Characterizing Membership Growth and Decline in the Reformed Church in America 2000-2005

Mainline denominations have experienced continuous losses in membership since the 1960s and 1970s, a trend that gives no sign of abating today.1 According to a November 2006 report released by the General Assembly Council of the Presbyterian Church (USA), all mainline Protestant denominations studied saw membership declines in the ten year period 1994–2004. The membership declines ranged from 4% (Reformed Church in America (RCA) and Evangelical Lutheran Church in America (ELCA)) to 21% (Disciples of Christ).2

Trying to understand why the steady downward trend in church membership has been occurring has been the focus of many researchers. The main justification for the research is that if membership declines were better understood, perhaps denominations would be able to allocate resources more effectively in the effort to fulfill the mission and vision of the church. A majority of these researchers first sought to understand the dynamics of growth and decline in the late 1970s and early 1980s when the declines in membership in mainline denominations first began. Two main schools of thought on the reasons for growth and decline in specific congregations soon emerged. Note: That we reserve the term “church” to describe the RCA as a denomination, and use the term “congregation” to describe specific, organized groups of believers.

The first school of thought is that factors external to congregations are responsible for the loss in membership.3 External factors include the demographics of the area surrounding a congregation: whether the population of the area around a congregation is growing or shrinking or whether the ethnic/racial profile in the area surrounding a congregation is changing or remaining the same. They also include the religious context in which a congregation operates: how many other congregations are nearby, what denominations are represented among them, and how many people in the area are members of other congregations. The second school of thought suggests that membership growth and decline in mainline congregations is a product of internal factors.4 These include the conservatism or liberalism of a congregation, its worship style, and its emphasis on particular types of ministries (education, evangelism, etc.).

A more realistic view, however, maintains that it is neither internal nor external factors taken separately, but rather a complex interplay of both that determines growth or decline in church membership.5 This in fact is the viewpoint of Roger Nemeth and Donald Luidens. In an article devoted to an examination of the demographic changes in the RCA, these sociologists discuss the declining membership of the denomination from the 1960s through the 1990s with reference both to external (birth rate, regional population changes, changing gender roles) and internal (congregational size and giving) factors.6 More recently, these two authors have collaborated with Corwin Schmidt and Jim Penning to broaden their research to include an analysis of the demographics of the Christian Reformed Church (CRC), the sister church of the RCA.7

The RCA Today

In 2003 the RCA focused its efforts to combat membership loss through the adoption of “Our Call,” an initiative that invites the church to devote its energies and resources over the next decade to the creation of new congregations and the revitalization of existing ones. The following study constitutes the response of a statistician and two of his students to this invitation. We propose here a statistically sound and objective analysis of factors associated with membership growth and decline of RCA congregations during the period 2000–2005. We begin by examining the state of existing congregations in 2000 with reference to several of the external and internal factors suggested by researchers as contributing to church growth and decline. On the basis of our measurement of these factors, we then attempt to predict whether a congregation would grow or decline by 2005. Membership growth and decline is assessed for both confessing members as well as for members who were baptized, but not confessing (mainly youth). Our results reveal that the RCA saw a 6.8% decline in total membership over the five year period 2000-2005. These results at the same time show that membership loss varied dramatically by region, with steeper declines in the Synods of the East and the Far West than in the Midwest.

Previous studies have shown that some factors related to church growth and decline are only significant in particular regions.8 Thus, we examine potential growth-related factors in a way that allowed us to assess whether factors are related to growth/decline in the RCA overall, or only in particular regions. Our findings demonstrate that while some factors (e.g. smaller congregations tending to lose members faster than larger ones) are implicated in membership loss across the RCA, others are related to membership change only in certain areas of the United States (e.g. median community income related to membership change only in the Far West).

In this analysis we look at thirty-one internal and external factors, and demonstrate statistically that eight factors are the most significantly related to congregational membership change over the five-year period. While we purposefully refrain from making substantive conjectures about the causes or implications of the findings, we do hope that the information we provide below proves useful to RCA leaders who will be interested in the types of congregations that are growing and the types that are shrinking.

