Supersized: Analyzing the trend toward larger churches

November 28, 2006

Megachurches are increasingly difficult to ignore. By the latest count there are approximately 1,200 Protestant churches in the United States that have a weekly attendance of at least 2,000. The rise of megachurches raises the question: Why now? Why have very large churches proliferated in recent decades? Did something happen in the 1970s or 1980s to make the number of very large churches start to increase in those years?

With a lot of help from denomination-based researchers, interlibrary loan and dedicated students, I examined church size in 12 Protestant denominations back to about 1970, and over a longer period for seven of these denominations. Overall, I tracked the church-size distribution in American Protestantism from about 1900 to the present.

Three results stand out. First, across the Protestant spectrum, there are more very big churches. This increase might not be surprising in the case of growing denominations such as the Southern Baptist Convention and the Assemblies of God, but the same trend is evident in denominations with declining membership over this period. The number of very large churches has increased, for example, in the Episcopal Church, the Evangelical Lutheran Church in America and the United Methodist Church. The number of very large Protestant churches has increased in almost every denomination on which we have data, and it does not matter whether the denomination is big or small, liberal or conservative, growing or declining.

The rate of increase in the number of very large churches seems to have picked up in the 1970s, but the trend toward more very big churches did not begin in that decade. For example, the number of Presbyterian churches with more than 2,000 members increased from 5 in 1900 to 74 in 1983. The number of Episcopal churches of that size increased from 7 in 1930 to 33 in 1960; for Missouri Synod Lutherans the number goes from 2 in 1900 to 23 in 1967. Conventional wisdom on this subject says that the number of very large churches has increased recently—and the results bear that out—but conventional wisdom fails to recognize that this is not entirely a post-1970 trend.

The long-term nature of this trend suggests that simple population growth and increased population density are partly responsible for the proliferation of megachurches. If a town or city or suburb of a certain size is needed in order to support a 2,000-person church, then the more communities of that size there are, the more 2,000-person churches there will be.

A second observation is that the very biggest churches are getting bigger. This too is a long-term trend that did not begin in the 1970s, though again it seems that the rate at which the largest churches are getting bigger has accelerated since that decade.

A long-term trend of this sort could occur by a demographic process of growth proportional to size. If churches simply keep growing by means of births to current members, for example, then the biggest churches would constantly get bigger.

This simple demographic explanation doesn’t work, however, because it assumes that yesterday’s biggest churches would also be today’s biggest churches. The biggest churches would be bigger than they were before, but the same churches would still be at the top of the heap. However, this is not the case.

Yesterday’s very biggest churches are not today’s very biggest churches. Many of today’s biggest churches grew very rapidly; their size is not the result of a steady, long-term increase. And this has been true for at least 100 years. Across all the denominations examined and across the entire 20th century, the half-life of being one of the 20 biggest churches in a denomination is 20 to 30 years. That is, of the 20 biggest churches in a given year, only half of them will still be on that list 20 years later, only one-quarter still on the list 40 years later, and only two or three still on the list 60 years later.

It is not that these very large churches peak and then shrink dramatically, although some do. Rather, the biggest churches of the moment are overtaken by a new cohort of churches that have caught the latest cultural wave and ridden it to the top, and then those churches are overtaken by the next wave, and so on.

The most interesting development, however, is a third trend: people are increasingly concentrated in the very largest churches. Most churches are small, but most church members are part of large congregations. The median church has fewer than 100 regular participants, but the median churchgoer attends a congregation with 400 regular participants. Even though there are relatively few large congregations, large congregations contain a disproportionate share of the churchgoing population. The biggest 1 percent of Protestant churches, for example, contain approximately 15 percent of all the people, money and staff. The biggest 20 percent of churches have between 60 and 65 percent of all the people, money and full-time staff.

You may have heard of the 20/80 rule of social life: 20 percent of the people do 80 percent of the work, 20 percent of the organizations control 80 percent of the resources, and so on. In Protestantism it’s more like a 20/60 rule, but the basic phenomenon is the same. People and resources are heavily concentrated in the biggest churches. This has been true of American religion for a long time, but the level of concentration is increasing.

The accompanying graph shows the percent of people in the largest 1 percent of churches in 12 denominations, seven of which are tracked from early in the 20th century to the present. This is an astonishing picture because every denomination shows the pattern of increased concentration from 1970 to the present, with no end in sight. Denominations vary in how concentrated they are, but all of them show the same trend toward increasing concentration since about 1970.

