Printable Version
VOL 2, No 6
March 2000
What Do You Mean by Average?
by Arthur E. Farnsley II
Over the past two decades, congregations have become an important focus of the social study of religion. Analyses of individual belief are yielding to studies of group practices, resources, and organization.
More recently, congregations have featured prominently in discussions about civic life. Since the passage of welfare reform legislation in 1996, congregations, secular non-profits, and government agencies have made a concerted effort to develop formal, programmatic partnerships. Government and foundation officials have invited congregations to play a larger role in social service delivery and community development. Their premise is that congregations can assume, to some degree, the role played by government in providing welfare and building civic community.
The Polis Center�s Project on Religion and Urban Culture has received several requests to describe what congregations are doing and to estimate their available capacity, based on our research. Our initial analysis of data collected on more than 300 congregations in Indianapolis suggests that current discussion about available capacity exaggerates what congregations can actually do.
Baseline Data
If congregations are the question, how does one begin to supply some answers? While people may know a lot about their own congregation, few know much about other congregations. In particular, most people know very little about congregations as organizations, and are unable to generalize meaningfully. Even large differences among organizational types�such as the fact that Catholic parishes tend to be arranged geographically while Protestant churches are usually not, or that inner-city neighborhood have many more, much smaller congregations per capita�are not part of the public discourse.
In the course of everyday conversation about religion, or in the day-to-day practice of faith, such questions matter little. But they matter a great deal in civic affairs, as well as in sociological analysis. If congregations are to be treated as organizations with the potential to deliver human services or build civil community, then we must ask significant questions about their capacity to perform these tasks.
Having spent three years trying to answer such questions, we know that good data on this subject is hard to get. We have done some survey work, but more of our data collection has been face-to-face, rather than by phone or mail. We have attended worship services and meetings to get reliable estimates of attendance and involvement. We have talked to everyone who would talk to us.
The deeper one gets into the data, the keener one�s insights become about the problems with everyone else�s data. (The limitations of one�s own data every researcher knows already.) For instance, Ram Cnaan, professor of social work at the University of Pennsylvania, conducted research on Indianapolis as part of his excellent study on behalf of Partners for Sacred Places. Cnaan analyzed the work of 25 congregations housed in historic buildings, looking at every form of community service they provided, including providing space for public meetings, volunteer help, and cash outlays. He concluded that, on average, the congregations he observed put $144,000 per year in cash and services into the local community. Of that $144,000, $33,000 was actual cash outlay built into the budget. The average congregation in his study had 517 members.
On the basis of the Cnaan study, The Indianapolis Star ran an editorial claiming that the average congregation in Indianapolis spent $140,000 on the community.
After studying 113 congregations in six cities�including 25 in Indianapolis�researchers have put a price tag on the community services provided by urban congregations...The average congregation spends $ 140,000 a year to subsidize its community programs and provides over 5,300 hours of volunteer support.
__"Worth of the Church," Indianapolis Star, December 5, 1997
If this were true, the 1200 congregations in Indianapolis would be contributing more than $170 million dollars per year, in cash and services, to the community. Our data suggested this sum could not be accurate.
Another example: about a year later, Elizabeth Boris, a researcher from the Urban Institute in Washington, D.C., presented results from an Urban Institute study of faith-based service providers in the nation�s capitol. It found that the median amount spent by faith-based providers was $15,000. The median size of the congregations, which made up the bulk of the faith-based providers, was 400 members. By this time, we had proceeded far enough in our research to be suspicious of a median membership of 400. And that made us wonder, too, about the $15,000 number.
Samples, Means and Medians
Whatever else might be said about the numbers collected by Partners for Sacred Places and the Urban Institute, they drew conflicting pictures. The discrepancy in the "cash outlay" figures�Cnaan�s $33,000 versus Boris�s $15,000�was significant enough to make us wary of making broad generalizations based on either study.