Methods

Databases. Each year the RCA publishes statistical tables based on self-reported demographic information for each congregation in the RCA for the prior calendar year. These tables are available in print form in the annual RCA “orange books” (The Acts and Proceedings of the General Synod of the RCA) with tables from more recent years also available in electronic form via the RCA’s website (www.rca.org). The orange books are a major source of the data used in this study.9

Sample. In 2000, there were 939 RCA congregations, of which 896 were physically located in the contiguous United States. Forty-one RCA congregations were in Canada and 2 in the Virgin Islands. By 2005, 34 of these 896 contiguous United States RCA congregations had been disbanded, released, or merged, leaving 862 congregations.10 Additionally, two congregations, Crystal Cathedral (Garden Grove, CA) and Fairhaven (Jenison, MI), were assigned the status of “mega-churches” (congregations with worship attendance over 2000) and eliminated from the analysis, leaving 860 congregations.11 Four additional congregations, part of the Collegiate Corporation in New York City, were also eliminated from the analysis because they only report statistical information in the aggregate. Thus the sample used for the majority of this analysis consists of 856 congregations, which operated in the RCA from 2000 to 2005, which were not involved in any mergers or closings and re-openings, and which did not have the status of a mega-church or corporation during that time.

Response Variables. Members are classified into two groups: confessing and baptized. Confessing members are those who have professed their faith and are typically high school aged or older. Baptized members are those who have not professed their own faith but have been baptized into the church. Typically, baptized members are high school aged or younger. We assign to each congregation a baptized membership and confessing membership change score which is calculated as the number of members in 2005 divided by the number of members in 2000. Thus, a congregation with 105 confessing members in 2005 and 100 confessing members in 2000 would have a confessing membership change score of 1.05, corresponding to 5% growth. Similarly, if that congregation instead had only 90 confessing members by 2005, their confessing membership change score would be 0.90 corresponding to a 10% decline in membership.

Predictor Variables. The factors by which we attempt statistically to predict a membership change score in a congregation (including baptized and confessing) are classified into three groups: demographic context, religious context, and institutional context.

Demographic context factors. Seventeen demographic context factors are included in the analysis. All factors except one were obtained from the US Census 2000 and represent the value corresponding to the zip code of the congregation of interest. The demographic context factors are: (1) the proportion of employed workers who commute at least 30 minutes to work each day; (2) the proportion of population aged 25 and older with at least a bachelor’s degree; (3) the proportion of the labor force that is employed; (4) the percent of the adult population that is below the poverty line; (5) the median home value; (6) the average family income; (7) the population density (people per square mile); (8) the percent of the population that is white; (9) the percent of the population that is aged under 9 years; (10) the percent of the population that is aged over 65 years; (11) the average family size; (12) the percent of the population that is married; (13) the percent of the population that is widowed; (14) the percent of the population that has never been married; (15) the percent of the population whose primary ancestry is Dutch; (16) the percent change in population from 1990 to 2000; and (17) the percent of the county that voted for Bush in 2000. The Bush variable, unlike the other sixteen variables, was measured at the county level using vote counts from CNN.12

Religious context factors. Six factors relating to religious context are included in the analysis. All these factors were measured at the county level and obtained from the Religious Congregations and Membership Survey year 2000 (RCMS, 2000). The six factors are: (1) the percent of the county that is a member of a congregation; (2) congregation density (the number of congregations per 1000 people in the county); (3) the percent of the county that is a member of a congregation affiliated with a mainline denomination; (4) the percent of the county that is a member of a Catholic parish; (5) the percent of the county that is a member of a RCA congregation; and (6) religious plurality (the number of identified denominations within a county).

Institutional context factors. Eight factors are included that measure specific characteristics of RCA congregations in 2000. These factors were obtained for each congregation from the RCA orange books. The eight factors are: (1) The number of baptized members in 2000; (2) the number of confessing members in 2000; (3) the percent of active members who attend worship in a typical week; (4) the relative size of the congregation’s Sunday school; (5) the relative number of people who attended new member classes in 2000; (6) giving per member (a congregation’s total income divided by the number of active members); (7) the percent of a congregation’s total income given to missions; and (8) the number of years that the congregation has existed.

Region. We focus on nationwide trends within the RCA, but also seek to identify regional differences in church membership growth. Preliminary analysis shows that differences in membership growth 2000–2005 can best be described by classifying the seven regional synods of the RCA into three groups:

This regional breakdown is depicted in Figure 1.

Figure 1: Graphical representation of the regions of the RCA

Graphical representation of the regions of the RCA

Stability Scores. After determining which factors are related most significantly to membership growth and decline (see “Results” section), we create stability scores for each congregation to summarize its predicted change in membership in the period 2000– 2005. Stability scores are defined separately for baptized membership change (baptized membership stability score) and confessing membership change (confessing membership stability score), and also reflect some regional differences.

