Searching “happiness” on the NBER working papers database was a very good idea. There are a ton of cool papers floating around in there, and today’s paper combines models of human consumption-smoothing behavior with… you would have never guessed it… your temple/mosque/church/[insert religious place here]!
One of the authors is Professor Luttmer, with whom I’m taking Econ 21. Cool!
Anyways, I think this paper is pretty funny because it basically says that membership and contribution to a religious organization is a decision that, like all of the others we make, is motivated by our own benefit. We’re selfish! And we can’t help it! In this case, it’s a form of consumption/happiness smoothing. Contribute to a religious organization during good times, and it will support you during bad times. It makes sense, right? (Technically one would hope that it would support you during bad times whether or not you loosen your purse strings as the collection plate goes around… but I digress.)
The results of the paper are lovely, I think, and the cultural implications are fascinating. The idea is that changes in income will pass through to changes in consumption and happiness level, and if you interact change sin income with religious participation, you should be able to see smoothing effects. Basically, using two different household surveys, the researchers find that for WHITE families, contributions to religious organizations prevent about 40% of an average consumption shock. For BLACK families, participation in religious services (not contributions) prevents about 66% of average happiness loss from an income shock.
My main question, which the paper does address, is whether people who participate and contribute to religious organizations are inherently different from their atheist counterparts. Clearly, they will be, and perhaps these differences are correlated to differences in consumption/happiness smoothing behaviors. The paper uses a nearest-neighbor matching method, which I’m not totally familiar with, in addition to controlling for the host of variables present in the two surveys. I’m going to go do some reading on nearest-neighbor matching (I suspect it’s similar to propensity score matching, which I don’t totally buy, because it’s hard to match people on every dimension except for one as significant as religion, and claim that they are the same in every other way) and revisit this paper.
Nevertheless, I found it interesting that the smoothing effects of religious participation differed by culture, and I thought the topic as a whole was an interesting one to put into the context of a rational decision-making model.