The Other Half of Health:
Patterns in Declining Churches

Cynthia Woolever
Hartford Institute for Religion Research
Hartford Seminary
woolever@hartsem.edu

Society for the Scientific Study of Religion Annual Meeting, Rochester, New York
November 4, 2005
Session G-10: Mobilizing the Religious Capital of Clergy and Laity

Abstract

Congregational research often considers factors related to their numerical growth and vitality. For example, previous research by Bruce and Woolever (2004) demonstrated that measures based on the input of worshipers in congregations were significant predictors of health in congregations. However, less is known about the least vital congregations—those falling in the 20th percentile or lower on congregational vitality measures. Based on a national random sample of worshipers in 409 congregations in the U.S., this analysis demonstrated that patterns of lay participation and experience differ significantly between Catholic parishes, mainline Protestant churches, and conservative Protestant churches. Catholic parishes score the lowest on congregational vitality, conservative Protestant churches the highest, while mainline Protestant churches’ scores fall in between. What is associated with low congregational vitality in each faith group? The model presented here included three sets of factors: (1) congregational demography—the average age and income of worshipers; race and gender of worshipers; (2) congregational characteristics—the size and growth in worship attendance; and (3) institutional measures—conflict in the previous five years and the congregation’s financial situation. Logistic regression showed that these three sets of factors are not helpful in distinguishing between healthy and unhealthy Catholic and conservative Protestant congregations. However, several of these factors are significantly associated with unhealthy mainline Protestant congregations. Mainline Protestant congregations with older worshipers, fewer women (i.e., lower than the national average of 60%), stable or declining worship attendance, no conflict, and a poor financial situation were more likely to score in the bottom one-fifth of all mainline congregations. These findings for mainline Protestant churches support previous work that identifies (a) the positive relationship between vitality and some congregational conflict; and (b) the detrimental effect of young worshipers’ absence. Further, the results contradict the perception that greater participation by women negatively affects congregational vitality. Finally, the set of factors related to congregational vitality or growth (e.g., size) are not necessarily related to unhealthy congregational life.

Other Congregational Measures

Congregational Demography

The four measures were calculated by aggregating across worshiper responses within each congregation. Average income of worshipers was also aggregated to the congregational level. However, since worshipers reported his or her total household income on a scale from 1 (“less than $10,000”) to 6 (“$100,000 or more”), their responses were first recoded to the mid-point of each income range.

Congregational Characteristics

Size: One person in each congregation completed a congregational profile survey and reported the average worship attendance for the current year and each of the last five years. Because the survey was given in April 2001, worship attendance figures for that year were incomplete, and a number of congregations did not report the figure. Congregational size was based on average worship attendance in 2001, if available, or 2000 if not.

Congregational growth in worship attendance: Numerical growth was measured as the percentage of growth or decline in worship attendance between 1996 and 2000 (attendance in 2000 minus attendance in 1996, the difference divided by attendance in 1996) or as a percentage of growth or decline between 1996 and 2001. Some congregations did not report attendance figures for 1996 and/or 2000. In such cases, growth was calculated based on the earliest and latest reported attendance figures. A log transformation of the congregational growth variable was used to correct for skewness and kurtosis.

Institutional Measures

Conflict in previous 5 years: This measure was a discreet variable coded 1 if the key informant indicated there had been conflict in the previous five years serious enough such that a congregational split occurred, members left, or the pastor left.

Responses were coded 0 if no such conflict was reported.
Financial situation: This measure was a categorical variable based on key informant responses where 0=financial situation is a serious threat to continued existence, 1=declining financial base, 2=stable financial base, and 3=increasing financial base.

View the tables for this report