The Star made the mistake of generalizing to all congregations what Cnaan had claimed about congregations housed in historic buildings. Those historic congregations, it turns out, were overwhelmingly more Catholic and mainline Protestant than congregations in general, so they were skewed toward groups with more members, and with a higher social class and greater income.
Although the researchers were careful to circumscribe their conclusions, editorialists and others leaped to conclusions that were unfounded and even untrue. We knew that in the service world of model programs and best practices, people were pointing to the very few large, well-organized programs and saying, "This is a model." In the terminology of sociology, those cases were "anecdotal." We were concerned that atypical congregations in specific social circumstances were being used to shape the activity of congregations that had very different circumstances. In a more insidious way, the raw numbers themselves were creating problems.
We have looked at 300 congregations in 17 neighborhoods of Indianapolis. Our sample is representative of each kind of neighborhood and of the 1200 congregations in the city as a whole.
We have found that the mean congregation has roughly 400 members. That is, when we added up all the congregational members in Indianapolis and divided by 1200 congregations, we got 400 members per congregation. This is what most people mean by "average." However, the median congregation�midway in size between the largest and smallest congregations�has closer to 150 members. Put another way, fully one-half of all the congregations in Indianapolis have fewer than 150 members.
These distinctions are important for two reasons. First, a few very large congregations are propping up the mean average. Indeed, one must reach the 70th percentile before congregations have 400 members or more. This means that only 30 percent of congregations are at least as big as the "average."
Second, this discrepancy between mean and median teaches us to be careful when we hear about averages. If the median size of congregations in the Urban Institute survey was 400, then half the congregations in that survey had more than 400 members. That means the survey was full of congregations much larger than we have found in Indianapolis. Although this does not prove that the $15,000 spending number is correspondingly too high, it suggests caution. The same might be said of the 517 members in the mean congregation studied by Partners for Sacred Places.
We noted the same differences in ways of looking at congregational budgets. The mean congregation in our sample has an annual budget of roughly $260,000. But the median budget is only about $125,000�i.e., half of all congregations have budgets below $125,000 annually. Only one in five have budgets exceeding $250,000. Again, only the top 20 percent are at least as big as the "average."
Other questions concerning congregational capacity followed suit. The mean number of full-time, paid staff at a congregation was 2.75. But the median number was, as one might expect, 1. The practical realities lying behind this statistic are telling. Fully 30 percent of congregations have only one full-time, paid staff member, but another 27 percent have none. Another 12 percent have two. This means that nearly 70 percent of congregations have zero, one, or two full-time staff members, yet the mean is 2.75.
Numbers, numbers, numbers�
Our research reveals that the top tier of congregations�the largest one-fourth or one-fifth�have a very large share of the members and control a very large share of the money. As reported in an earlier Research Notes, in some urban neighborhoods the single largest congregation accounts for as much as 90 percent of all social service spending by congregations. In the total sample, only 20 percent of congregations spend as much as the mean congregation spends on social services. When only one-fifth of a group is "average" or above, there is something misleading about the term.
Of course, a host of other variables influence the definition of typical or average. We arranged our data by theological groupings and found, not surprisingly, that mainline Protestant churches are larger than either the mean or the median and that Catholic churches are much larger�generally nine or ten times as large as the median of all churches. Their budgets follow accordingly.
Our data suggests that everyone interested in congregations as organizations�from civic leaders to readers of the morning paper�should ask hard questions about any information in which the majority of the congregations being discussed are Catholic or mainline Protestant. These data are not inherently wrong, but they cannot be used to generalize usefully about other congregations. In Indianapolis, for instance, only 300 of the 1200 congregations are from the Catholic tradition or the Protestant mainline. These groups combined make up one-fourth of the congregations but have roughly half of all members.
The same problem applies with evangelical and independent congregations, which include a few megachurches with several thousand members. There is good reason to consider treating megachurches and Catholic parishes in separate categories; the size of their memberships and budgets, as well as the range of their programs, obviously distort any discussion of what constitutes "average."
