UPDATE: A version of this article has just gone up at the Guardian's Cif.

Robert Putnam has been in London, channeling the wisdom of social capital at Number 10, as well as talking at St Martins-in-the-Fields yesterday evening. That venue is the big clue to his latest findings. It could be summarised thus: if you want big society, you need big religion.

In the US, over half of all social capital is religious. Religious people just do all citizenish things better than secular people, from giving, to voting, to volunteering. Moreover, they offer their money and time to everyone, regardless of whether they belong to their religious group.

It could be, of course, that the religious already have the virtues of citizenship. However, Putnam believes the relationship is actually causal, not just a correlation. Longitudinal studies show as much. So why?

He argues it's not to do with belief, but with being part of a community of belief. An atheist with several church-going friends will be a better citizen too. In fact, church-going friends are what he calls 'supercharged' when it comes to citizenship. Working out just what a religious community gives would be key to generalising the findings beyond faith.

If that can be extracted from the religious context at all - which was the point raised by Phillip Blond, Putnam's interlocutor last night. To my mind, Blond asked exactly the right question: can you imagine a secular network that matches a church network for its pro-sociality?

Putnam could not. He doesn't know what makes faith communities civically exceptional. Not even networks of environmentalists, that share interests beyond themselves too, score so well. In short, it doesn't look as easy to separate the content of faith from the community of faith, which is intuitively as you'd expect, as communities of faith arise and are sustained for reasons of faith.

(As an aside, I felt that Putnam has a fairly naive understanding of faith communities. He stressed several times that he is not a theologian, just a fact finder - though, of course, facts can't speak for themselves but must be interpreted. Perhaps some further religious education would help...)

Blond had a deeper critique too. Putnam was insisting that it's being part of a community of faith that matters, regardless of the nature of the faith that community represents. But mightn't the reason why an American Buddhist finds they have a lot in common with an American Christian, say, be not because they are both religious, but because they are both American. In other words, Blond continued, to be American is, in a lower case sense, to be Christian - to be ethically shaped by the religion that predominantly informed the founding of America, and has arguably championed the virtues of equality, social concern and participation more than any other. Putnam replied that research in non-Christian cultures would have to be done to determine the case.

Putnam thinks that the evidence shows the link between civic engagement and religiosity in the UK is pretty much the same as in the US, notwithstanding that British religiosity is obviously far less pronounced. So what was he saying at Number 10?

The first principle of any government should be, Do No Harm. From that it follows, if you're convinced by the evidence that religiosity is pro-social, that the politician should make sure they don't undermine people's habits when it comes to belonging to communities of faith.

The other big policy implication of his talk seemed to be to encourage the development of bridging social capital (links with others), as well as bonding social capital (links to your own) - which is to say, encourage faith groups to integrate. But, he added, there's a crucial further point here too.

The evidence is that strong faith communities make for strong interfaith and wider social links too. In other words, it's a mistake to assume that inward looking groups aren't also, on the whole, civically outward looking. So the French are wrong to insist a good citizen can't wear the veil. It also followed from what Putnam said, though he did not say so explicitly, that faith schools are, generally, good for society. In short, the link between bridging and bonding is not zero-sum, as is perhaps often assumed.

There are many other things you can do to build a big society, Putnam stressed. It's just that he's been looking at religious communities recently. However, supporting them is a big one.