Baptized membership stability score. For a baptized membership stability score, each congregation gets points according to the following algorithm. Congregations get one point for each of the following factors in 2000: having more than 200 baptized members; having more than 500 confessing members; having less than 30 denominations in the county; and having at least 75% of members attend services on a typical Sunday morning. Congregations receive an additional point if more than 100% of members attend services on a typical Sunday morning (this represents non-members attending). Additionally, congregations in the Far West receive one point each for being in a zip code with a median income less than $40K and in an area that experienced more than 30% growth in the period 1990–2000. The sum of these points is the congregations baptized membership stability score. The higher the number the more likely a congregation was able to retain or increase the number of baptized members in the period 2000–2005. (See Figure 5a.)

Confessing membership stability score. We also define a confessing membership stability score. Congregations get one point for each of the following factors in 2000: having more than 200 baptized members; having at least 40% of members participate in Sunday school (with an additional point if at least 60% participate); and having at least 75% of members attending services on a typical Sunday morning. Congregations receive an additional point if more than 100% of members attend services on a typical Sunday morning (this represents non-members attending). Additionally, congregations in the Midwest and Far West receive one point if they are in an area that has between 500 and 2000 people per square mile and two additional points if they are in an area that has less than 500 people per square mile. The sum of these points is the confessing membership stability score. The higher the number the more likely a congregation was able to retain or increase its number of confessing members in the period 2000-2005. (More detail is provided below; see especially Figure 5b.)

Analysis. The analysis presented below depends on a relatively sophisticated statistical analysis. By means of it, we are able to pinpoint which of the 31 factors (17 relating to demographic context, 6 to religious context, and 8 to institutional context) are most directly related to church growth and decline in the RCA overall or in specific regions.

To clarify this point, consider that many of the 31 factors, if examined in isolation, will be found to be related to church membership growth or decline. However, realizing that factors do not act in isolation, we examined the data statistically to find the factors that were the most strongly associated with church membership change by way of a statistical technique called multiple regression.

To better understand our analytic method, consider the following example. Consider two of the 31 factors: size of new member class in 2000 and percent attendance in 2000 and their relationship with the response variable, baptized membership change. There are three possible relationships between the two factors and one response variable as visualized in figure 2 shown below. Double headed arrows represent relationships between factors. For convenience, we have labeled the three arrows “A,” “B,” and “C.”

Figure 2: Visualizing relationships between factors

Visualizing relationships between factors

If we only looked at size of the new member class factor and its relationship with baptized membership change we would see that there is a relationship (congregations with larger new member classes in the year 2000 were more likely to grow over the five- year period under study). What we don’t know is whether or not this relationship is observed because there is a direct relationship between size of new membership class in year 2000 to baptized membership change or whether it is indirectly related. A direct relationship between size of new member class in year 2000 and baptized membership change is conceptualized by arrow C. If this arrow were actually present, this would mean that size of new member class in year 2000 has a direct impact on baptized membership change in the year 2000.

An indirect relationship between size of new member class and baptized membership change would be described by arrows A and B, together. Specifically, if congregations with larger new member classes are more likely to have higher percent attendance (arrow A), and, in turn, congregations with higher percent attendance are more likely to have relationship with baptized membership change (arrow B), then new member class may have an indirect relationship with membership change.

Multiple regression, the analytic method used here, allows us to pinpoint which of the 31 factors under study are most likely to have direct relationships with baptized and confessing membership change (the response variables). After analyzing the example data presented above using multiple regression we find that arrow C really does not exist; however arrows A and B exist, indicating that new member classes have an indirect relationship with membership change.

Here is a bit more detail on the statistical analysis. Our preliminary analysis uses a nested analysis of variance (ANOVA) to examine the proper level of hierarchy relative to regional differences. For our main analysis, all predictors were examined for association with the response variables in a quantile regression predicting the median. Subsequently, significant variables are entered into a multiple (multivariable) quantile regression model and eliminated systematically (backwards elimination) until all variables reach the a=0.01 level of significance thus indicating a strong direct effect on the response variables.13

Results: Active Membership Growth and Decline in the RCA 2000-2005

In 2000, the RCA had 295,666 total members, of which 38,976 were identified as inactive, leaving 256,690 active members. Of these 256,690 active members 75,527 were baptized members, but not confirmed (confessing) members of the church. Because of the more or less uniform practice of infant baptism in the RCA, profession of faith does not occur until the teen years. Consequently, the vast majority of these 75,527 baptized members are pre-teens.

