Is the Recession Really Better for Evangelicals?

Paul Vitello's New York Times article that cited my research has started something of debate as to whether Evangelical Protestant churches truly are benefiting from the current recession. The debate began when Frank Newport of Gallup replied that polling numbers do not show any overall increase in weekly church attendance, a fact gleefully promoted by Jack Shafer of Slate and later challenged by Mark Silk. More recently, Tobin Grant of Christianity Today weighed in on the matter and concluded that there is no "evidence that evangelical churches would benefit more than other religious groups." His piece was of particular interest to me because he references my work and along the way makes this claim:

[Paul] Vitello should have interviewed a few more sociologists or economists, who would have told him that the link between recessions and revivals is just myth. This may have helped avoid the back-and-forth that has resulted in confusion over a topic where we need much more clarity.

Unfortunately, Grant himself adds to the confusion by making this brazen claim. Had he read my paper fully or knew the literature better on this issue he would have been more cautious in his conclusions. There is both evidence and theory that suggest a link between the business cycle and religiosity. Moreover, this evidence and theory also points to evangelical Protestants benefiting more from recessions than their mainline counterparts. Here is a quick overview.

Stephen Sales in a 1972 study titled "Economic Threat and the Determinant of Conversion Rates in Authoritarian and Nonauthoritarian Churches" found that the conversion rates for more conservative denominations to be countercyclical while for more liberal denominations they are procyclical. J.H. McCann did a study in 1999 titled "Threatening Times and Flucutations in Church Membership" that similarly used the authoritarian-nonauthoritarian classification scheme. His conclusions mirrored Sales. Although their studies were not limited to Protestant denominations, there is a close mapping in them between the authoritarian and Evangelical denominations. My own study also looked at membership patterns for a sample of 25 Protestant denominations over the years 1968-2004. I found a systematic countercyclical component to the membership of the evangelical Protestant denominations during this time. Specifically, evangelicals grew significantly faster in recession years, when unemployment picked up, and when the stock market tanked. This was not the case for the mainline Protestants. Grant, however, tries to dismiss these findings:

Beckworth's findings were limited to changes in total annual membership of 24 Protestant denominations. Beckworth acknowledges that this is merely a proxy for other types of religious activities such as attending worship. But as CT readers well know, church membership means different things in different denominations. Some have been members since infancy. Others attend the same church for decades and never become members. Membership also changes for reasons other than changes in religiosity. When a diocese, synod, or other group of churches leaves in protest, the membership numbers decline dramatically... At other times, church membership numbers may be inflated and need to be corrected.

The problem with Grant's critique is that while everything he said may be true, it cannot explain why for 30+ years the above mentioned economic measures were systematically associated in a countercyclical fashion with evangelical Protestant membership and not with mainline Protestant membership. The easy answer is that while all these other idiosyncratic factors may have been influencing membership, so was the economy. In fact, this understanding is consistent with my findings: about 1/3 of the variation in evangelical Protestant membership growth is explained by the economy. That leaves the rest to the Grant's other factors.

To be clear, a key assumption in these findings is that increased religious participation by U.S. Protestants will be manifested in increased attendance at churches and, in turn, in increased church membership. While increased religious participation could also be manifested in other ways, it should been seen at a minimum in increased church attendance and eventually in increased membership if some of the new attendees formally join the church. Yes, ther factors will also be influencing membership numbers, but given what the data shows this assumption is reasonable.

Now even if you accept Grant's critique of the membership findings, you then have to wrestle with the other findings in my paper. Primarily, the finding that during the last recession in 2001 employment status was a significant determinant of weekly attendance at church. This data was based on a pew survey and controlled for a number of confounding factors, including the 911 effect on attendance. What I found is that if one were unemployed there was a greater probability that he/she attended church weekly. Moreover, being unemployed only had an effect on evangelical Protestants.

This part of the paper also helps shed light on Newport's finding that overall attendance has not gone up. My findings for this period showed about 42% of population attended church weekly, almost identical to Newport's finding for this year. But this number by itself it does not necessarily shed light on the effect of the recession. What is needed to see what part of the population that is adversely affected by the recession is attending church on a weekly basis. That is how I found a higher rate of attendance in the 2001 recession. I suspect Gallup would too if they sliced up their data this way.

Finally, There are compelling a priori reasons why why evangelical Protestants should benefit more from recessions than mainlines Protestants. First, evangelicals sell a different product than mainlines. Of interest here, is that they sell more certainty. Mainline churches are more likely to have fewer absolute beliefs, question the Bible more, and raise more questions about God's role in human affairs. Evangelical Protestants, on the other hand, are more likely to sell a powerful God that can get you a job and bring you financial peace. For someone who is unemployed and trying to feed a family, the evangelical message is far more compelling. Second, evangelicals on average come from a lower socioeconomic background than mainlines. This means they are more likely to have a higher opportunity cost for religion. As a consequence, during economic booms religious participation becomes too costly while during recession it becomes cheap for them. The opposite holds true for mainline Protestants. (See here for more on these channels. There is also a social capital formation channel I discuss in the paper.) These stories imply that evangelical Protestants have more to gain during an economic downturn.

Let me close by noting that while I believe there is evidence and theory to support a link between the business cycle and Protestant religiosity, only time will tell is it will be borne out in this recession. I am, however, certain that calls to dismiss this link at any level are not well founded.

Update: The recession is the not the same everywhere in the United States. As I noted here, some states like Texas and Oklahoma are still adding jobs while states like California and Florida are a workers nightmare. Consequently, the benefit to evangelical churches should be occurring more in these states.