IF YOU want to avoid an argument over religion at your next dinner party, you might suppose it safe to invite an economist or two. They, of all people, could be expected to stick to Mammon. Or maybe not, if a new paper by Robert Barro, one of America's best-known economists, and Rachel McCleary, a colleague at Harvard University, is any guide. It explores the influence of religious belief and observance on economic growth.

To be sure, this is not the first time that economists have held forth on subjects that many people consider private. Gary Becker, a Nobel laureate at the University of Chicago, is well known for applying economic theory to questions such as whether marriage is economically efficient, whether drug addiction is rational and how couples decide how many children to have. Mostly, though, economists have so far left religion alone. However, other students of society have not been shy of offering ideas about religion and wealth. A century ago, Max Weber, a founder of sociology, observed that the Protestant work ethic was what had made northern Europe and America rich. Recently, Niall Ferguson, a British historian at New York University, argued that today's economic stagnation in Germany and other European countries owes much to the decline of religious belief and church attendance during the past four decades. The Protestant work ethic, he thinks, is dead.

If there is a link between religion and economic performance, then economists ought to have something to say about it. To test the connection, however, economists need figures, ideally covering many countries and many years. This is where Mr Barro and Ms McCleary come in. They have collected data from surveys of religious belief for 59 countries in the 1980s and 1990s—whether people say they believe in God, heaven and hell, and whether they attend services at least once a month—and have tried to tease out whether these have any direct effect on GDP growth.

This is not easy, because in theory the causal arrow could run either from religion to economic growth or in the other direction. On the one hand, adherence to religion might foster behaviour that helps a nation to become rich. The Protestant ethic emphasised thrift and willingness to work hard. Honesty might generate trust. Openness to strangers could lead to openness to trade and immigration. On the other hand, economic growth might bring about changes in a country's religiosity. As European countries have become richer during this century, fewer people have gone to church. In America, however, the downward trend has stopped, or even gone into reverse, in recent decades and Americans are far more religious than Europeans. It is also thought that church attendance declines as people become better educated and as they move from the country to the city. Both these shifts tend to accompany economic growth.

In order to sort this out, the authors needed to find measures of religion which are not themselves affected by GDP growth. To do this, they draw on a fundamental tenet of orthodox economic faith: that more competition is better. In some countries, religion is (or was) banned altogether or discouraged. In others, a single strand is sponsored by the state. In others, there is a free-for-all.

Earlier studies, say the authors, suggest that religion is more likely to flourish where there is less state control and a greater diversity of belief. So they use this to test the impact of religion on growth. Up to a point, their findings corroborate some common perceptions. More prosperous countries seem to have lower rates of church attendance, although America—the best instance of a country of competing sects rather than a state religion—is a conspicuous exception. More urbanised countries tend to be less religious. However, contrary to what many people think, religion seems to have a stronger hold in countries with better educated populations.

The most striking conclusion, though, is that belief in the afterlife, heaven and hell are good for economic growth. Of these, fear of hell is by far the most powerful, but all three indicators have a bigger impact on economic performance than merely turning up for church. The authors surmise, therefore, that religion works via belief, not practice. A parish priest might tell you that simply going through the motions will bring you little benefit in the next world. If Mr Barro and Ms McCleary are right, it does you little good in this one either.

Indeed, Mr Barro and Ms McCleary go further. They find that church-going, after a certain point, is (in an economic sense, anyway) a waste of time. They argue that higher church attendance uses up time and resources, and eventually runs into diminishing returns. The “religion sector”, as they call it, can consume more than it yields.

All this is intriguing, but does religion make much difference? Japan, where there are many sects but little fear of hell, has grown far faster since the second world war than the Catholic Philippines. Officially atheist China is growing at a cracking pace. Presumably, what your religion is matters as much as whether you are religious: has anyone theorised about a Catholic work ethic? And if religion does have an effect, is a 20-year period long enough to find it?

Equally, religious people would say that prosperity is not the purpose of faith. Few these days would even see it as a reward for their belief and behaviour, but as at most a by-product, and then not necessarily a welcome one. Recall Jesus's words about the rich man, a camel and the eye of a needle. So is Mr Barro's and Ms McCleary's rough, preliminary research good news for the faithful? Or is it just a little disconcerting?