond the evidence cited in my piece: the British Social Attitudes and (recently cancelled) Citizenship surveys.

Instead, he draws on several polls when taking issue with my assertion that Muslims are not markedly different from the general population in terms of their feelings of belongingness to Britain. One is an ICM poll which suggests two fifths of Muslims want Sharia Law introduced. Then, polls by NOP and Populus indicate three out of every ten Muslims want to live under strict forms of Sharia law, as practiced in Saudi Arabia. These findings, it is argued, should give us pause for thought when making assertions about the attitudes of British Muslims.

And here enters my point: Goodman is right when arguing we need to consult all of the available evidence, but there are also good reasons why any future enquiry into this issue would be mistaken to rely heavily on this type of evidence. Polls provide incredibly useful snapshots of public opinion, and their architects perform an important role. Sometimes, however, they rely on small samples and introduce the risk of over-representing specific types of people, for example Muslims who hold particularly strident views and are keen to air those views to pollsters.

On sensitive issues like religious identity, polls sometimes ask misleading questions and fail to compare groups facing similar situations and grievances. In the aftermath of the Danish cartoons, for instance, Muslims were asked lots of questions about their attitudes toward freedom of speech. Their responses, however, were then often compared with members of the general population who were not experiencing similar insults to their religion. Why not compare their responses with those of devout Christians after similar faith-based insults, such as the furore that accompanied Jerry Springer the Opera? Also, it is often unclear in commercial polls how respondents understand questions to which they are responding.

But there is a more fundamental point here. Implicit in arguments that jump on polls like those above is a popular though misleading assertion that – for Muslims – religious identity somehow represents an alternative to (in this case British) citizenship. The assumption is that Muslims who endorse various aspects of their faith are choosing this religious identity as an alternative to their national identity. However, as the academic Maria Sobolewska demonstrates in her chapter in The New Extremism in 21st Century Britain, religious identity actually appears to have little to do with Muslims’ sense of alienation, or their supposed ambivalence toward the British state and society. Religiosity among Muslims neither exacerbates their sense of alienation, nor acts as some kind of "replacement identity".

Analysing the attitudes of over 1,700 Muslim respondents to the Citizenship Survey (which employed more established and consistent measures than the temporal and often vague measures in polls), she paints a different picture. Contrary to the popular view, most Muslims happily agreed Islam is the most important thing in their life, AND that their primary loyalty rests with the British state, and that they belong to Islam AND Britain. For most, belonging to Islam and belonging to Britain was not a mutually exclusive choice.

Nor was there much evidence that more religious Muslims were more excluded or alienated than those who said they were less so. Even polls support this point (although sadly their findings are often ignored in tabloid media). For example, upwards of 90 per cent of Muslims say Islam is very important to them (e.g. NOP/C4 2006, Populus/Policy Exchange 2006), while similar percentages openly declare their loyalty to Britain.
To underline the point, Sobolewska compares the views of Muslims on three widely accepted indicators of political exclusion and support for democracy: (1) their trust in institutions like councils, parliament and police; (2) feelings of belonging to Britain; and (3) whether they feel they can influence decisions affecting the country. Her findings? Muslims were actually more trustful of institutions than other religious groups – including Christians. On belonging to Britain, they showed no great difference to the Christian and mainly white majority. In fact, Muslims were more likely than other minority religious groups and also the non-religious to feel that they belong to Britain. And lastly, Muslims were the least likely group to say they did not have any influence over decisions that affect the country. The overall conclusion was that Muslims appear far more integrated and attached than popular claims would have us believe, and there is little evidence they are turning to religion as an alternative identity.

To return to Goodman, he is absolutely right that any attempt to promote cohesive local communities will struggle to resolve grievances that are fuelled by events in the international arena. But he also sidesteps my latter point that not only were some cohesion-based initiatives moving in the right direction, but they were also backed up by decades of academic research. Unfortunately, many have since been cancelled or starved of funding. We all share a consensus that these trends are too important to be left for extremists to manipulate. But it is also paramount that debate over their causes and consequences remains firmly anchored in the wider evidence base.