Table 1: Breakdown of RCA membership for the years 2000 and 2005

Type of Member Year 2000 Year 2005 Net Change 2000 to 2005 Percent Change 2000 to 2005
Active Baptized 75,527 67,886 -7,641 -10.1%
Confessing 181,163 168,450 -12,713 -7.0%
Inactive 38,976 39,121 +145 +0.3%
Total 295,666 275,457 -20,209 -6.8%

Table 1 shows how membership changed across the RCA in the period 2000–2005. Overall, the RCA experienced a 6.8% decline in membership over this five-year period, a decrease of 20,209 members. This loss was due to a 10.1% decline in baptized members and a 7.0% decline in confessing members.

The numbers in Table 1 reflect the entire RCA. However, we focus on the 856 congregations physically located in the United States, had originated before 2000, and were still active in 2005. Among these 856 congregations there were 234,194 active members in 2000, of which 166,026 were confessing and 68,168 baptized, but not confessing. By 2005, the number of active members had decreased to 214,114 (an 8.6% loss), of which 154,869 were confessing (a 6.7% loss) and 59,245 were baptized, but not confessing (a 13.1% loss).

While there was an overall decline in membership among these 856 congregations, Tables 2a and 2b show that this was not due to precipitous declines in membership in a select number of congregations; rather, most congregations experienced a moderate or significant loss of membership over the five-year period. Specifically, 60.3% (516/856) of RCA congregations nationwide experienced moderate or significant loss of confessing members and 58.3% (499/856) of RCA congregations nationwide experienced moderate or significant loss of baptized members. Especially notable is the fact that 42.3% (362) of RCA congregations experienced a significant loss (>25%) of baptized members over the five-year period. Tables 2a and 2b also show that while loss of members is more an issue in the Eastern and Far West regions, the Midwest is also losing members.

Table 2a: Growth and loss of confessing membership in RCA congregations 2000–2005

East Midwest Far West Overall
Significant Loss
(At least 25% decline)
31.9%
(122)
19.9%
(80)
32.4%
(23)
26.3%
(225)
Moderate Loss
(5-25% decline)
34.8%
(133)
34.5%
(139)
26.8%
(19)
34.0%
(291)
Stable
(Between 5% decline and 5% gain)
11.5%
(44)
19.6%
(79)
11.3%
(8)
15.3%
(131)
Moderate Growth
(5-25% growth)
14.1%
(54)
17.9%
(72)
18.3%
(13)
16.2%
(139)
Significant Growth
(>25% growth)
7.8%
(29)
8.2%
(33)
11.3%
(8)
8.2%
(70)
Total 100%
(382)
100%
(403)
100%
(71)
100%
(856)

Table 2b: Growth and loss of baptized membership in RCA congregations 2000–2005

East Midwest Far West Overall
Significant Loss
(At least 25% decline)
47.1%
(180)
37.7%
42.3%
(30)
42.3%
(362)
Moderate Loss
(5-25% decline)
10.2%
(39)
20.8%
(84)
19.7%
(14)
16.0%
(137)
Stable
(Between 5% decline and 5% gain)
14.1%
(54)
13.9%
(56)
12.7%
(9)
13.9%
(119)
Moderate Growth
(5-25% growth)
8.9%
(34)
12.7%
(51)
9.9%
(7)
10.8%
(92)
Significant Growth>
(>25% growth)
19.6%
(75)
14.9%
(60)
15.5%
(11)
17.1%
(146)
Total 100%
(382)
100%
(403)
100%
(71)
100%
(856)

Figure 3 further explores confessing and baptized membership losses by region. The East and Far West regions had 23.5% and 22.3% reductions respectively in the number of baptized members over the five-year period, while the Midwest had an 8.0% reduction. Reductions in the number of confessing members followed a similar pattern. The East experienced a 13.6% reduction, the Far West a 10.1% reduction, and the Midwest a 3.4% reduction. Thus, while nationwide active membership saw an 8.6% decline, there were overall decreases of 16.3%, 13.7%, and 4.8% in the East, Far West, and Midwest, respectively.

Figure 3: Percent loss in baptized and confessing members in the RCA 2000–2005

Percent loss in baptized and confessing members in the RCA 2000–2005

In the last 40 years the East has seen dramatic declines in active membership, dropping from nearly 150,000 members to just over 50,000. These declines account for nearly all the overall declines in RCA membership over the thirty-five year period as both the Midwest and Far West regions saw marginal increases in active (baptized and confessing) membership. This is depicted in Figure 4 below.

