In Times of Stress, Can Religion Serve as Insurance?

The New York Times, "Economic Scene" , September 08, 2005

From layoffs to bad health to killer hurricanes, economic well-being is subject to all sorts of nasty shocks. To protect themselves, people accumulate savings and buy insurance.

But these formal strategies are not the only ways to guard against sudden hard times. Social networks -- families and friends -- also provide critical safety nets.

Economists are increasingly interested in the role of such networks in economic life. They want to understand "the way that an individual's social connections somehow translate into an economically valuable tool," Rajeev Dehejia, a Columbia University economist currently visiting at Harvard, said in an interview.

In particular, he and other economists have recently begun to examine how, in addition to any spiritual or otherworldly benefits, religious affiliation may serve as a kind of insurance.

With two co-authors, Thomas DeLeire of Michigan State and the Congressional Budget Office, and Erzo F.P. Luttmer of Harvard, Professor Dehejia takes up the question in a new working paper for the National Bureau of Economic Research. (The paper, titled "Insuring Consumption and Happiness Through Religious Organizations," is available here.)

The paper uses two sets of data, measuring two different kinds of well-being, to address the question. In the first, the economists use the government's authoritative Consumer Expenditure Survey to see how a household's spending changes when income changes. Since the survey records whether respondents contribute to religious groups, the researchers use that as their marker of religious participation.

"About 40 percent of households make a contribution to a religious organization and these contributions represent about 1.2 percent of household income," they write.

Looking at spending before and after income goes up or down, the economists found that religious households have less volatile consumption. Their spending fluctuates about 30 percent less than spending by similar, nonreligious households.

Interestingly, the researchers note, this insurance effect shows up most among people who give relatively little to religious groups. "Our baseline results," they write, "are driven to a large extent by households that contribute less than $400 annually." That is less surprising than it might seem, because the households that benefit are themselves relatively poor."

The effect is strongest for whites, where religious affiliation "buffers consumption against roughly 35 percent of the impact of income shocks," they write. The result for blacks is not statistically significant but points in the same direction, perhaps, the economists speculate, because black congregations are more likely to provide nonfinancial support like child care or meals.

The second part of their study looks not at spending but at a more direct measure of well-being: how respondents assess their personal happiness.

After all, Professor Dehejia said, "If you take the doctrine of religion seriously, then what religion actually does is it doesn't insure your income but it changes how you react to the negative shocks you face."

Here, the economists use the National Survey of Families and Households which, among other questions, asks respondents to rate their happiness on a seven-point scale. Because the survey interviewed the same people in 1987-88 and 1992-94, the economists can compare how the same people responded over time.

Unlike the consumption survey, which records only donations, this one has more detail on religious beliefs and activities. (It does not, however, measure gifts.)

Using religious attendance as their measure of affiliation, the economists get a strikingly different result by race. While whites get no significant "happiness insurance" from religious affiliation, blacks get a lot.

"The median level of religious attendance reduces the happiness impact of income shocks by blacks by about 75 percent," they write. The effect is greatest among those with a high school education or less.

The puzzle, said Professor Dehejia in the interview, is not why blacks benefit but why whites do not. "Why is it that among whites you see a significant insurance effect on consumption but not on happiness?" he asked. Perhaps, he speculated, whites feel guilty about the help they get.

The results are about the same for religious belief as for attendance, since the two are so closely connected. The economists note, however, that attendance matters most. "Just believing is not sufficient; one needs to participate in a religious organization to get happiness insurance," they find.

As for Hurricane Katrina, Professor Dehejia noted that the disaster had probably disrupted congregations too much for them to provide much financial support.

But thinking of the hurricane as "a huge negative income shock," he said, "If you believe our results literally, what they would say is that religious individuals should be less hurt in terms of their sense of well-being, that religion is going to help people come through it."