Much of what we know—or think we know—about American religion is shaped by surveys and polls. The news media loves these polls, and so do many pastors. The latest instance is a Gallup poll showing that, for the first time in eight decades of polling on the issue, less than 50 percent of Americans “said they belonged to a church, synagogue or mosque.”

At first glance this would seem to confirm the trend toward religiously unaffiliated Americans, or the “nones.” And there is no reason to doubt that the rising number of nones is part of the story here. Even evangelical churches like the Southern Baptist Convention, once seen as impervious to the cataclysmic declines affecting the mainline denominations, have been on a pattern of (slow) decline for more than a decade.

But, as I have suggested before, we should take religion polls with a grain of salt. They usually tell us about some trends on the religious landscape, to be sure, but they are almost always open to widely varying interpretation. Polls are at their best when there is little wiggle room for interpretation in the data. For example, you can be pretty certain what the answer means when a pollster asks, “Did you vote for Joe Biden?”

Any time the answers are qualitative, as they usually are with religion, we should be more skeptical. Even something like “church membership” is an elusive category. Does that mean you consider yourself part of a denomination, or a local church? A religious tradition, or a particular congregation?

Gallup lists the actual question as “Do you happen to be a member of a church, synagogue, or mosque?” I imagine that the implication of this question might be clearer for “synagogue or mosque” than “church,” which to some respondents could mean a denomination or religious tradition, although the pollsters seem also to have asked about respondents’ “affiliation” with a religious tradition.

Still, for more parish-based Christian traditions, “being a member” might mean “I was baptized there, married there, and I expect my funeral will be there.” But that would not exactly make you an “active” member of a congregation.

There are also plenty of people who actively attend churches who are not members, either because membership is not emphasized, or because they have some doctrinal or denominational hang-up that keeps them from joining. Thus, you might have many people who identify as “members” but who rarely attend, while others know they are not members, but they’re present every Sunday.

As my Baylor colleague Philip Jenkins recently noted, the phrasing of religion questions can produce wildly different results. Recent polling questions about religious affiliation in the U.K. were phrased slightly differently, and produced data showing that either 50 percent of British people have “no religion,” or 25 percent. That’s a massive difference!

Then there’s the problem that people who may be members of a certain church may not have attended there for years. (We can at least assume that the poll is not reflecting the thousands of people who are still on church membership lists even though they are, in fact, deceased!) “Member” ideally should connote activity, but it doesn’t always mean that, as every pastor knows.

There is also a small but significant number of people who tell pollsters that they do not have a religion, but they are church members. The number of such people is usually about 5 percent to 10 percent of the so-called “nones.” This is a puzzling group, to be sure, but some of them are presumably Christians who would say “I don’t have a religion, I have a relationship with Jesus.”

Finally, there is polling’s dirty little secret: response rates. The Gallup poll is typical in that it does not seem to publicly report its response rate, but you can assume that it was less than 10 percent, if not less than 5 percent. Response rates in the era of landlines used to exceed 85 percent. Pollsters will insist that they have ways to account for this problem. We can assume that Gallup, which is one of the most reliable agencies, is trying their best to do so. Still, the people being polled are the types of people who respond to polls, a vanishingly rare breed today.

But to return to our question: if church membership rates are cratering, why is this the case? This poll undoubtedly was affected by the pandemic, although a lot of the data seems to come from the pre-pandemic years of 2018-19. However, the pandemic will have made the question of church membership even more muddled than before. If a person worships online every week, but they never physically attend a service, are they a member? This issue will not go away once COVID subsides.

Next, it seems clear that with every passing year, more senior adults are dying who assumed that you should be a church member to be a good citizen and a good American, even if a person is not particularly devout. They are not being replaced by younger Americans who share that conviction. The number of people who attend church by default is apparently plummeting. Whether this is mostly good or bad for congregations is unclear. It might be bad for numbers and finances, but it will probably lead to a more uniformly committed church body in many cases.

Virtually all polls show lower rates of religious affiliation for younger generations, eliciting the typical concern that the younger generation has “left the faith.” (Hang around churches very long and you will hear dicey statistics touted about how huge numbers of kids in youth ministries will “leave the faith” in their 20s.)

But virtually all polls are also snapshots of people at one moment in time, and you can bet that large numbers of those young people—especially those who grew up in church—will become active church members again, often at life transition points such as marriage, or when they start having kids. If you really want to worry about a demographic trend, note the plummeting rates of marriage and especially childbearing in America, statistics that have a strong correlation with church attendance.

The overall picture of declining church membership should be of interest, but not special worry to Reformed and evangelical believers. We’re not so much concerned with “mere” church members, but “regenerate” church members. And evangelicals have been at their best—such as during the First and Second Great Awakenings—when they had to work hard at drawing people into church with crystal-clear proclamation of the gospel, and with caring service to the needs of congregations.

A lot of the church decline has to do with the death of cultural Christianity in America. This development is of concern, of course, and may encourage increasingly prominent cultural roles for rabidly anti-Christian views. There is also a risk that churches may supply an ever-dwindling amount of “social capital” that has been a major benefit for attendees since the colonial period. Church attendance, on balance, leads to better outcomes in gauges of human flourishing, such as family stability, supportive friendships, charitable giving and service, and so on.

On the other hand, the death of cultural Christianity represents an opportunity for evangelical churches, as Russell Moore and others have pointed out. As Moore has argued, much of that so-called “Christianity” is really just “a means to an end—even if that end is ‘traditional family values’—[and] is what J. Gresham Machen rightly called ‘liberalism’ . . . it is an entirely different religion from the apostolic faith handed down by Jesus Christ.”

If nominal, utilitarian, civil-religious “Christianity” is mostly what’s fading away with the cratering of American church “membership,” then I say good riddance.

Sign up here for the Thomas S. Kidd newsletter. It delivers unique content only to subscribers.