Field of Science

Religious belief and religious involvement have opposite effects on support for democracy

One of the challenges with doing surveys is that the answers you get can depend on the order in which you ask the questions. For example, if you ask people about their religious beliefs, then their minds will be primed to respond to later questions in a way that fits with their beliefs.

This is a problem for surveys, but it also offers a novel research opportunity, as a recent study by Ben-Nun Bloom and Gizem Arikan, political scientists at The Hebrew University of Jerusalem, has shown. They're interested in how religious beliefs affect support for democracy - I blogged a study of theirs back in October.

They recruited university students from Israel (all of them Jewish) and Turkey (all Muslims), and asked them about their religious belief, religious social behaviour, and support for democracy (using standard questions pulled from the World Values Survey).

But the cunning thing was that they varied the order in which they asked the questions, ending up with three different groups who had each been asked a different set of questions first.

What they found was that asking questions about religious belief (i.e.belief in God, heaven, life after death, etc) first significantly reduced support for democracy - and this effect held regardless of how religious the individual was.

On the other hand, asking questions about religious social behaviour (i.e. attending services, having friends of the same religion, etc) first increased support for democracy! Again, this was independent of how important religion was to the individual's social life.

The positive effect of religious behaviour on support for democracy was a little stronger in Turkey than Israel, but otherwise the basic trends were the same in both countries.

So this puts an additional twist on their earlier results - which used data from the World Values survey to show that religious beliefs decrease and religious socialisation increase support for democracy.

What they suggest is that continual exposure to religious ideas and messages results in a kind of life-long priming. Religious socialisation encourages group cohesion, which might affect support for democracy. Religious beliefs trigger thoughts of traditionalism, security, conformity - which might act to reduce support for democracy.

Clearly, religion and democracy are both complicated beasts, and so there is not going to be a straightforward relationship between the two!

But this study also just goes to show how malleable is the link between beliefs and behaviours. Back in 2009, I reported on another study which found that what people tell you about their attitudes to risk depend on whether you ask first about their religion and gender.

If you ask these questions first, they will give you an answer about attitude to risk that fits the social stereotype. But if you ask about attitude to risk first, you get answers that are far less stereotypical!

Now, social scientists and psychologists are well aware of these issues, and in the studies I review on this blog they are usually careful to ask about religious beliefs after the experiment, and not before.

But I do often wonder about all those studies that use data from large surveys. Usually the questions on religion are buried within questions on all sorts of other topics. So who knows how reliable they are!
Ben-Nun Bloom, P., & Arikan, G. (2012). Priming Religious Belief and Religious Social Behavior Affects Support for Democracy International Journal of Public Opinion Research DOI: 10.1093/ijpor/eds030

Creative Commons License This article by Tom Rees was first published on Epiphenom. It is licensed under Creative Commons.


  1. In my career, I have done many sociological and political surveys and supervised many students who did surveys. While surveys are a great tool, they're just a tool, i.e. not a guarantee of success -- nor truth, for that matter. Many skilled workers have hit their fingers with a hammer or cut themselves with a saw.

  2. That's really interesting. I wonder if people who make surveys typically think about the order in which they ask the questions.

    Do you know how big the effect was? If it is really significant, then it certainly seems like it needs to be considered very carefully at the planning stages. It also seems like something that would have to be considered when people do a meta-analysis.

  3. At the same time, in a big questionnaire like the WSV there are so many different questions that all sorts of different things will be getting primed simultaneously. This does not mean that the order becomes unimportant but that the effect probably becomes impossible to predict. The only response that comes to mind in that case is to have the order of the questions be randomised for each respondent.

  4. It'd be interesting to see the same questionaires put to rank unbelievers. Does the independence from strength of belief and involvement extend to those who have no religious belief or involvement at all?

  5. excellent and great doubts.
    It just shows that our illusion of a singular self that believes something is delusional. We have multiple beliefs, multiple selves, all fluxing depending on cues and uses. We are more flexible than we'd like to imagine.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS