We are in the Islamic holy month of Ramadan. During this time, as prescribed by one of the five "pillars", or obligations, that make up the foundation of Muslim life, hundreds of millions of followers are abstaining from eating and drinking (and a host of other activities, like smoking and sex) between dawn and sunset.
This fasting does not burden all the world's Muslims equally. For Muslims who live in the Northern Hemisphere, the day is longer at this time of year than it is in the Southern Hemisphere - which leads to longer fasts.
But this isn't always the case. The Islamic calendar is lunar (around 354 days) and does not incorporate leap years, so Ramadan rotates slowly through the Western calendar, year after year, 11 days at a time. In a decade and a half, Ramadan will fall during the shorter days of winter in the Northern Hemisphere, while Muslims in southern locales like Australia and South Africa will face longer fasts.
This cyclical feature of Ramadan turns out to be a handy tool for social scientists. As we demonstrate in a working paper for the United States' National Bureau of Economic Research, it helps us address a longstanding question: Does religion - specifically, the practice of religious rituals - affect economic growth?
It may seem obvious that engaging in religious practices (attending services, praying, studying sacred texts) would slow economic activity, as such activities take time and effort that could be directed elsewhere. Indeed, several recent empirical studies have found a negative correlation between such religious behaviour and economic growth.
But correlation is not causation. The causal effect of religious practices on economic activity is much more difficult to establish, not least because there are so many other variables that influence economic activity. How can we tell what specific impact religious practices might have on economic growth? We need to somehow isolate that effect.
This is where the Ramadan cycle is helpful. Because the length of fasting during the holy month varies idiosyncratically but predictably as a function of the lunar calendar and latitude (factors that are presumably unrelated to those that affect economic behaviour), Ramadan presents a kind of naturally occurring experiment: Religious practice is precisely varied and everything else is left in place.
Analysing data taken from many countries over many years, we found that a more intense Ramadan (one with longer fasts) did indeed have a negative effect on growth in predominantly Muslim countries. This finding was robust, holding when growth was measured by GDP: by GDP per capita; in yearly rates; and when aggregated in up to five-year periods. We also found no effect of a more intense Ramadan on GDP growth in non-Muslim countries, which increased our confidence in the finding.
According to our numbers, if the average daily Ramadan fasting were to increase to 13 hours from 12, economic growth would decrease by about 0.7 of a percentage point.
Our evidence indicates that the negative economic effect of Ramadan goes well beyond the short-term fall in productivity that you would expect from the fasting itself. Instead, it seems that a more intense Ramadan leads Muslims to choose to work less or, more precisely, to work in more flexible, if less productive, activities. At the same time, and quite possibly related, Muslims report to care less about work and money, and more about religion.
On the other hand, using data from the World Values Survey, a global research project focusing on people's beliefs, we also found a striking effect of Ramadan on Muslims' self-reported well-being. Following a particularly intense Ramadan, Muslims were more likely to state that they were happy and satisfied with their lives. This may have been a result of the spiritual practice itself, or of the socialising that Ramadan entails, or of a changed perspective on what is important in life - or of a combination of these and other possibilities.
More broadly, what this specific example of Ramadan shows is that religious practices can change people's economic behaviour in ways that are difficult to summarise as "good" or "bad" - even leaving aside the many other possible benefits and costs that our study was not designed to address.
Still, and at the risk of being a tad simplistic, we are tempted to conclude that this year's long days of fasting may make Muslims poorer than they would have otherwise been, but happier nonetheless.
NEW YORK TIMES
Filipe Campante and David Yanagizawa-Drott are associate professors of public policy at Harvard Kennedy School.