Figure 4: Active (baptized and confessing) membership in the RCA 1965–2000

Percent loss in baptized and confessing members in the RCA 2000–2005

While this thirty-five year trend has continued in the East over the last 5 years (16.3% decrease), both the Midwest and Far West regions are now experiencing declines in active membership as well (4.8% and 13.7%, respectively).

Variables related to baptized and confessing membership change. Statistical analysis shows eight factors of congregations in 2000 most significantly related to confessing and baptized membership change. These eight factors are displayed in Table 3.

Table 3: The eight factors significantly related to confessing and baptized membership change

Factor Related to change in confessing members Related to change in baptized members
Demographics
Population Density Yes (Only in Midwest and Far West)
Regional Income Yes (Only in Far West)
Regional Population Change Yes (Only in Far West)
Religious Context
Regional Religious Plurality Yes
Institutional Context
Number of Confessing Members (2000) Yes
Number of Baptized Members (2000) Yes Yes
Sunday School Size in 2000 Yes
Attendance Rate Yes Yes

Three of the 8 variables related to membership change are demographic, one is a religious context variable, and the remaining 4 are institutional variables. The following analysis will help us better understand the relationships between these eight variables and membership change.

Population Density. Population density is related to confessing membership change in the Midwest and Far West regions of the RCA. Table 4 shows that congregations in urban areas (5000+ people per square mile) were hit particularly hard in terms of loss of confessing members with a median loss of 33% in the Far West region and 22% in the Midwest region.

Table 4: Relationship of population density and confessing membership decline

Population density of area surrounding the congregation Median change in baptized members Number of congregations
Midwest
Rural: Less than 500 people per square mile -4% 238
Suburban: 500-2000 people per square mile -9% 82
Dense Suburban: 2000-5000 -18% 63
Urban 5000+ -22% 20
Far West
Rural: Less than 500 people per square mile 7% 20
Suburban: 500–2000 people per square mile -3% 10
Dense Suburban: 2000–5000 -25% 20
Urban 5000+ -33% 21

Median Family Income. In the Far West, median family income of the community surrounding the congregation was related to baptized membership changes over the five-year period.

Table 5: Relationship of median family income and baptized membership decline in the Far West region

Median family income of area surrounding congregation Median change in baptized members Number of congregations
Less than $40K 0% 17
$40K–$60K -24% 19
$60K+ -11% 25

Congregations in middle income ($40K–$60K) areas saw the largest decline in baptized members, with a median decline of 24%.

Population Growth 1990-2000. Also in the Far West, congregations in areas that experienced limited growth in the 1990s saw substantial losses in percentages of baptized members 2000–2005, as shown in Table 6.

Table 6: Relationship of population change 1990-2000 and baptized membership decline in the Far West region

Population change 1990-2000 of area surrounding congregation Median change in baptized members Number of congregations
Less than 10% growth -50% 15
10-20% growth -35% 23
20-30% growth -12% 11
30%+ growth -5% 22

Religious Plurality. Religious plurality (number of distinct denominations represented in a county) was significantly related to a decline in baptized members nationwide, with congregations in counties with few denominations experiencing smaller declines than those in counties with many.

Table 7: Relationship of religious plurality and baptized membership decline

Religious Plurality (Number of different denominations represented in county) Median change in baptized members Number of congregations
Less than 30 different denominations -8% 241
Between 30 and 50 -12% 338
Fifty or more -25% 277

Congregation size (number of confessing members). Congregation size, measured by number of confessing members in 2000, was related to baptized membership change in the period 2000–2005. Specifically, larger congregations experienced the smallest losses in membership, as shown in Table 8a.

Table 8a: Relationship of congregation size (Number of confessing members) in 2000 and baptized and confessing membership decline

Congregation size in 2000 (# of confessing members) Median change in baptized members Number of congregations
Very Small: Less than 50 -13% 144
Small: 50-100 -28% 220
Medium: 100-200 -12% 212
Large: 200-500 -13% 212
Very Large: 500+ -7% 68

Congregation size (number of baptized members). Congregation size as measured by the number of baptized but not confessing members was related both to confessing membership and baptized membership change as shown in Table 8b. Specifically, larger congregations saw smaller declines in both the number of baptized and confessing members. The exception to this was congregations with no youth members in 2000, which on average saw no change in baptized members five years later. (Of course there were none to lose!)

