Network Norwich and Norfolk > People > James Knight > What the pollsters forgot to tell you

What the pollsters forgot to tell you

JamesKnight300Regular Network Norfolk columnist James Knight looks beyond the headlines created by a recent poll on Christianity in the UK to highlight what happens when statistics are taken at face value.


The recent poll carried out by Ipsos MORI for the Richard Dawkins Foundation for Reason and Science concerning the extent to which people are Christians has grabbed the headlines recently, with many Christian commentators or apologists responding with a retort. My main inclination to comment has only been in relation to what I'm now going to say in this article. The main area of focus has been the following statistic:


"At the time of the 2011 Census, just over half (54%) the public thought of themselves as Christian, compared with almost three-quarters (72%) in the 2001 Census."


Polls are said to be questions that tell us how people feel about certain issues, but usually they are not. What they actually are is a set of questions designed to reflect the outcome that the poll commissioners wish to see. If you want to convince the public that they feel a certain way about something, all you have to do is frontload the questions to direct people towards the answers you wish to see. 


Let me offer an example. Imagine you're the Prime Minster and you want to bring back compulsory national service - you may want to mobilise military troops in a particular county beset by civil war, or you may want to get people off unemployment benefits, or something of that kind. What you'd like is a referendum that shows the nation is behind you - after all, if the majority of the electorate are seen to be backing you then your decision to bring back national service is not much of a political risk. The poll to see if people in favour is really a poll that you want to show people 'are' in favour.  It might look something like this:


Politician: Sir, are you worried about the rise in crime among teenagers?


Person being surveyed: Yes.


Politician: Do you think there is lack of discipline and vigorous training in our Comprehensive Schools?


Person being surveyed: Yes.


Politician: Do you think young people welcome some structure and leadership in their lives?


Person being surveyed: Yes.


Politician: Do they respond to a challenge?


Person being surveyed: Yes.


Politician: Might you be in favour of reintroducing National Service?


Person being surveyed: Possibly.


Politician: Yes or no?


Person being surveyed: Yes.


VoteLogoNow say you wanted to reach the opposite conclusion - by altering the question inversely you could get the survey to reflect a ‘no’ majority by frontloading the questions differently. For example:


Politician: Sir, are you worried about the danger of war, the unrest in Middle East, North Korea and Islamism?


Person being surveyed: Yes.


Politician: Are you unhappy about the growth of armaments and nuclear non-proliferation?


Person being surveyed: Yes.


Politician: Do you think there's a danger in giving young people guns and bombs and teaching them how to kill?


Person being surveyed: Yes.


Politician: Do you think it's wrong to force people to take arms against their will?


Person being surveyed: Yes.


Politician: Would you oppose the reintroduction of conscription?


Person being surveyed: Yes.


What you’ve seen in the above illustration is the same people being surveyed, but a very different consensus based on the direction of the questioning. It's very rare that polls are conducted because commissioners are interested in people's opinions. Whenever you see a poll, expect that whoever requested its commissioning is looking to prove a point they had already preconceived. I expect the Richard Dawkins Foundation for Reason and Science didn't merely have a curiosity regarding people's religious beliefs - they wanted to show a nation that is losing its faith and becoming more secular as times goes on.  Remember one thing though, proponents of secularism need not be unbelievers - many Christians are secular politically in that they want church and state politics kept separate. Given the foregoing, these following Ipsos MORI conclusions are not as revealing as the Richard Dawkins Foundation would have you believe:


1) Three quarters (74%) strongly agree or tend to agree that religion should not have special influence on public policy, with only one in eight (12%) thinking that it should. 


2) More oppose than support the idea of the UK having an official state religion, with nearly half (46%) against and only a third (32%) in favour. 


3) There is overwhelming support for religion being a private, not public, matter. Asked how strongly they support the statement that governments should not interfere in religion, 79% strongly agree or tend to agree, with only 8% strongly disagreeing or tending to disagree. 


