Disappearing Act

Over the months since it came out, I was the very least important person pouring scepticism on last year's Bible Society report 'The Quiet Revival', with its startling and much-discussed claim that the proportion of people aged 18 to 24 who attended church at least once a month had risen from 4% in 2018 to 16% in 2024, and yesterday polling company Yougov withdrew the research the paper had been based on. It had been affected by fraudulent responses, they said, that would normally be screened out of results but because of 'human error' hadn't been. The Bible Society played the injured party, stating that Yougov had repeatedly assured them the data was reliable. Now tens of thousands of new Christians, and the hopes they engendered in both institutional and less mainstream church bodies, have faded away into the sociological mist. 

In a way I suppose it's quite encouraging that it's taken less than a year for this to happen. But it's very revealing that so many people (not just the Bible Society) were prepared to go over the top on the basis of research which absolutely was not supported by experience and, more importantly, contradicted any other data set that you might care to mention. News stories covering the 'Revival' were always able to find churches where lots of new people seemed to be turning up (the Church Times was careful to balance a Pentecostal Free Church in the South Wales valleys with an Anglo-Catholic congregation in London), but that's always the case. My colleagues enthusiastically reported people trickling in here and there, but it all amounted to nothing out of the ordinary. Still, that's only anecdotal, and anecdotal evidence reveals only what's possible, not what's representative. But every other actual, quantified survey pointed in the same direction: religious observance in the UK has recovered from the pit it collapsed into during the Pandemic, but not even to the level it was at before it, let alone anything more. It's striking, then, how reputable organisations can so readily overlook problems with information if it suits them. 

To be fair to the Bible Society and others, this was a big survey - 13,000 people. It looked credible. If the problem was really that fraudulent results hadn't been extracted in advance of publication - enough fraud to skew those figures - how compromised is public-opinion sampling more generally? Is it actually that easy to deflect it into unreality?