There are two big problems with Paul's research inasmuch as it can be said to provide evidence for the effects of religiosity on a society: one, he doesn't measure the actual effects of religion, and two, he gives no evidence that the two sets of data he examines are at all connected.
This could be a case of a scientist playing too far away from his home pitch. Whereas in paleontology one normally looks at concrete finds or facts as isolated instances, in modern sociology it is almost universally dynamics over time that are considered the basic research unit. So to present two sets of statistics - numbers of religious people and number of, say, teen pregnancies - is by no means to prove that there is any causal link between them. Indeed, SF says as much in his own writeup. More importantly, however, it is entirely pointless to consider these number out of historical context, unless you can show that the levels of both church attendance and teen pregnancies has remained at a constant level in all post-industrial societies, which is patently absurd.
The US has not always been as religious as it is today. Although it did always have a higher rate of evangelism than many European countries, it was by no means a markedly more religious society than Italy, say, or Portugal. In the last 20 years a specific stream - you could call it a sect I suppose - of Christianity has steadily gained in popularity, contributing to a dramatic rise in church attendance numbers. At some point in the last 20 years, church attendance in Italy and the US would have been about the same (never mind for the moment the fact that the size and diversity of the US demographic landscape would have made a nonsense of that anyway); at this point, what was the ratio of other social ills per capita in these two countries?
It is equally important to examine the trending of the two phenomena alongside each other. What were the numbers like before the point at which the two countries started going in opposite directions (churchgoing up in the US, down in Italy)? Which of the two had the higher rate of teenage pregnancy in, say, the fifties, and how have they changed since? Was, for example, the rise in teen pregnancies in the US proportionate to the rise in church attendance? And was the opposite effect observed in Italy? We could get a bit of a shock if we find out that in fact religious affiliation started rising at the point when the numbers were the bleakest, because that would tend to indicate that religiosity is a reaction, or a corrective action, of a nation embattled by social ills, as opposed to being a cause at all.
Staying on the subject of teenage pregnancy, it is in fact on the wane in the US(1). The numbers are also down in the UK(2), as is church attendance(3); but the UK still has the highest teenage pregnancy rate in western Europe. What does that last fact mean? Does it mean anything at all? On the face of it it is significant that an increase in secularism has not wrought the same dramatic change in the numbers in the UK as it did in Scandinavian countries, for example - but surely in order to to examine that we have to set aside religion as the only factor and interest ourselves in other social dynamics and economic factors affecting the different countries.
Meanwhile Danish scientists have recently claimed that the decline in teenage pregnancies is more due to a decline in the fertility of young males(1) than any education or church based initiatives. If this is true to a statistically significant extent, then all the data we collect is effectively meaningless. Even if it isn't, we still haven't looked at contributing factors such as the availability of cheap and effective contraception, living standards and levels of poverty and education, the general permissiveness of social mores in non-sex related matters etc.
I've kept with teen pregnancy because it's an easy example to find statistics about. But it is not unique in its complexity. Briefly touching on violent crime, for example, we can see that despite the fact that Jesus would almost certainly be anti-gun, the rate of gun ownership in the religious US is very high, as is the rate of violent crime. In order to get any insight into possible connections between the two, we would need to properly understand the ratios of churchgoing folk owning guns, as well as whether they justify this ownership on religious grounds or on grounds of compatibility with generally more conservative opinions, which might in fact be at the basis of both behaviours.
I am not advancing any argument in favour of religion as a force for good, although it's not impossible to make one (the role of Christian movements in the abolition of slavery is well documented, though more recent examples are harder to come by). Certainly religiosity has not done the US any favours in terms of its foreign policy. But religion is so intertwined with the life and mores of a population that it would be difficult if not impossible to draw easy conclusions about its isolated impact on any trend across an entire society. We may as well say that as Japan is the most ethnically homogenous post industrialised society, and has the highest rate of education and technological market penetration, then ethnic diversity is bad for the education system.
(1) http://www.slate.com/id/2140985/?nav=fo
(2)http://www.nhsinherts.nhs.uk/hp/health_topics/teenage_pregnancy/teenage_pregnancy.htm
(3)http://www.timesonline.co.uk/article/0,,2-1511237,00.html