Table 8b: Relationship of number of youth in 2000 and baptized and confessing membership decline

Number of youth (baptized members) in 2000 Median change in baptized members Median change in confessing members Number of congregations
No baptized members 0% -19% 56
Very Small: 1-25 -25% -16% 231
Small: 25-50 -27% -13% 162
Medium: 50-100 -18% -9% 178
Large: 100-200 -15% -8% 143
Very Large: 200+ -7% -2% 86

Sunday school size. The relative size of a congregation’s Sunday school program in 2000 was significantly related to change in confessing members, as shown in Table 9. Specifically, congregations with larger Sunday school programs (for their size) saw smaller declines in membership.

Table 9: Relationship of Sunday school size and confessing membership decline

Sunday school size (as a percentage of active membership) Median change in confessing members Number of congregations
Less than 20% -19% 124
20-40% -10% 449
40-60% -9% 201
60+% 0% 80

Attendance Rate in 2000. The percentage of active members who attended a typical Sunday service in 2000 was a significant predictor of both baptized and confessing membership change in the period 2000–2005. Congregations with low attendance rates were the most likely to lose members over the five-year period, as shown in Table 10.

Table 10: Relationship of typical Sunday worship attendance and confessing membership decline

Percent of members that attend on a typical Sunday morning Median change in baptized members Median decline in confessing members Number of congregations
Less than 50% -30% -15% 161
50-75% -17% -12% 405
75-100% -5% -4% 163
More than 100% 0% -3% 127

Stability Factors by Region. Table 11 illustrates the findings from Tables 4–10 by identifying those factors of congregations which gave them “stability” in membership in the period 2000–2005.

Table 11: Stability Factors

Factor Stability in the number of confessing members Stability in the number of baptized members
Demographics
Population Density Rural and Suburban (Less than 2000 people per square mile) (Midwest and Far West only)
Regional Income Median family income less than $40K (Far West only)
Regional Population Area with 30%+ growth
Change 1990–2000 (Far West only)
Religious Context
Regional Religious Plurality Less than 30 different denominations in the county
Institutional Context
Number of Confessing Members in 2000 At least 500 confessing members
Number of Baptized Members in 2000 At least 200 baptized members At least 200 baptized members
Sunday School Size in 2000 At least 40% of active members participate in Sunday school
Attendance Rate At least 75% of members typically attend Sunday services At least 75% of members typically attend Sunday services

Figures 5a and 5b below show how stability scores based on factors of a congregation in 2000 predicted membership change five years later. Stability scores were calculated as a sum of particular factors related to membership stability as described above. Specifically, stability scores provide a quick and definitive method of classifying a congregation’s ability to grow in the period 2000–2005.

Figure 5a: Baptized member stability score by region

Baptized member stability score by region

Figure 5b: Confessing member stability score by region

Confessing member stability score by region

Conclusion

As with all mainline denominations, the RCA has seen significant declines in membership since the 1960s. However, until recently these overall declines could be explained almost entirely by a loss of members in the East region of the RCA (as shown in Figure 3). Over the last five years membership loss has continued in the East, while there have also been membership losses in the Far West and, to a lesser extent, the Midwest.

In this study we sought to identify membership stability factors common to all RCA congregations in the U.S. in the period 2000–2005. Specifically, congregation size (large number of both confessing and baptized members), relative size of Sunday school, attendance rate and religious plurality (number of distinct denominations in the county) were related to baptized and/or confessing membership change over the five-year period. Additionally, three demographic factors were identified (population density, median family income, and population change) and related to church membership change in specific regions of the RCA (specifically the Far West and the the Midwest). Figures 4a and 4b show how the stability factors identified can quite definitively distinguish shrinking from stable congregations.

Consistent with previous research, we have shown that factors associated with church membership growth and decline are both external and internal. In our study we found four of each. However, while all four of the internal factors were implicated in membership change for the RCA overall, only one of the external factors was implicated in all regions of the RCA. The other three external factors were related to membership change only in particular regions of the United States.

We note that two of the four institutional variables related to membership change (number of baptized members and number of confessing members) were related to congregation size. Specifically, larger congregations had more stable memberships over the five-year period compared to smaller congregations. Additionally, we note that the other two institutional variables related to membership change (relative size of Sunday school and attendance rate) reflected a congregation’s membership participation level in Sunday school and Sunday worship, as well as the ability to include non-members in Sunday school and Sunday worship.

