What makes churches grow?
This is a question that some of the folks who work here at the Presbyterian Center are concerned about.
It’s also a question of interest to many folks around the country involved in religious congregations or even in secular organizations, like ours, some of which are facing membership decline.
Finally, it’s a question that receiving a church newsletter in the mail a couple of months ago really hammered home to me.
For a couple of years my family was active in a West Central Florida mainline Protestant congregation: Kirkwood Presbyterian Church, in Bradenton, Florida.
Founded in 1980, Kirkwood had experience solid growth throughout most of its short history, including during the two years that my family was there.
But a couple of years ago the church’s founding pastor, Pastor Bill – the only pastor the church had ever known – retired.
It probably should have come as no surprise to me to read in this newsletter about the church’s financial problems.
I also went to the Presbyterian Church (U.S.A.) Web site and found that Kirkwood has suffered a couple of membership dips in recent years, including a rather sharp one in the past year.
Knowing that even mighty Kirkwood was suffering membership and growth problems renewed my interest in understanding church growth.
I’ve talked with a number of you all about the work my Research Services colleagues and I have done trying to identify factors that influence how much churches grow. We’ve looked particularly – but not exclusively – at Presbyterian Church (U.S.A.) church growth.
These are some of the sources of data we’ve used.
- Census 2000 data gives us information about the characteristics of the communities in which congregations are located.
- For Presbyterian Church (U.S.A.) congregations, the Session Annual Statistical Report gives valuable information about congregational characteristics, including membership totals.
- For the 500 or so Presbyterian congregations that participated in the 2001 U.S. Congregational Life Survey, responses by worshipers to the U.S. Congregational Life Survey worshiper’s survey give vital information about congregational characteristics.
- Finally, responses by congregational leaders to the U.S. Congregational Life Survey congregational profile and leader survey give information about characteristics of pastors.
One of the first steps researchers would undertake with these data would be to use a statistical technique known as regression analysis. In its simple form, regression analysis assesses to what extent the values for one variable predict the values for another variable.
Here are the simple regression analysis results for a half a dozen of the factors that best predict net percentage membership change between 2000 and 2005. I want to direct your attention to the top three factors in particular.
The half a dozen plus factors and the standardized single regression coefficients for them are:
- Prioritization of children’s and youth ministries score: .274
- Percentage of worshipers age 65 or older: -.26
- Shared leadership practices score: .241
- Prayer as percentage of time the leader spends in paid ministry during a typical week: -.146
- Salary and housing benefits: .145
- 1995-2000 net percentage membership change: .142
- Percentage of residents age 25 or older in the zip code in which the congregation was located with bachelor’s degrees: .122
- Total membership: .104
- Leader’s age: -.052
Prioritizing children’s and youth ministries is a scale that combines responses to several different U.S. Congregational Life Survey worshiper’s survey questions. Congregations that prioritize children’s and youth ministries are more likely to grow.
You’ll note the negative sign next to the statistic for percentage of worshipers age 65 or older. That’s because there is a negative relationship. Congregations with a lot of older worshipers are LESS likely to grow. Younger congregations are MORE likely to grow.
Finally, shared leadership practices is another scale that combines responses to a number of different worshipers’ survey questions. These include questions about how good a match you think there is between your congregation and your pastor and how much church leaders pay attention to people like you when they’re making decisions.
The next step would usually be to run a multiple regression analysis. This technique includes values for a bunch of variables and asks how well, collectively, they predict – in this case, church growth. It also sorts through the individual variables and tells you which ones predict best.
As soon as you try multiple regression analysis with these data, which we have done, you run into two challenges. First, the best we’ve able to account for is about a quarter of the growth. That’s not terrible, but it’s not great. For us, Presbyterian church growth still remains somewhat unpredictable.
Second, the values of many of these variables are interrelated. And that’s really a no-no in multiple regression analysis. For example, it turns out that congregations with a lot of older people tend NOT to prioritize children’s and youth ministries. If you throw caution to the wind and include both prioritization and percentage of older people in a multiple regression, older people drops out.
It’s not crystal clear how to interpret this. But two things are clear:
Prioritization of children’s and youth ministries is a better predictor of growth than percentage of older people; and
There are some congregations with a lot of older worshipers that DO prioritize children’s and youth ministries, and there are some younger congregations that do NOT.
To illustrate, let’s go back to Kirkwood. This is the inside of the sanctuary at Kirkwood – obviously, when there aren’t any people in there.
Kirkwood is a snowbird church. Bradenton is a big retirement area. In the summer Kirkwood might have a couple of hundred people in worship. In the middle of the winter, in both services put together, there might be over 1,000 in worship.
This was one of the most memorable Kirkwood rituals while we were there. At the start of worship Pastor Bill would walk around the room with a cordless microphone and ask newcomers to introduce themselves. But he’d also ask oldtimers returning after a while to identify themselves. He might give me the mike and I might say: “I’m Bill Adams, and this is my wife Martha. It’s great to be back at Kirkwood. We’ve just gotten back to town from Baraboo, Wisconsin. Go, Packers!” And all of the Wisconsin people would clap. Actually, lots of people would clap.
You might think that a congregation with this profile would NOT emphasize children’s and youth ministries. But you would be wrong. I know more about the youth ministries. While we were there, there were active middle school and high school youth groups and bustling Sunday school classes. The church paid, modestly, several talented, enthusiastic people to help lead these. And the church had my son – not always the most enthusiastic church participant – help out with the sound system for the 11 a.m. services. This was one way they involved children and youth in worship.
What does knowing a little bit about the church growth data and the Kirkwood experience tell us?
Demography is not destiny. “Objective” factors only go so far in influencing church growth. Just because you’ve got a pastor who’s the right gender, or a congregation that’s the right age, or a community with the right education level – doesn’t necessarily mean your church is going to grow. An old people church might CHOOSE to prioritize children’s and youth ministries, which we know can be a factor in growth.
There’s a little bit of a Is the glass half-full or half-empty issue here. A colleague of ours, knowing these data, has quipped that: Presbyterian churches – even those in advantageous situations – still find a way to screw up growth. I prefer to look at things the other way around. Even for churches that are in “objectively” disadvantageous situations, there is still room for maneuver, still opportunities, still possibilities for growth and decline.
I’d like to think that – if Kirkwood Presbyterian Church – on Easter morning 2005 – in that sanctuary – could somehow persuade my son to be baptized as a Christian and to be confirmed – churches around the country – of whatever denomination, in whatever situation – should be able to find a way to grow.
I look forward to hear your comments and questions, starting now.
P.S. What might this mean for Toastmasters clubs like ours? I don’t think you can apply lessons about church growth one-to-one to the experiences of secular organizations. I do, however, see a couple of possible implications. Our club has people with a pretty broad range of ages in it. I have, however, seen clubs with mainly younger people or mainly older people. Judging from these data, we might become a little nervous about the growth prospects of a club with mainly 80-year-olds. It’s not exactly clear why having a lot of older people might be bad for church growth – or what a older club should do about that – but there does seem to be a connection.
With shared leadership practices, it seems clear that – as with churches – in a club in which a lot of different people are involved in activities and club decision-making – and that all the offices are filled, and the officers are all active – that this kind of club would be more likely to grow. Let’s stay active and stay mixed-age if we want to grow!
-- Perry