I'm a Christian, but I am largely in agreement with the above statements. The same is true of religious education - even amongst Christians there are opposing views on the merit of faith schools, with many Christians preferring greater scholastic diversity and a more eclectic mix of pupils.  That said, one disturbing result was that even though "More Christians oppose (38%) than support (31%) the teaching of 6-day creationism in state-funded school science lessons." - it is very worrying that nearly a third of Christians supposedly espouse the teaching of pseudoscientific nonsense in schools.


Declining church attendance gets a lot of attention, but as I've said before, that's only really happening in CofE churches - the Pentecostal churches are significantly increasing their congregations with an impressive externally focused church ethos. 


Next we have this from the Ipsos MORI poll - "When asked where they seek most guidance in questions of right and wrong, only one in ten (10%) said it was from religious teachings or beliefs, with over half (54%) preferring to draw on their own inner moral sense." This is the sort of result that I think misleads, and here's why. The question was "When it comes to right and wrong, which of the following, if any, do you most (stress most, my emphasis) look to for guidance?" The options were:


1) My own inner moral sense


2) Parents, family or friends


3) Religious teachings and beliefs


4) Philosophy and reason


It is fairly obvious that these four options are not mutually exclusive – they are perfectly compatible, and even Christians ought to feel that their inner moral sense is what drives their pursuit of rectitude – after all, most Christians probably believe that God gave them the moral sense with which to see God in Jesus.  Without an inner moral sense we wouldn’t be able to see Jesus’ morality as being the most God-like the world has seen.  Therefore it is either disingenuous or naive to set them up mutually exclusively in a multiple choice scenario.


The upshot of all this is that most polls are frontloaded, most statistics only cover the parts of the issue that give exhibition to the argument the statisticians want to express, and most belief systems are far too complex and multi-dimensional to be reduced to a few off-the-peg clichés and succinct phases. 


I am not saying that Christian belief isn’t on the decline in the UK – it may well be going through a period of diminution.  What I am saying is, the Richard Dawkins Foundation poll commissioned by Ipsos MORI doesn’t carry the weight that they claim – not just because the sample space isn’t broad enough, but mainly because the questions do not give enough credence to the complexity of the answers. 


What is likely to happen is that if the new-wave atheists believe that they are singing from the same hymn sheet (pun intended) as most of the UK you may see in the future their brand of strident secularism becoming more and more like a militant religion.  The response to a wavering Christianity isn’t a strident atheism, it is a better, more love and grace imbued Christianity. One might argue that the country needs that kind of love and grace spread about fairly urgently. 



The views carried here are those of the author, not of Network Norwich and Norfolk, and are intended to stimulate constructive debate between website users. We welcome your thoughts and comments, posted below, upon the ideas expressed here. You can also contact the author direct at  

James is a Christian writer and local government officer based in Norwich. 
You can access his current collections of columns here 

Meanwhile, if you want to find out more about Christianity, visit:


Paula Kirby 22/02/2012 21:14
I write as the person at RDFRS UK who worked closely with Ipsos MORI throughout the commissioning of this poll, and I would like to correct a few misunderstandings in your article, James.

First, this was not a Richard Dawkins Foundation poll commissioned by Ipsos MORI, as you state in your penultimate paragraph. It was an Ipsos MORI poll, commissioned by the Richard Dawkins Foundation. That's a very important difference!

Ipsos MORI are not only one of the most highly respected names in opinion polling in their own right, they are obliged to follow the highest professional standards, in the setting of the questions, the selection of the sample respondents, and the analysis and presentation of the responses.

First: the design of the questionnaire. This survey was handled by Ipsos MORI's specialist Race, Faith and Cohesion Research team, who - as you might expect from their name - specialise in this area of research, understand their subject matter well, and are used to designing questionnaires that are appropriate to it.

All questions were scrutinised carefully to ensure they conformed with strict professional standards and, in particular, to ensure that they were fair, clear and not leading. In addition to being put together with great care in association with the Race, Faith and Cohesion Research team, it is Ipsos MORI's standard policy that all surveys have to be approved by an independent senior panel within the company, to make absolutely certain that the highest standards have been complied with.