The main limitation to the preceding analysis is that there exist several factors potentially related to membership change which were not considered in this analysis. Possible factors include many institutional level variables, such as worship style, pastor personality and style, participation in church revitalization networks, as well as other demographic context variables, such as birth rate. Not only might these variables have been related to membership change, but they might also have explained more adequately why particular variables in the preceding analysis seem to be related to membership change. For example, if birth rates are low in population dense areas of the Midwest and Far West, it may well be that the fact that congregations in population dense areas are less likely to grow can be better attributed to birth rate than to population density. It is important to note, however, that the preceding analysis identified the eight out of the thirty-one factors most directly related to church membership change in the RCA over the five-year period studied (For more on why this is the case, please see the Analysis section.)

In sum, in the period 1965 to 2000, the synod of the East accounted for virtually all loss in membership in the RCA. However, during the last five years both the Far West and, to a lesser extent, the Midwest, have lost members while losses in the Eastern synods continue. Especially concerning is the fact that 42.3% (362/856) RCA congregations had more than a 25% loss in baptized (youth) members in only a five-year period. Additionally, eight factors of stable congregational membership have been identified: four were related to size and participation levels in a particular congregation, and four to the demographic and religious context in which a congregation exists. Five of the eight factors are found to be common nationwide, while the remaining three are specific to particular regions of the United States.

Appendix A: Congregations Disbanded and Released Between 2000 and 2005

Classis Status Congregation City State
Albany Disbanded Christ Church Watervliet New York
California Disbanded Corona Community Church Corona California
Disbanded Rancho Capistrano Community Church San Juan Capistrano California
Cascades Disbanded Servants of Christ Reformed Church Federal Way Washington
Central California Disbanded Christ Church of the Bay Area San Mateo California
East Sioux Disbanded First of Clay County Everly Iowa
Nassau-Suffolk Disbanded Willston Park Willston Park New York
New Brunswick Released Magyar New Brunswick New Jersey
Pleasant Prairie Disbanded New Song Community Hiawatha Iowa
Red River Disbanded Hope Reformed Church Carrollton Texas
Disbanded Christ Community Church Dallas Texas
Disbanded Valley Ridge Community Colleyville Texas
Passaic Valley Disbanded Lakeview Heights Clifton New Jersey
Disbanded Christ Community Highland Lakes New Jersey
Florida Disbanded Hudson Reformed Church Hudson Florida
Disbanded Christ Community Church Palm Springs Florida
Disbanded Palm Grove Reformed Church Holiday Florida
Central Iowa Disbanded Calvary Des Moines Iowa
Schenectady Disbanded Buskirk Reformed Church Buskirk New York
Rocky Mountain Disbanded Church of the Rockies Denver Colorado
New York Disbanded Union Reformed Church of Highbridge Bronx New York
Muskegon Disbanded Central Community Church Muskegon Michigan
Minnesota Disbanded Minnesota Valley Community church Prior Lake Minnesota
Greater Palisades Disbanded Central Avenue Jersey City New Jersey
Delaware Raritan Disbanded Fourth Reformed Church Somerville New Jersey
Dakota Released Presbyterian-Reformed Church United Willow Lake South Dakota
Chicago Disbanded Peace Reformed Church Mount Prospect Illinois
Disbanded New Life Community Alsip Illinois

Appendix B: Congregations Merged between 2000 and 2005

Classis Closed Church City Combined Church City State
California Oasis Community Church Victorville Crosswinds Community Church Hesperia CA
Southwest Michigan First Reformed Church Kalamazoo Second Reformed Church Kalamazoo MI
Greater Palisades The Reformed Church on the Hill Cresskill Clinton Avenue Reformed Church Bergenfield NJ
Chicago Green Oak Reformed Church Oak Lawn Christ Community Palos Heights IL

