Birth dearth: Demographics of mainline decline

October 4, 2005

Since the publication of Dean M. Kelley’s classic study Why Conservative Churches Are Growing in 1972 it has become axiomatic for many people that the mainline denominations (Methodist, Episcopal, Lutheran, Presbyterian) are in decline while conservative denominations (Southern Baptist, Assemblies of God, Pentecostal and holiness churches) are growing. Most observers of this phenomenon—such as Jeffery Hadden, Richard John Neuhaus, James Davison Hunter and Thomas Reeves—assume that the reason conservative churches are growing is that they are attracting converts from the mainline churches. The conservatives, it is said, have a strong appeal because of their emphasis on traditional teachings.

Some scholars have dissented from this view. Wade Clark Roof and William McKinney in American Mainline Religion (1991) note that higher fertility rates and a younger age distribution give conservative denominations a demographic advantage. Robert Wuthnow in his book Christianity in the 21st Century (1993) and Darren E. Sherkat in his article “Counterculture or Continuity? Competing Influences on Baby Boomers’ Religious Orientations and Participation,” in Social Forces (March 1998), also take note of the demography of denominations. However, none of these authors quantifies the relative contributions of demography and church switching to mainline decline. The notion that mainline decline is due to excessive liberalism has remained alive and well.

For example, in The Empty Church: Does Organized Religion Matter Any More? (1998) Thomas Reeves asserts that a major reason for the numerical decline of mainline churches is their failure to retain their own children as members once they reach the age of decision. He approvingly quotes Stanley Hauerwas: “God is killing mainline Protestantism in America and we goddam well deserve it.” Reeves adds that Kelley was correct when he said the strict demands of faith and practice that the conservative churches make on their members are the reason for their success. He contends that “liberal Protestantism . . . has become so secularized and indistinct that it cannot compete successfully.”

Our purpose is to dispatch this “excessive liberalism” argument and supplant it definitively with the demographic one. Using data from the General Social Survey (arranged by cohorts) we can confirm that the mainline’s share of the Protestant population declined from 60 percent to 40 percent across cohorts born between 1900 and 1975. But the conversion rate from mainline to conservative denominations remained steady at around 13 percent for all those cohorts. An explanation for the decline of the mainline must be sought elsewhere.

The surest source of new adult members for a denomination is in the pews already—among the children of the current members. For most of the 20th century conservative women had more children than mainline women did. In fact, the existence of larger conservative families explains 70 percent of the mainline decline (recast as conservative growth). Further analysis shows that the remaining 30 percent of the decline came from a precipitous drop in conservative-to-mainline conversions.

We can posit other explanations for the conservatives’ increase: a higher apostasy rate for mainliners than for conservatives, and a greater inflow from outside Protestantism to the conservatives. We tested these possibilities, and there was no evidence that either of these differentials was large enough to account for as much as 1 percentage point of the denominational shift. The fertility explanation seemed to hold the most promise.

The so-called demographic imperative means that in a population made up of two groups, the one with the higher rate of natural increase will increase its share of the total at the expense of the group with the lower rate of natural increase, other things being equal.

Conservative fertility has indeed been higher since the early days of 20th century, when conservative women had on the average about one more child than did mainline women (see the graph on the next page). The two groups converged somewhat during the baby-boom years, when mainline fertility increased dramatically. In the years after the baby boom, the mainline rate declined earlier than did the rate of the conservatives. Only in recent decades has the fertility of the two groups become similar. The law of the demographic imperative predicts that under such circumstances the conservative share of the population will increase unless some other factor counters it.

We created a model that allowed us to calculate what the decline of the mainline would look like if the higher fertility of the conservatives were the only influence on the relative size of the two types of denomination. We compared our hypothetical curve with a second curve which describes the actual proportion of Protestants who were in the mainline denominations during the first 75 years of the 20th century (see the graph on page 27).

What would the Protestant population look like if nothing but demographic rates changed? The answer is that it would look remarkably like it does in real life. The demographic model accurately predicted the sharp downturn in mainline population through the baby-boom cohorts. Our model’s only error is that it predicted that the mainline decline would level off ten to 15 years sooner than it actually did.

Thus the shift of the Protestant population away from the mainline toward the conservative denominations is overwhelmingly due to conservatives’ higher fertility, which accounts for 70 percent of the change.

The declining propensity of conservatives to convert to the mainline accounts for the 30 percent of mainline decline that fertility rates cannot account for. As intermarriage of conservatives to mainliners has declined across the century from 33 to 20 percent, most conservative exogamists have remained conservatives, and this propensity has increased in recent birth cohorts.

Additional aspects of conservative-to-mainline conversion—most notably those associated with intergenerational social mobility—have also slackened. Overall, conversion from conservative to mainline denominations fell from 30 percent among the earliest cohorts to less than 10 percent among those raised conservative in the late 1980s and early 1990s.

Higher fertility and better retention thus account for the conservative denominations’ rising share of the Protestant population and the corresponding decline in the proportion of Protestants who belong to mainline denominations.

The various popular “culture wars” and “strictness” theories which have been used to explain the decline of the mainline are irrelevant. Mainline denominations’ support for homosexuality and abortion and modernist interpretations of scripture do not account for the change. The raw power of demography does a much better job.

Why was the reproduction rate of the conservatives higher than that of the mainline for so many decades? Presumably the opposition of the mainline leaders to laws banning contraceptive devices (prevalent nationwide in the 1920s and 1930s) reflected or influenced child-bearing practices. Thus beneath the demographic explanation there is also a cultural one, though not the cultural explanation that most of the “culture wars” writers have advanced: it took most of the 20th century for conservative women to adopt family planning practices that have become dominant in American society. Or to put the matter differently, the so-called decline of the mainline may ultimately be attributable to its earlier approval of contraception.

Predicting the future is precarious at best, but our evidence suggests that the trends underlying the mainline’s decline may be nearing their end. The demographic momentum, as it affects cohorts, is spent. Our model predicts less than a 1 percent decline in the proportion of Protestants who belong to the mainline among cohorts that will come of age during the next decade. Unless conservative Protestants increase their family size or mainline Protestants further reduce theirs, this factor in mainline decline will not be present in the future.

The other key predictor—the falling rate of switching from conservative to mainline denominations—is reaching an end point of its own. Having fallen from 21 to 9 percent, the conservative-to-mainline switching rate cannot continue to fall much further, simply because it cannot fall to less than zero. Exhaust both sources of change and change will stop unless and until a third variable comes into play.

Nonetheless, a word of caution is in order. We have focused on cohorts because changes in behavior show themselves most clearly in the succession of cohorts. We need to keep in mind, though, that the cross-section of Protestants in any particular year for the next half century or more will still include the people born during times when there were different fertility rates. The demographic momentum of differential fertility will remain present until the cohorts born in the early 1970s—who are in their 30s in this decade—pass away. So the Protestant population will continue to shift in the conservative direction for many years to come, even if no further changes in underlying behaviors occur.

Michael Hout, Andrew Greeley and Melissa Wilde are sociologists at, respectively, the University of California-Berkeley, the University of Arizona, and Indiana University.