This figure shifts the focus away from true megachurches. In some denominations, the largest 1 percent of churches includes those with only 500 or so attenders on an average weekend. This is much smaller than the megachurch of popular parlance. This means that megachurches are to be understood as one manifestation of a much broader phenomenon: churchgoers are increasingly concentrated in the largest churches.

Note the overall shape of the lines on this graph. Each denomination on which we have data over the longer term was relatively highly concentrated before 1930, and concentration everywhere decreased until about 1970, after which it started to increase, again everywhere. The earlier period of relatively high concentration is interesting and unexpected, but I linger on the increase since 1970 in order to convey what this graph reveals. In every denomination on which we have data, people are becoming increasingly concentrated in the very largest churches, and this is true for small and large denominations, for conservative and liberal denominations, for growing and declining denominations.

This increased concentration may have social and political implications. Increased concentration of members makes religious institutions more visible, since one 2,000-person church is more visible—if only because of the size of its building—than ten 200-person churches. Increased concentration also increases the potential for social and political influence. Since one 2,000-person church is easier to mobilize for social or political action than ten 200-person churches, a politician is more likely to address one 2,000-person church than ten 200-person churches, and the pastor of one 2,000-person church probably gets an appointment with the mayor more easily than any of the ten pastors of the ten 200-person churches. Increasing concentration also affects intradenominational politics and the development and diffusion of worship practices. Also, increased concentration can create the illusion of religious revival when in fact it’s the social organization of religion that has changed, not the overall level of religious participation or commitment.

Explaining this concentration trend is more difficult than documenting it. The usual ways of explaining the rise of megachurches do not quite work, at least not in their simple versions.

It sometimes is said that the secret to megachurch success is that megachurches have figured out how to attract the unchurched. But overall church attendance is not increasing. The only study I know of that compares very large churches with smaller churches concludes that there is no difference between the two in the percentage of new members who were not previously involved with a church. So the increasing concentration of people in the very largest churches is not a consequence of megachurches tapping into a previously uninvolved population. Increased concentration is occurring mainly because people are shifting from smaller to larger churches, not because people are shifting from uninvolvement to involvement in big churches.

Nor can we explain this trend by reference to some constant advantage of size. It is true that there are certain attractions to worshiping as part of a big group. But the advantages should have been apparent long ago, with the appearance of the first big churches, not just since 1970.

Recall the shape of the concentration trend. It looks like someone flipped a switch in 1970, reversing 70 years (or more) of deconcentration and initiating a trend in the opposite direction that continues today. We need an explanation that accounts for change in the 1970s and 1980s, not one that implies long-term, steady increase in concentration.

A third possible explanation invokes suburbanization, the proliferation of automobiles and decreasing travel costs. The problem here is that the trend lines do not line up properly. American society became steadily suburbanized throughout the 20th century, with the fastest suburbanization occurring between 1945 and 1970. American society continued to become increasingly suburban after 1970, but at a slower rate than before. Thus, the most rapid suburban population increases were accompanied by decentralization of religion, perhaps because major denominations self-consciously built medium-size churches in the burgeoning mid-century suburbs.

Similar patterns exist for the increases in paved roads, the proportion of households with cars, and the costs of operating a car. These trends were all strongest before 1970, which suggests that religious concentration should have started to increase decades sooner than it actually did. We could posit a 25-year lag or a threshold that suburbanization, car ownership and travel costs had to reach before generating religious concentration. This possibility should not be ruled out, but to make such an explanation compelling requires confronting this inconvenient misalignment in the timing of an effect and its purported cause.

A fourth possible explanation is that megachurches are a new, innovative organizational form designed by religious entrepreneurs, perhaps those associated with the church growth movement, who were particularly attuned to post-1970 society and culture. The main problem here, however, is that in key ways today’s megachurches are not a new organizational form. Very large churches in the 19th and early 20th centuries exhibited many characteristics typical of today’s megachurches—rapid growth under a gifted leader, high-quality preaching and music, multifunction buildings, extravagant theatrical displays, many and varied small groups and activities, and auditorium-style worship spaces with stages instead of pulpits and little or no Christian symbolism. We cannot attribute the post-1970 proliferation of megachurches to the invention of a wholly new organizational form.

A fifth possible explanation is that recent cultural change has given a new advantage to an organizational form that has existed for a long time. Perhaps people today—with baby boomers leading the way as usual—are more comfortable with bigness, more attracted to spectacle and more drawn to a church in which they can choose to be anonymous, or in which they can choose between anonymity in a big crowd and intimacy in a small group.

