The Muslim faith and public opinion: the polling evidence

In this article I don’t propose to look at the Quran, upon which I am no expert although I have read sizeable chunks of it. I merely intend to look at the polling evidence on this subject. Does this evidence support or oppose the claims that ‘Muslims’ pose a threat to the UK?

There are two substantial pieces of work on this subject which I’ll consider in this article: an ICM poll for Channel 4 published in April 2016, and a ComRes poll for Radio 4 Today published in February 2015.

There are a few factors to bear in mind before looking at the data:

1. Polling of Muslims is incredibly difficult in the United Kingdom, and therefore potentially biased.

Roughly 5% of the population identify as Muslim. A standard opinion poll sample is about 1,000 people. If a polling company were to contact people at random, therefore, it would have to contact 20,000 people in order to find 1,000 Muslims. In practice, it would be far more because not everybody is happy to take part in opinion polls.

Finding a random sample of 1,000 Muslims is very difficult, and therefore a simple workaround is to only poll in those areas where at least 20% of the population are Muslims. This skews the results of a poll: it’s possible, for example, that those who hold extremist views might prefer to segregate, and therefore be more likely to live in communities where there are more other Muslims. Nevertheless, the polling challenges are what they are. In the case of ComRes, the sample also included some people who had previously identified as Muslim in previous surveys and were happy to be re-interviewed.

2. Without a control group, it’s harder to draw conclusions

If a survey were to find that “15% of British Muslims believe” then it would be meaningless unless we’re able to put it into context. Perhaps 20% of the general population might hold those beliefs, or maybe it’s only 2%. For the results to mean anything, it’s important to ask the same questions to a representative sample of the general population. That way it’s possible to make informed comparisons. If the samples find “15% of British Muslims believe X, but only 2% of the whole population believe X” then we’d be able to tell the difference.

The ICM / Channel 4 study had such a control group, and is therefore much more useful for the purpose of understanding what Muslims believe compared with the general population. The phrasing of a question can subtly affect responses.

3. Have the questions been answered honestly by the respondents?

A brief ‘sanity check’ is worthwhile. Given that some of the far-right claim Muslims generally engage in ‘taqiyya’ (a doctrine which enables Muslims whose life is under threat to be dishonest about their true beliefs in order to save their lives) we should take this claim seriously and look for evidence of this either way in the data. Is there any evidence that people are ‘telling an opinion poll company what they want to hear’?

4. A small difference between Muslim opinion and the control group opinion could just be random statistical noise.

Imagine that you toss a coin 1,000 times. You won’t necessarily get exactly 500 heads and 500 tails, even if the coin is completely fair. You might get 508 heads and 492 tails, for example. Broadly speaking you’re likely to be somewhere in the 470-530 region 95% of the time.

If you repeated the coin experiment twenty times, then on average you’d find that one of those twenty times you’d get a result that’s outside these parameters.

What does that mean for polling? Well, if you’ve chosen people at random, you’re likely to be within 3% of the true answer (as long as your sample isn’t biased – see point 1) the vast majority of the time. But if your control sample says “65% of Muslims believe Y” and “67% of the whole population believe Y”, how can we be sure there’s actually a difference? It could be that there’s no difference at all. Small differences may be no difference at all.

So, moving on to the surveys of opinion:


This survey was of precisely 1,000 British Muslims.

Without an adequate control group, this poll must be treated with a certain pinch of salt. 95% of respondents feel a loyalty to Britain (although when the question is asked the other way around, 6% feel a disloyalty to Britain). Often samples don’t add up to exactly 100% due to rounding off errors. They’ll sum to 99% or 101%. In this case though, it’s two separate questions.

‘Feel loyalty to Britain’ is 95% to 4% (1% don’t know)

‘Feel disloyalty to Britain’ is 6% to 92% (2% don’t know)

Presumably a small number of people are conflicted, and feel both a loyalty and a disloyalty to Britain at the same time.

Is this statistic worrying? To take the ‘worse’ of the two, 6% of British Muslims feel disloyal to the country in which they live. Again, the true figure is likely to be lower due to the way the sample was obtained. But how many non-Muslims feel the same way? If we don’t know, the statistic is meaningless.

Likewise, 20% said that ‘Western liberal society can never be compatible with Islam’ and 72% disagreed. It’s a poor question: it’s possible to object to ‘liberal’ but not to ‘Western’. Some Christians might answer an equivalent question in the same way: those who are anti-abortion, anti-gay marriage, anti-pornography, etc., might say that the society is not compatible with their version of Christianity.

93% agreed that ‘Muslims in Britain should always obey British laws’ (6% disagreed).

In this article I’m going to skip over some of the irrelevant questions to this piece. How accepted Muslims feel they are in British society, or how relevant they see the Muslim Council of Britain, are of interest – but they’re not specifically relevant to this article, which is already far too long.