Secondly, the sample. The survey was carried out using Ipsos MORI's face-to-face Capibus methodology (, to ensure that the sample size was large enough to be statistically reliable, and that the people polled were genuinely representative of the nation as a whole. This is Ipsos MORI's bread and butter, and is based on a deep understanding of statistics. I'm afraid you cannot simply dismiss it. This is how the best opinion polling is done. The headline figure of 54% having ticked the Christian box in last year's census has a 95% probability of being correct to within 2% either side. The accuracy of the remaining data fluctuates from question to question, because of different base sizes, but it all has a 95% probability of being correct to within 2-3% either side. This is the norm in national opinion polling: the standard to which all properly conducted surveys are performed.

Thirdly, the results. Ipsos MORI themselves produce the data tables from their results, and they have been made public, both by us and by Ipsos MORI themselves ( No one needs to depend on our (or the media's!) statements about them.

As with all their clients, Ipsos MORI vet everything we say about the poll and ensure that nothing we say goes beyond what is genuinely shown by the data. Unfortunately, they can't do the same with others who comment on it! For example, Bishop Michael Nazir-Ali, speaking on Newsnight last week, took the figure that showed that 49% of people who ticked the Christian box in last year's census do NOT believe that Jesus is the Son of God, and then claimed that meant that 51% DID. The actual figure for those who DID was 44%, with the difference consisting of those who selected 'Don't know' or 'Prefer not to say', which is rather different from the impression given by the retired bishop. Ipsos MORI drill their clients very carefully to ensure we do not make similar misrepresentations. (Actually, the bishop made the further error of thinking our data showed that 49% of the POPULATION believed Jesus is the Son of God; whereas in fact ALL our results, other than the headline figure of 54% having ticked the Christian box, relate ONLY to people who recorded themselves as Christian in the census.)

Something else that I want to clarify. Despite the way the media have presented the story, nothing in our press releases (see made any comment about people 'not being real Christians' or anything at all along those lines.

The poll was commissioned because, ever since the 2001 census seemed to suggest that 72% of the population were Christian, that figure has been used in some quarters to claim that Britain is a Christian country and that there is therefore widespread support for religion having special influence in public life - eg. reserved seats for bishops in the House of Lords, hospital chaplains being paid for by the NHS rather than churches, obligatory 'broadly Christian' worship in state schools in England and Wales, etc; and also to suggest widespread opposition to liberal social policies such as legalised abortion, gay marriage, etc. In other words, the 2001 census 'Christian' figure was used by those with a narrow, socially conservative Christian agenda to give the impression that there was widespread support in the nation for their policies.

Many of us (including many Christians) suspected that they were unjustified in doing so, and that many UK Christians probably did not share their attitudes to these social issues, or their devotion to the bible, which Christian lobbyists openly state underpins their attitudes. However, suspecting is one thing: having evidence to show whether or not that hunch is correct is quite another.

So this poll was conducted as close as possible to the 2011 census, to explore what those who ticked the Christian box this time REALLY think. And the results are clear. The overall number of people who ticked the Christian box has dropped dramatically from 2001 and is now probably in the order of 54% (down from 72% in 2001); and when asked why they had been recorded as Christian in the census, 2/3 of THEM gave reasons which had nothing to do with personal Christian belief. I won't repeat the core findings here: they're summarised in the press releases on, and you'll find the underlying data via a link there too, or using the Ispso-MORI link I posted above. But the overwhelming picture is of a group of people who, by and large, rarely go to church, rarely pray, rarely read the bible, do not derive most guidance on matters of right and wrong from religious teachings, and less than half of whom do not believe Jesus is the Son of God. If you look at what they mean by 'God' in the first place, you'll see that significant numbers of them don't even define God in specifically Christian terms.