It is difficult to rule out some version of this argument, but it also is difficult to rule it in. It is easy to generate plausible hypotheses about cultural change, but it is difficult to specify a relevant cultural change and then point to evidence for that change that is independent of the rise of megachurches itself and therefore not question-begging. Making a cultural-change explanation persuasive requires finding that missing link.

I suggest another explanation: the increased concentration of people in the very largest churches is caused in part by rising costs that make it more and more difficult to run a church at a customary level of programming and quality.

Churches suffer, I think, from “Baumol’s cost disease.” This is a phenomenon identified in the mid-1960s by economists William J. Baumol and William G. Bowen. The basic idea is simple: if there is increasing productivity and efficiency in some sectors of the economy, and if wages increase in those sectors, then wages also will increase in other sectors, or else talent will move to the sectors in which wages are increasing.

However, some kinds of activities cannot be made much more efficient. It probably takes about as much preparation and effort to produce Hamlet or perform a Beethoven symphony in the 21st century as it did centuries ago. Activities that have at their core human effort, training, practice, attention and presence cannot be made much more efficient. No technological invention or social innovation makes it possible to reduce the level of input into such activities and still get the same level of output, so enterprises organized around such activities cannot be made more efficient without a reduction in quality.

Churches are subject to Baumol’s cost disease. Like schools, universities, theater companies and symphony orchestras, churches face ever-rising real costs with no significant opportunities to reduce those costs by becoming more efficient. The only options in such a situation are to sacrifice quality or increase revenue.

Rising real costs are not a problem for an enterprise, not even one for which efficiencies are unavailable, so long as revenues keep increasing at the same rate that costs increase. Congregations have followed that pattern for a long time. Americans are famously generous to their churches, and financial data from 20 denominations between 1920 and 2000 show that the amount of money gathered by the average congregation went up in real terms every decade since 1920, except during the 1970s when unusually high inflation eroded any real increase in donations. If you looked only at this trend, you would think that the average church ought to be in reasonably good financial shape, since churches enjoyed ever-increasing real revenue streams throughout the 20th century.

But increases in donations are not often compared to the rate at which the costs of running a church have increased. Beginning in about 1970 the rate at which donations increased stopped keeping pace with the rate at which the costs of running a church increased.

It is difficult to establish this claim definitively. In an initial effort to track the relationship between rising revenue and rising costs, I used the inflation-adjusted median salary of a full-time male pastor as a proxy for cost. (This is not a perfect measure, especially not after 1970, when increases in the cost of benefits started to outpace wage increases.) Comparing real revenue increases with real cost increases in this way suggests that from 1940 to 1960 increases in revenue for the average congregation far outpaced increases in clergy salaries. But that gap narrowed considerably from 1960 to 1980, and the lines crossed between 1990 and 2000, indicating that median real clergy salary increased at a higher rate than real donations between 1990 and 2000. These results are not definitive, but they do suggest that churches faced a qualitatively different, more severe challenge to balancing revenues and costs after 1960 or 1970 than they faced before 1960. And the high inflation of the 1970s probably exacerbated this situation.

When cost increases outpace revenue increases, churches cut corners and reduce quality by deferring maintenance, declining to replace youth ministers when they leave, replacing retiring full-time ministers with half-time pastors, and so on. In short, churches find it difficult to maintain the same level of programming and quality they had before. And this will be true even if the church loses no members. If costs rise faster than revenues, a 200-person church will be unable to produce the same level of programming and quality it produced before, even if it remains a 200-person church. Moreover, the minimum size at which a church can be economically viable will increase. The result is that people will be pushed out of smaller churches that no longer meet their minimum standards and into larger churches that still do.

Scholars and journalists who have written about megachurches have focused almost exclusively on the attractions of large churches. The rising-cost explanation calls attention to forces that are pushing people out of smaller churches as well as the factors pulling them into large churches. Since newly large churches are populated mainly with people who previously attended smaller churches, understanding increasing religious concentration means understanding what is behind the flow of people from smaller to larger churches.

This cost-increase explanation for increasing religious concentration remains speculative, but the concentration trend itself seems incontrovertible. Like it or not, we are in the midst of a significant change in the social organization of churches, one whose causes and consequences we do not yet fully grasp.

This article is adapted from a longer essay in the Review of Religious Research (June 2006).