78% of Muslims consider publishing pictures of the prophet Mohammed to be personally offensive; this is specifically banned under Islam. The statistic is unconcerning (for example, if churchgoing Christians were asked how they feel about depictions of Jesus which contradict the Bible, they would express offence too – see for example the outrage over Jerry Springer: The Opera) in and of itself. 68% believe, however, that acts of violence against those who publish such pictures can never be justified.

The concerning figure is that 11% said that ‘organisations which publish images of the Prophet Mohammed deserve to be attacked’, and that 27% have ‘some sympathy behind the motives’ for the attacks on Charlie Hebdo.

There is a huge difference, of course, between saying that someone ‘deserves’ something to happen and actually doing it.

In a different context, roughly half of the British people support the death penalty. Some of those who oppose the death penalty might believe that murderers deserve the death penalty, yet still oppose it for other reasons (like the risk of wrongful convictions).

The phrase ‘Those who murder innocent children deserve to die’ would likely elicit a very high level of agreement in an opinion poll, but the number who would actually be prepared to kill them would be far lower.

Here’s where the opinion becomes less clear: many Muslims would consider that such actions ought to be criminal. They would believe a legal consequence to such actions is deserved, but that is very different from taking the law into your own hands.

With regard to integration, just 13% say they ‘prefer’ to socialise with Muslims rather than non-Muslims.

82% considered themselves to be ‘practising’ Muslims (the usual caveat here: the true figure could be lower due to the way that the data was collected).

Just 8% said that they ‘know people’ who feel ‘strongly sympathetic’ towards those fighting for ISIS or Al-Qaeda.

94% said that if they knew someone in the Muslim community were planning an act of violence, they would report them to the police. Indeed, many terror plots have been foiled for precisely this reason.

The most worrying statement is this: ‘Muslim clerics who preach that violence against the West can be justified are out of touch with mainstream Muslim opinion’ was only shared by 49% (and opposed by 45%).

The phrase ‘mainstream Muslim opinion’ is a difficult one to assess. Does it mean ‘in the UK’? If it is taken in that way, it’s more concerning. Does it include both Sunni and Shia? Does it consider Salafism, for example, to be ‘mainstream Muslim opinion’? A Deobandi Sunni Muslim might well answer in the affirmative to that question if they considered Salafist philosophy to be within the mainstream, yet they (and everyone they know) oppose it.

Nevertheless, on that particular question I do have a concern over the result.


The sample size for the Muslim poll was 1081, and the control group was 1008.

This is a substantial piece of work. It is a 615-page document, which I have read in full, but for length I’ll be selective.

1. It shows that 86% of British Muslims identify as British. Higher than the 83% for the population as a whole (this is where we see the value of having the control group).  This is also higher than their identification with their own country of birth (79% – which may be the UK in some cases anyway) or with their parents’ country of birth (73%).

The way this question was asked is likely to elicit a lower response than the ComRes ‘loyalty to Britain’ question, because it provided other options such as ‘loyalty to your local area’ and ‘loyalty to your parents’ country of birth’. It was possible to select more than one answer, but easier to choose the other options than to admit ‘disloyalty’ to a country.

Nevertheless, Muslims in Britain saw themselves as more British than the control group. However, the difference is only 3%. Therefore, statistically, there is insufficient evidence to draw an inference that there is any difference between the two – it could be mere random variation.

2. It’s worth noting that on various questions, the control group were in fact more outspoken than the Muslim group. The general population were more likely to say that anti-Muslim prejudice has increased (61% in control group, 40% in Muslim group), and less likely to feel that they can impact on decisions affecting their local area (43% in Muslim group compared with 35% in the control group).

3. Attitudes towards homosexuality show a marked difference between the Muslim group and the control group. Indeed, this is the biggest difference in opinion.

28% of the control group considered it acceptable for a ‘homosexual person’ to be a teacher in a school, compared with 75% of the general population.

It appears that there is a softening over time of attitudes within the Muslim community on such issues.

7% of the control group strongly disagreed that homosexuality should be legal, so such views are not unheard-of in the wider population.

However, within the Muslim community the figures were much higher. 47% of those not born in Britain strongly disagreed that homosexuality should be legal. 27% of those born in Britain shared that view.

So, we see a very clear decline and move towards integration with those born in this country.

4. The ‘integration’ question: just 1% said they wanted to be subject to Sharia Law and government. 49% wanted to integrate in all aspects of life, 29% wanted to integrate on ‘most things’ whilst retaining a separate identity and 17% wanted to only partially integrate.

There is, however, greater support (23%) for Sharia law in ‘some areas’. This declines amongst those born in Britain compared with those not born in Britain (19% to 27%).

5. 7% of British Muslims support the establishment of ‘a caliphate’. As it’s possible to support the principle yet oppose ISIS, and regardless there’s no suggestion of seeking to create a caliphate HERE, it would seem that the numbers who hold to that particular worldview are likely to be an order of magnitude lower than 7% of British Muslims. It is worth noting that 2% of the control group also held to this view.