Why is this significant?
These questions about relative size and resources should help civic leaders think more constructively about the role congregations can play. The lesson is not that only big, rich groups can get anything done; small congregations can and do have substantial impact. The lesson of means and medians is to use caution in making generalizations or in recommending "model" programs.
When congregations think of themselves as organizations�perhaps when they are making strategic plans or evaluating their annual efforts�they should maintain a realistic frame of reference. Are they smaller than average? Do their members give more or less than the average? How do their community ministries stack up? Serious answers to all these questions require narrowing the frame of reference so that the "average" used for comparison is meaningful.
It sounds tempting for congregations to say, "We should just be ourselves and not worry about how we compare to any other group." But other kinds of organizations don�t behave in this way, nor should they. Any human activity conducted in the presence of others uses those others as a mirror by which actions are observed, evaluated, and modified. Most congregations are not islands. They want to relate to the world in ways that make sense for themselves and for others, and that requires making comparisons.
Like all organizations, congregations can use data to think critically about their activities. But they must ask how the available information applies to them. Congregations benefit most from data when they can figure out what the generalized information tells them about their specific situation. The question is not so much, "How are we like or unlike the average?" but "What range of groups does that average include and what does that information mean for us?"
Conclusion�����
When researchers first turned to congregations, they looked at them in business management terms. But a renewed emphasis on their role as community-serving, community-building organizations requires better research into congregations as distinct types of organizations. The Polis Center�s Project on Religion and Urban Culture is still compiling and analyzing data, but we feel confident that we can draw a detailed and accurate picture of the situation in Indianapolis.
We hope that what we are learning about Indianapolis has value for other venues. Other cities, considering their own circumstances, may or may not find Indianapolis to be an appropriate yardstick. But while our particular findings are likely to be context specific, the questions we are learning to raise must be raised everywhere.
We can accurately describe the mean and median sizes of congregational memberships and budgets in Indianapolis, but the numbers require interpretation. The majority of congregations are smaller and have less money than early studies would lead us to believe. Denomination and theological tradition have a strong influence on capacity as well, but have yet to be sufficiently examined.
Some congregations play an important role in social service delivery and community building, and a similar capacity could be built in other congregations. But how many such congregations are there, and how can we know which congregations have the potential to do more? If we expect congregation to assume a larger role in providing public services, we must begin with realistic expectations, based on a fair accounting of the enormous breadth and variety among congregations as organizations.
ROUNDTABLE
On December 15, Research Notes hosted a roundtable discussion at the Indianapolis Center for Congregations. Participants had been provided beforehand with the text of this issue of RN, and were invited to respond to the issues raised in the paper. Diana L. Etindi is a research fellow in the Welfare Policy Center at Hudson Institute. Judith Cebula is religion writer for the Indianapolis Star. Art Farnsley, director of research at The Polis Center, wrote the paper under discussion. Kevin Armstrong is senior public teacher at The Polis Center. The following is an edited version of their discussion, which was moderated by Armstrong.
ARMSTRONG: Once again, congregations are taking a prominent role in the larger community. This brings us to the question: What is an "average" congregation? Are congregations like the children of Lake Wobegon, where "all are above average"? What are the implications for public policy? Art has presented in his essay a description of the average congregation in terms of attendance and budgets. He also speaks of why the distinction between the mean and the median is important. Did Art�s figures surprise you, or confirm what you already knew?
ETINDI: I had read the Ram Cnaan study, and felt that it really skewed the statistics because of his focus on congregations in historic buildings. What Art presented was more accurate regarding the general church population.
ARMSTRONG: From where you sit, what are the implications of those kinds of measurements?
ETINDI: This is hard. My personal faith makes it hard for me to think in terms of congregational work as a business. I believe in a God who can multiply loaves and fishes and feed a multitude with very, very little. I don�t think that we should look only at numbers. Robert Birdsell, who is president of the National Center for Neighborhood Enterprise, has said that, "Too often we reduce our dreams to match our resources instead of expanding our resources to match our dreams." Congregations should look not only at their resources, but at finding resources. Then they could focus on their dreams and what God places on their hearts to do.