Does this mean they're 'not really Christian'? At RDFRS UK, we're not very interested in defining what Christians must believe. That's up to you (and the existence of 34,000 different Christian denominations suggests there's no ready agreement, even among yourselves), and up to any individual: if someone thinks of themselves as Christian despite just thinking Jesus was a good man, wasn't the Son of God, wasn't physically resurrected, etc etc, that's up to them. That said, there certainly are still churches who would not accept such a loose definition of 'Christian' (take a look here, for instance - but be warned, they take the bible SO seriously, that they take their website down on the 'Sabbath' - But again, that's up to them too. Our interest is not in defining 'Christian'.

What we CAN say, though, is that the majority of people who ticked the Christian box in the census this time around certainly do not hold the kind of beliefs that motivate those lobbyists campaigning for bible-based social policy and a greater voice for religion in public life. And that any attempts to claim in future that such policies have the support of the nation because 'the Census shows we're a Christian country' will be totally unjustified.

But we didn't just ASSUME that low levels of literal Christian belief and low levels of Christian practice must mean opposition to the policies demanded by organisations like Christian Concern, for instance. We asked. And again, the questions were designed by Ipsos MORI to ensure that they were not leading or biased.

And here the results were even MORE clear. Quite apart from the specifics, with more 'Census-Christians' supporting than opposing legalised abortion, legalised assisted suicide, equal rights for gays in all areas of their lives, etc etc, on the GENERAL issue of the role religion should play in public policy, 74% said that religion was a private matter and should not have special influence in public policy; and only 2% thought that religious beliefs should exempt someone from having to comply with the law.

So what emerges is very clearly that most UK Christians, as recorded in the 2011 census, support neither the specific policies agitated for by Christian lobby groups, nor the general principle that religion should have a special voice in public policy. (Eric Pickles, Baroness Warsi and David Cameron, I do hope you're taking note.)

Finally, just two more brief clarifications in what has already been an overlong screed.

The first is that no one should confuse secularism with atheism. Theism is the belief in a god or gods, atheism is the lack of belief in a god or gods, and secularism is the view that the state should be NEUTRAL in the matter of belief in a god or gods. We are not pushing for state atheism, however willing we might be to challenge religious claims. We - along with other secularists, including, as you very rightly and importantly point out, many who are themselves religious - simply see state neutrality in matters of religious belief as the only fair and democratic approach, given that so many UK citizens now follow non-Christian religions or, increasingly, no religion at all. It is very encouraging that our poll suggests strong support for that principle, even among those who ticked the Christian box in the last census.

The second is that no one is trying to silence the religious in general, much less Christians in particular. In a free and democratic country it is important that as many people as possible speak up and have their views heard, and we would encourage everyone to do that, whatever their stance on religion. The issue is that, at present, religion (and especially Christianity) has, not just influence, but SPECIAL influence: through the 26 reserved seats in the House of Lords, for instance, through the automatic special deference that is shown to the professional religious, and, of course, through this government's avowed intention to 'do God'. What we want to see is no one being either privileged or disadvantaged, purely on the basis of their religion or lack of it. If that is 'strident' or 'aggressive' , then 74% of UK Christians are 'strident' and 'aggressive' too. But of course, it is not. It is simply fair. And our Ipsos MORI poll - which was conducted to precisely the same high, impartial and objective standards of every other Ipsos MORI poll - shows that most Christians agree with us about that.

Paula Kirby 23/02/2012 15:45
Oops. Where I wrote in my post above, "and less than half of whom do not believe Jesus is the Son of God", I should have written, "and less than half of whom believe Jesus is the Son of God", of course. (The stats are that 44% of 'Census Christians' DO believe it and 49% of them don't.) The original 'do not' in that sentence was a leftover from an earlier form of wording that I missed when checking.
James Knight 23/02/2012 19:42
Thanks for the reply Paula.

On your first point, yes, apologies, I had a concentration slip at the end - looking back I meant to type 'conducted' by Ipsos MORI not 'commissioned'.

But that aside, you haven't "corrected a few misunderstandings in my article" - you've only really made a few unrelated points - some of which I do actually agree with, as do many Christians I know. A particularly relevant point is our mutual discomfort with the socially conservative Christian agenda in giving the impression that there was widespread support in the nation for their policies.