6. 11% of the general population, but only 6% of Muslims, ‘sympathised’ with the organisation of radical groups.

Again here, ‘radical groups’ are undefined. There exists a level of concern over this response, but it is a low level of concern.

It’s worth noting that on a more defined question – whether committing minor crime as part of a protest is acceptable – both the Muslim group and the control group showed exactly 18% agreeing that it is.

1% of British Muslims (compared with less than 1% of the whole population) said that they ‘completely sympathise’ with making threats of terrorist actions as part of political protest, and a further 5% ‘sympathise to some extent’ (compared with 2% of the whole population).

7. 4% of British Muslims said that they would ‘sympathise’ with terrorist actions (compared with 1% of the population as a whole). Less than 1% said they would ‘completely sympathise’ (identical to the control group).

Any figure greater than zero is by definition troubling. However, it is important to put this into some kind of context. During the Troubles in Northern Ireland, 26% would ‘sympathise’ with Republican terrorism and 27% with loyalist terrorism. That’s of the whole population rather than of just a small subset of the population, too.

8. Throughout the poll, the figures for those born in Britain are generally more integrated than those not born in Britain. This suggests that as each generation goes by, integration is likely to increase rather than decrease.

9. Muslims are less likely (44% to 60%) than the control group to support violence to protect their family, but more likely (24% to 7%) to support violence organised by groups to protect their own religion. However, it is worth noting that 83% of the Muslim group are practising Muslims whereas a much smaller percentage of the control group are regular worshippers, and that control group by definition includes those who are non-religious.

Muslims are also slightly more likely (22% to 17%) to support violence against the police to fight injustice, and against governments 920% to 16%) to fight injustice.

10. Muslims generally (64% to 17%) support the right of Muslim girls to wear a niqab to school; the control group (37% to 47%) generally oppose.

11. 31% of Muslims (9% of the control group) believe it is acceptable for a Muslim man in the UK to have more than one wife.

12. Likewise, 39% of Muslims (5% of the control group) believe that wives should obey their husbands. Once again, there is a generational difference: the figure for those Muslims aged 65+, for example, is 56%.

13. Muslims generally feel more positively about Catholics and Protestants (69.7 to 64.8 and 69.7 to 68.2) – results are for a mean score out of 100 – than the control group. The ‘thermometer’ methodology makes it harder to assess which differences are statistically significant and which are not.

14. They feel less warmly towards those who are not religious (60.7 to 69.1) than the control group.

15. They feel less warmly towards the Jewish community (57.1 to 63.7) than the control group.

16. Muslims are less likely (26% compared with 46% in the control group) to consider that anti-Semitism is a problem in the UK.

17. Muslims are more likely (35% compared with 9% in the control group) to consider that Jews have too much power in the UK, (31% to 7%) that Jews have too much power over the government, and (39% to 10%) that Jews have too much power over the media.

These differences are also reflected in a string of other questions.

18. 2% of the Muslim group (less than 1% of the control group) ‘completely sympathise’ with stoning those who commit adultery. A further 3% ‘sympathise to some extent’ (2% in the control group).

19. Overwhelmingly British Muslims think that they’d be treated the same as everyone else in the UK (doctors: 91% say that, 3% think they’re treated better and 3% that they’re treated worse. Schools: 88% say that, 3% think they’re treated better and 3% that they’re treated worse, etc.) – a good sign for future integration. Even with legal system and police, the figures are still overwhelming though slightly lower as would also reflect the general population.


a) There is no concern that polls are not being answered accurately. I mention this only for the sake of completeness as it should be self-evident, but if there were any attempt to mislead or subvert, the clear dividing lines on homosexuality (amongst others) would not be there.

b) The samples may be unrepresentative of all Muslims, but if it they are skewed, it’s more likely to be in the direction of finding Muslims in the UK to be less integrated than they are.

c) The Muslim community in the UK is gradually becoming more integrated over time.

d) Those who support terrorist actions are a tiny minority of all Muslims.

e) The Muslim community seems to contain proportionally far more people who oppose homosexuality and subscribe to views which are anti-Semitic than the general population.

f) The overwhelming majority of British Muslims consider themselves to be British, and would report anyone they knew to be planning a terrorist attack.

g) It is often pointed out that even a small percentage of the Muslim community still equates to a large number of people. Such calculations still lead to overestimates (in addition to the sampling issues mentioned, there is a huge difference between sympathising with motives and actually seeking to carry out a terrorist attack).

h) The need to work with the Muslim community as a whole is clearly indicated, and cannot be done with an ‘us versus them’ mindset.

i) Suggestions from those who hold to anti-Islam beliefs that we as a nation are ‘at war’ with Islam, or that our way of life is under significant threat, are simply nonsense.

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *