Since I couldn’t find a convenient single dataset with publicly recorded MP religious identification I sat down and burnt a significant portion of the weekend making one (available as a google sheet here).
Why do this? We’ve had a week of speculation - some of it mine - about how hard/easy it is for publicly religiously identifying people to get into politics, none of it in any way based on real world data despite wikipedia being replete with it. I really do *hate* speculation that doesn’t even make a nod to evidence (particularly when it’s mine). (Also, I thought it would be quicker)
How did I do this? Mostly from scraping wikipedia. Sat down and went through the first 50 UK 2019 MPs manually to get reasonable keywords then made a little scraper in Python to go look for those keywords in all the MP pages. Sadly, even once I’d excluded “Christian (Wakeford)” and "Church(ill)” the results were something I had to go through manually (*now* maybe there’s enough data to train an neural network to do it for me!). Frequently, wiki entries were ambiguous so I’d google around a bit/follow links - for each entry I saved the most unambiguous source I found. I also finished up by using a small set of lists of publicly religious MPs (compiled by religious newspapers / parliamentary groups for specifically religious MPs).
Note - I used the 2010-2019 BES Constituency Results with Census and Candidate Data dataset to get winning MPs so it’s just the 632 from Britain (i.e. NI excluded) - have the results for NI, but they are unsurprising (although surprisingly hard to find a clear link to a public declaration of religious identification from some members of the DUP!).
Headline observations from having pored through a couple of hundred wiki entries looking for unambiguous public statements of religious identification?
British MPs are generally quite disinclined to talk about their faith/lack of it. Some people are immediately going to assert that this is because secular society hounds the religious etc etc. But, actually, if we take this 1997 anonymous MP questionnaire as a very rough baseline - just to establish that we’re expecting probably something quite a bit more than 30% of MPs to be nonreligious - then it’s the “nonreligious” MPs who are least likely to publicly identify. Note - have only included people as “nonreligious” if they’ve stated that explicitly as an identification.
The broader point is that British MPs - as much as British people - are generally disinclined to raise their private religion [ed. where’s your cross country comparison?]. I was able to find only 22% of MPs publicly identifying as any religion/none - but it was obvious that this is not a random sample. By which I mean that you have a small minority of MPs whose faith is such that they *actively want to tell you about it* - e.g. strong skew towards evangelical faith - then you have people who seem more obligated by circumstances e.g. they are visibly part of ethnic group associated with a specific minority religion.
You can also see a lot of MPs who aren’t personally religious but who want to assert connections with a specific community - quite a lot of “I’m not religiously X but I am culturally X”. That division is clearest with Judaism/Jewishness but you can see people trying to make a similar distinction Islam/Muslimness (Sajid Javid appears to get quite tetchy at the implication that he’s not Muslim just because he’s not religious). This is most overt among ethnic minority MPs but you can see it among Catholics and even some very right-wing Conservative MPs.
It’s particularly interesting that MPs generally don’t want to bring up their faith because - at least for Conservative MPs - you can see evidence some circumstantial evidence that they believe it would be beneficial to do so. By which I mean that Conservative MPs in leadership elections start talking about their faith. More than that, look at this Liz Truss quote - "I share the values of the Christian faith and the Church of England, but I'm not a regular practising religious person." (tough one to parse as Christian: Yes/No, I took some consolation when I found a very conservative Christian advocacy site struggling with this as well).
Two points I want to stress here
this dataset has very, very obvious response selection issues (we’re not seeing a random sample of religious identification so much as Who Feels A Desire/Obligation To Disclose / Who Got Leant On By The Speaker To Write A Public Roots Essay On Their Faith)
a lot of the narratives thrown around this week about How Hard It Is To Religious In Politics / That Ethnic Minorities Get Less Scrutiny are obviously silly even just at the data collection stage
Overall, gathering this data it seems like there are 3 clusters of Religious MP Identification
Evangelical/Ideological (“I want to tell you the Good News/I have some extreme opinions on gay marriage/faith healing that I justify in terms of my faith”) - must be noted that there are a bunch of Conservative MPs clearly trying to identify with this group in every way except faith (people voting alongside, framing positions as about religion, telling you that they are ‘culturally christian’ but not actually personally a believer).
Ethnicity/Community (“I am so happy to be here an X in parliament, I grew up in an X household and I’m a proud member of the X community”). I’m exaggerating here for effect - there are plenty of practicing members of minority ethnic groups but they do tend to be younger and more moderate.
Randomly Publicised (“I didn’t particularly want to talk about my religion, but I turned up to one Christians On The Left/Conservative Christian Fellowship thing and got added to the list”)
Note on gay marriage - there would be much less public data here if it weren’t for gay marriage debate/legislation/votes. Not only do a lot of people only go on the record to say that they are Christians when arguing against it - but also for it.
So what does the raw data look like? Definitely in need of some further processing, in this first round I just tried to get the most detail that was available but this is probably too much precision to be of use (and Caroline Ansell gets a whole category to herself - did not think I’d have to resolve what to do when someone identifies as both Catholic *and* Evangelical).
Most obvious thing here is that minority faiths/preferences of every (okay, many) persuasion - Catholics, Christian Evangelicals, Muslims, Humanists - are well represented. The group least represented (proportionate to UK census results) is the non-denominational nonreligious and just behind them Christian/Christian CoE
What else can be done with this data? Given how partial and heavily selected it is, not much to think about the 78% of parliament that doesn’t want to disclose religious identification. But we do have voting record data, so it should be possible to identify to use this to identify ‘religious’ patterns of voting (most obviously around things like gay marriage).
I also invite people to improve/expand the dataset - I could only bear to do it for 2019, but the process can be repeated for 2017/2015 with pretty much the same quality of data. You can also look at all *candidates* (although expect much less data to be available).
Good stuff - a real public service! A supplementary source (for atheists anyway) would, I suppose, be whether MPs choose to swear or affirm - https://www.parliament.uk/about/how/elections-and-voting/swearingin/ ?