Appendix C: Congregations Adopted and Organized between 2000 and 2005

Classis Status Church City State
Brooklyn Organized Ghanian Presbyterian Church of Brooklyn Brooklyn New York
California Adopted Abundant Mission Church Los Angeles California
Adopted Dasom Korean Church Long Beach California
Organized The Bridge Chino California
Organized East Hills Community Riverside California
Organized Oak Springs Community Church Murrieta California
Organized Iglesia Reformada la Senda Corona California
Central California Organized Lighthouse Community Church Lathrop California
Organized New Life Community Church Tulare California
North Grand Rapids Organized Clancy Street Grand Rapids Michigan
Queens Organized Sung Shin Kew Gardens New York
West Sioux Organized Rejoice! Community LeMars Iowa
Wisconsin Organized Iglesia Trinidad Milwaukee Wisconsin
Central Iowa Organized Plano Plano Texas
Organized Celebrate Community Church Knoxsville Iowa
Organized Prairie Ridge Church Ankey Iowa
Zeeland Organized Spring Valley Community Church Allendale Michigan
South Grand Rapids Organized RiverTree Community Church Wyoming Michigan
Rocky Mountain Organized Forest Ridge Community Church Monument Colorado
Organized Harvest Christian Community Wheat Ridge Colorado
Holland Organized Haven Shores Community Church West Olive Michigan
Florida Organized Comunidad de Fe Palm Springs Florida
Greater Palisades Organized Han Sung Cresskill New Jersey
Organized Glory Christian Hackensack New Jersey
Organized HunMaUm Reformed Church Palisades New Jersey
Dakota Organized Open Door Fellowship Corsica South Dakota

1 Lynn Japinga, “The Rain of God: RCA Growth and Decline in Historical Perspective,” Lecture Series, No. #2, Visiting Research Fellows Program, Van Raalte Institute, Hope College. April 20, 2004.

2 Perry Chang, “Recent changes in membership and attendance in mainline protestant denominations,” Research Services, A Ministry of the General Assembly Council, Presbyterian Church (USA), Louisville, KY. November 2006, www.pcusa.org/research/resports/denominational_size.pdf, (27 June, 2007).

3 See e.g., Dean Hoge and David Roozen, eds., Understanding Church Growth and Decline: 1950-1978, (New York: Pilgrim Press, 1979); David Roozen and C. Kirk Hadaway, eds., Church and Denominational Growth (Nashville: Abdingdon Press, 1993); Roger Stump, “The Effects of Geographical Variability on Protestant Church Membership Trends, 1980-1990,” Journal for the Scientific Study of Religion, Vol. 37, No. 4 (Dec 1998), 636-651; Richard Rathge and Gary Goreham, “The Influence of Economic and Demographic Factors on Rural Church Viability,” Journal for the Scientific Study of Religion, Vol. 28, No. 1 (March 1989), 59-74; and C. Kirk Hadaway, “The Demographic Environment and Church Membership Change,” Journal for the Scientific Study of Religion, Vol. 20, No. 1 (March 1981), 77-89.

4 See e.g., Dean Kelly, Why Conservative Churches Are Growing: A Study in Sociology of Religion (New York: Harper and Row, 1972) and Laurence R. Iannaccone, Daniel V. A. Olson, and Rodney Stark, “Religious Resources and Church Growth” Social Forces, Vol. 74, No. 2 (December 1995), 705-731.

5 See e.g., William McKinney and Dean R. Hoge, “Community and Congregational Factors in the Growth and Decline of Protestant Churches,” Journal for the Scientific Study of Religion, Vol. 22, No. 1 (March 1983), 51-66.

6 See e.g., Roger Nemeth and Donald Luidens, “The RCA in the Larger Picture: Facing Structural Realities,” Reformed Review, Vol 47, no. 2 (Winter 1993-94), 85-112.

7 Corwin Smidt, Jim Penning, Donald Luidens and Roger Nemeth, Divided by a Common Heritage: The CRC and the RCA at the Beginning of the New Millennium, Historical Series of the Reformed Church in America, (Grand Rapids: Eerdmans, 2006).

8 See, for example, C. Kirk Hadaway, “The Demographic Environment and Church Membership Change,” Journal for the Scientific Study of Religion, Vol. 20, No. 1 (March 1981), 77-89.

9 Prior to 2000 statistical tables were not available electronically, and so until that time Hope College sociologists Don Luidens and Roger Nemeth, with the help of many Hope College students kept a manually entered database of these statistical tables. This database was generously provided by Professors Luidens and Nemeth for use in this study. In addition to the RCA’s own statistical information, data was gathered from the year 2000 Religious Congregations and Membership Survey (RCMS, 2000) and the 2000 U.S. Census (Census 2000).

10 See Appendices A and B for a list of disbanded, released, and merged congregations. Appendix C provides a list of congregations added to the RCA during the same time period; there were 26.

11 For a definition of a “megachurch” see hirr.hartsem.edu/megachurch/definition.html (June 2007)

12 Available at www.cnn.com/ELECTION/2000/results/index.president.html (April 2007)

13 More details of the statistical analysis are available from the contact author upon request. He can be reached at: Nathan Tintle, Ph.D, Assistant Professor of Mathematics, Hope College, 27 Graves Place, Holland, MI 49423, tintle@hope.edu