New work by Pablo Brañas-Garza, Antonio Espín and Shoshana Neuman examines the effect of religiosity on the extent to which people's behaviour is pro-social. Based on games frequently used in experimental economics, the results, based on quite large samples of individuals, are striking. Active participation in religious activities appears to be strongly associated with more altruistic behaviour. Looking into the future, as adherence to faith declines, so (in the absence of countervailing measures) might the extent of pro-social activity.

The mechanism underpinning this finding is complex – it is likely to be partly the direct result of religious teaching, but partly also the result of the development of a sense of community that is engendered by frequently meeting neighbours at a religious institution. The latter should be seen alongside other developments – such as the growth of home entertainment – that together may reduce social cohesion.

Altruism, it seems, needs institutions. If our society lets the old institutions go more quickly than it lets go of its need for altruism, then new institutions will need to develop to plug the gap. These new institutions may be virtual – social networking, for example. But it remains to be seen whether the connection between the virtual world and the real world is sufficiently strong to serve the need.

In a fascinating study of the impact of corporate culture on company performance, Luigi Guiso, Paola Sapienza and Luigi Zingales have found a strong positive effect of managerial integrity and ethics. This should not be surprising, but it is immensely reassuring. In competitive markets, bad behaviour should be competed out of existence.

Òscar Jordà, Moritz Schilarick and Alan Taylor have produced one of the more interesting analyses of recession following financial crisis. As is by now well known, the severity of such recessions greatly exceeds that of a typical recession. Figure 6 of their study demonstrates this vividly. Following a financial crisis, output falls further and takes longer to recover than in a normal recession; this follows from the fact that investment - which is barely affected in a normal recession - falls drastically and recovers very slowly; this is reflected also in lending data, which show very slow recovery in the credit market.

The recent recession was triggered by an increase in debt. As the international evidence reported in Figure 1 of the study shows, this debt increase was primarily a characteristic of the private sector – indeed public sector debt had been falling as a proportion of GDP in the decade leading up to the crisis. Notwithstanding that, the ability of governments to respond to financial crisis necessarily depends on their own debt position; clearly this has deteriorated in the wake of the crisis, and contributes to the overall level of indebtedness. An adjustment was necessary – and the low levels of bank lending that have served to prolong the recession represent a part of that adjustment.

Another part is fiscal retrenchment. Hopes that this would of itself be expansionary (owing to a ‘confidence fairy’ effect) always appeared fanciful. The trick to recovering the government’s financial position must to phase fiscal policy so that retrenchment is achieved over a reasonable time horizon, but without proceeding so quickly as to damage the recovery that now seems under way.

Gordon McCord and Jeff Sachs have provided a fascinating analysis of the determinants of economic growth. This is an area in which I too have dabbled in the past, and many of their results - concerning the effects of landlock, openness etc. - reassuringly resonate with mine.

Surprisingly, though, McCord and Sachs find that the number of years of schooling typically received by men serves to dampen growth - though the schooling of women increases it. There are many possible explanations why female education should have a stronger positive effect on growth than male education - for instance, women may be more effective in passing their knowledge onto younger generations. But the finding that male education actually reduces growth appears perverse. I would argue that it is one that should be treated with caution - there is surely a high degree of collinearity between the amounts of schooling that men and women receive, and this could be distorting the result. This is a case of a study where further detail would be welcome.

The latest industrial production data, released this morning, reveal a drop in (seasonally adjusted) output in August. Nevertheless, feeding these data into my neural network forecasting model reveals that there remain grounds for cautious optimism about the likely trajectory of the economy over the next couple of years. The forecasts appear below.