CEBULA: I tend to think more pragmatically. Numbers help me. A church that has 50 people showing up on Sunday is not out there saving the city. Churches like Second Presbyterian, or Community Church of Greenwood, with thousands of members can do work in the community. The numbers help me to understand what is doable. There are some very small ministries doing valuable things. But money matters, and the number of people matters. If we believe that prayer can inspire people to go beyond what they think is possible, there is power in numbers even there.
ARMSTRONG: Many in the religious community have said, "This is not how we choose to look at ourselves." But others will judge a religious institution by its capacity. Do notions of what is average matter to funders, community organizers, and neighborhood development folks?
ETINDI: Definitely. When you are operating under a budget, it helps to know what can be expected. It keeps foundations from having unrealistic expectations of congregations. It also helps people to know where they might partner. Just to use urban-suburban partnerships as an example�a suburban church can put its resources behind what smaller churches in the inner city are doing, and that is a great combination. Numbers can help people to recognize where the strengths and weaknesses are, and what gaps they might fill in for one another.
CEBULA: I wonder: is the fundamental role of the congregation spiritual and moral development? If a small church wants to try social service work, do we have different expectations of them as opposed to a 3,000-member congregation? Would it be unfair if we didn�t give them a shot at it? Or do we have a template, the large suburban church, for doing that work, and that is what we expect from even the small church?
FARNSLEY: One of the things that got me going on this was when the Indianapolis Star wrote the editorial after the Ram Cnaan study came out, saying that churches put $140,000 into the local neighborhood every year. The Star said, "There are 1,200 churches doing this, on every street corner and in every neighborhood." And you know what? That is not so.
ETINDI: Not even close.
FARNSLEY: Congregations do not like to think of themselves as organizations. They do not like organizational analysis of what they are doing; they don�t like financial analysis. It makes people uncomfortable to do sociological analysis of something they think of as primarily theological or philosophical. But if people are going to be talking about congregations as resources and neighborhood actors, and talking about faith-based partnerships, then we need data to form realistic expectations. I worry that politicians might sell the public a bill of goods about what churches are capable of. We may be setting churches up to fail. We are pointing to Tabernacle Presbyterian and saying, as Henry Cisneros said in Higher Ground, "Now, there is a model urban ministry." If that is the model, we are talking about maybe half a dozen churches in Indianapolis. What will everybody else do?
CEBULA: Certain types of congregations appeal to some people and not to others. I happen to not like big churches, personally. I prefer a smaller, mid-sized congregation. I prefer liturgy that is quieter. I will go to the 5:30 mass at St. Thomas and not the 10:30 on Sunday morning. So if I have a particular taste or feel a need for a particular kind of worship experience, what about the church that is right across the street from Tabernacle Presbyterian?
FARNSLEY: It is a faithful few.
CEBULA: "A faithful few." I love the contrast. Right across the street. If I am in that neighborhood and have a choice, do I want to be in that big church? No, I want to be in the smaller community. If the small church wants to have a pantry or some kind of outreach to the neighborhood, but they don�t fit the model, doesn�t that squeeze them out of doing good work in the community?
FARNSLEY: One benefit of this sort of analysis might be that it encourages groups that are looking to build partnerships, such as government, to think smaller. Governor O�Bannon went to Eastern Star to kick off the FaithWorks program. But if that is the model then no one can follow the model. There is only one Eastern Star. It is a magnificent place, a tremendous ministry, but others might think, well, if that is what the state is looking for, we don�t have much to discuss. A lot of congregations could administer a $25,000 grant, but they couldn�t possibly hope to do $250,000.
ETINDI: Is it that small congregations aren�t doing much, or that they aren�t in the habit of recording what they do? Congregations aren�t used to doing performance evaluations.