But my criticisms of the poll and the inability of any group to package this subject into a few succinct customary phrases is something that is never addressed. Polls are said to be questions that tell us how people feel about certain issues, but when the complexity of those issues is not addressed properly there is a distorted conclusion.

This also is a very salient point - I said...

"It's very rare that polls are conducted because commissioners are interested in people's opinions. Whenever you see a poll, expect that whoever requested its commissioning is looking to prove a point they had already preconceived. I expect the Richard Dawkins Foundation for Reason and Science didn't merely have a curiosity regarding people's religious beliefs - they wanted to show a nation that is losing its faith and becoming more secular as times goes on. "

It is stretching the imagination for any member of RDFRS to suggest they weren't guilty of this. It is perfectly fine to attempt to circumscribe the socially conservative Christian agenda when it becomes too much of a strident force. But if specious methods are used then the criticism has little merit.

And I would remind you again that virtually every question (as indicated by my drawing attention to the morality question) is just far too simplistically administered to do any kind of justice to the philosophical gravitas of the subject of the Christian faith, let alone its influence in the broader societal context. This is my main problem with the poll. It's all very well trying to subvert the Christian influence with recourse to statistics - but if those statistics are the result of questions that demonstrate a hopelessly inadequate understanding of faith, its complexities, and the vast philosophical mileage one has to cover to get to the heart of the issues, then the route by which the RDFRS reached their conclusions about the general feeling in this country is fraught with misjudgements.

In the world of collecting statistics based on opinions (as opposed to facts) it is very difficult to construct a reliable and meaningful index that is not biased in one way or another. If you are interested in statistics, Paula, I have a potentially significant one that I cannot get atheists to tackle. Almost all of the new-wave atheists - the sort who identify themselves with RDFRS - call themselves former Christians.

If you want a really significant poll - rather than trying to stretch around a diverse country and find whether many are the kind of Christians that can be easily stereotyped, you should do a poll consisting only of RDFRS affiliates, and then tackle the question of why about 95% of them (I'd forecast, based on experience) are ex-churched.

Do I assume correctly that you are the Paula Kirby from the Washington Post? You see, part of my deliberations have been on the subject of why new-wave atheists have such a bee in their bonnet about the badness of faith - and it mainly seems to boil down to some sot of catharsis to do with their Promethean escape from the clutches of their church history. The language you use is just so reminiscent of someone who is hostile to her own past Christianity....

"If you value freedom, you should flee from religion as the antelope flees the lion. "

"Religion is the very antithesis of freedom, insisting on our complete subjugation to the unachievable demands of an invisible but supremely powerful overlord."

"There is no aspect of our lives, no matter how intimate, which religion does not unblushingly insist on its right to control. "

Moreover, new wave atheists also continually speak of the Old Testament as though that is the current standards of Christianity, when any Christian knows that Christ is the standard..

"The Abrahamic god even enthusiastically endorses the vilest of all negations of freedom: slavery. In Leviticus 25, there is a direct quote from this supposedly perfect deity, specifically permitting the Israelites to take and keep slaves, the only proviso being that they must be from the neighboring tribes and not from their own people. Straight from the horse's mouth, as it were, and hardly a shining example of freedom as a religious ideal."

"Religion delights in petty rules and the exercise of power over its followers. What theistic religion does not attempt to curtail believers' freedom with nonsensical decrees about foods that may or may not be eaten, fibers that may or may not be worn, days on which they may or may not work, coverings that must or must not be worn on their heads, books that must or must not be read, images that may or may not be created, words that may or may not be spoken, ideas they may or may not explore, actions they may or may not perform, rituals - whether physical or symbolic - they must perform in order to cleanse themselves of impurities of religion's own invention? "

How can you hope to have a balanced view of faith when you're so hostile to every part of it?

Thanks again for posting.

Best Wishes


 Recommended reading 
Not rated. (3 reviews / comments)
  • Write a review or comment
  • Site Search

     Norfolk services........ 


    Twitter-button  facebookbutton

    Sign up for our
    free e-newsletter

    Send us your latest local news and events