FARNSLEY: My own view is that less is happening than some people want to believe. Churches do have food pantries; they do give out clothing; they do all kinds of counseling on their feet. People just knock on the door. But these things are impossible to measure. People in ministries like these don�t think of them as administrative procedures.
ARMSTRONG: What does this mean, then, for the congregations? How can this kind of distinction help congregations to see themselves in relation to others?
FARNSLEY: There is no way to talk about capacity without narrowing it down to organizational capabilities, whatever the spirit that underlies it. But it worries me that congregations see Eastern Star on TV, and compare themselves unfavorably. They may think they are failing in some way because they have fewer members, or a smaller budget. In fact they may be typical for their type or denomination. When you see averages, you have to think of context. If you are a Methodist congregation you should be asking, what are other Methodist churches doing? If our data does mean anything to congregations, I hope it helps them to locate themselves on the spectrum; that it tells them more about themselves than some statistical "average."
ETINDI: It would be great if we had a study that broke congregations down into three or four membership levels: 200 or fewer, from 200-500, from 500-1,000, and more than 1,000. It would give congregations a clearer sense of how to compare themselves, in terms of what they might accomplish. If they judge themselves according to some studies that have been done, they see this huge contrast.
CEBULA: How could that research be done, though? How could you talk about analysis, especially with smaller churches that don�t have the structure?
FARNSLEY: The question is, what would you measure? Would you measure the number of people they served? I think a rejoinder to that would be, "No, we are talking about quality of service." A small church might provide the quality of service, the life-transformation, to three people that a big church didn�t provide to any of its 200 clients. But I say you can�t measure that.
CEBULA: Social scientists and journalists would say to do anecdotal stuff, look at four or five levels based on size. Do a story or two from each of those levels so that you give people a choice.
FARNSLEY: The ideal would be for a group to be able to think about model programs in terms of what is within its purview. "What other organizations are enough like us that we could use them as a model?" Religion is different from business in this regard. In business, at some point the big groups eat the small groups.
CEBULA: Especially lately.
FARNSLEY: When you buy a toaster you go where toasters are cheapest, even if that means that Bob�s Toasters goes out of business because of Wal-Mart. But it doesn�t work like that with churches; lots of people are very happy to go to Bob�s and be small and get personal service.
ARMSTRONG: We sort of dance around the notion that capacity is somehow related to theology. Denomination and theological tradition often influence the tendency of a congregation to be large or small. But from the outside I suspect that most civic funding and public policy folks pay less attention to those differences. Does it matter in understanding a congregation�s capacity?
ETINDI: Well, obviously, evangelical churches tend to have evangelism as their mission, as opposed to doing good works in the community. They consider their primary purpose to be the winning of souls�whereas other churches may consider their primary purpose to be contributing to the community, or doing good to their fellow man. The latter group would naturally incorporate more social service programs than the former.
ARMSTRONG: Did I hear you say�or almost say�that an evangelical church is more likely to be large?
ETINDI: No, actually I think that might not be the case. A number of the mainline churches have become quite large over the years, and tend to have more of a social mission. Some of the smaller groups have formed because they feel that the mainline churches are too socially oriented and have lost that basic mission of winning souls to Christ.
CEBULA: If I have grant money to give, or a contract to award, do I care if their ultimate mission is to convert or save? I think it does matter. We need to be conscious of it, if salvation is their fundamental motivation, not service. There is a distinction to be made, and I think it is an important one. It may be overt: "You need to come to Bible study before we give you a clothing voucher or job training." I think taxpayers would want to know if that was where their money was being spent. They may love the idea; they may be bothered by it. But what if a Muslim congregation were evangelizing Christians to become Muslim? Would some taxpayers be bothered by that?
FARNSLEY: Oh wait, I know that one. Yes! Some taxpayers would be bothered by that.
ARMSTRONG: We have talked about funders and public policy groups, but who else ought to be attentive to this kind of information? There are a variety of audiences that make assumptions about capacities, and who look at congregations as organizations. Can you think of others who are doing that currently?
CEBULA: Journalist make those assumptions all the time. What kind of Christmas stories do we write, what kind of Easter story, what kind of feel-good story do we do? The bigger church, the one with the public relations office, gets on our radar. Seminarians or people considering entering the ministry need to realize what is going to be expected of them in their work. Ministry isn�t just about preaching, teaching, or ministering anymore; neither is healthcare just about healing. Our systems have become much more complicated.
FARNSLEY: Does it strike you as odd that the upper quarter of congregations have the lion�s share of the members and the money? This is not a value judgement about the quality of care for their members, the quality of worship, but just a descriptive fact.
CEBULA: It doesn�t surprise me because in our culture we consolidate our school systems, we consolidate our newspapers and media, grocery chains�everything is that way. It also doesn�t surprise me that there are so many fascinating little churches.
ETINDI: A lot of the work that congregations do is not measured financially: mentoring people, for instance, who are moving from welfare to work. A lot of the work consists of becoming involved in people�s lives and being a friend, being someone who will check up on them and help them along. That doesn�t really require any money; just time and a compassionate spirit.
FARNSLEY: If congregations do good, good for them. That doesn�t need to be measured. But if we are talking about their relationship to social programs and public spending, it does have to be measured. I don�t know how many people have been successfully mentored by congregations in Indianapolis. But if it becomes a decision issue in foundations and public policy then we can�t say, "Well, some of what they do is not financial," or, "Some of what they do is spiritual." If these things are being figured into budgets and policies and programs, then they are like any other organization. We have to know how much, how many, and who are they are.
CEBULA: So then the question of whether "average" means "typical" become paramount.
FARNSLEY: At least it keeps you from making the mistake of saying, "There are nine congregations with 10 members each, plus one congregation with 910 members, and the average of them is 100." Because 100 doesn�t look anything like any of them.
CEBULA: It is important to realize that there are churches to whom it does not matter. They don�t want a contract. They don�t want grant money. The public policy people need to be okay with the fact that not every church is going to be involved.
FARNSLEY: You did a story about the governor offering technical services to congregations, in which you said there are 1,200 congregations in Indianapolis. While this is true, the number of congregations to whom these services will mean anything is more like 150. What matters is that you didn�t say�as the editorial did�"There are congregations like this on every street corner." There are 1,200 that are technically eligible, but we are not really talking about that.
ETINDI: For years, Catholic Charities, Jewish Family Services, Hoosier Family Services, and others have been accustomed to receiving money from the government for their work in such things as immigration, refugee services, and resettlement. For organizations such as these, to think in terms of getting other funds from governments is no big deal. They know how to go through the process of applying, and whom to contact. But for others, this is a brand new thing. Not having someone to help them through the process would make it almost impossible.
FARNSLEY: That is a great point. Catholic Charities and groups like that were set up as non-profit corporations, essentially separate from the denomination, because it was understood that they were going to play by federal government rules. They were going to have administrators; they were going to write grants; they weren�t going to be theological or sectarian or make people read the Bible before they got services. With Charitable Choice, it is not clear how much sectarian activity and religious teaching is permissible with government money. So, in fact, many smaller organizations�that would never set up a non-profit�see themselves in line for government contracts.
CEBULA: But they don�t have the guidelines to help them understand how this is different. Catholic Charities, Lutheran Family Services, the Jewish Federation�they formed because parishes and synagogues and churches came together and said, let�s pool our resources and provide for the needy. It was never the case that those institutions were going to do work on behalf of local congregations, or help only people of those particular faiths.
ARMSTRONG: Well, this is the time to say: any other questions or thoughts that we haven�t addressed?
CEBULA: Diana suggested helping congregations to locate themselves on a continuum. If we can�t do hard data, we can help congregations see themselves in other ways, by telling stories about real people, real action, real care giving.
ARMSTRONG: Thank you all.
