SLIDE 1 Religion and Polygyny in a Christian sub-Saharan setting: Combining the Top-Down and Bottom-Up Perspectives Victor Agadjanian, PhD Department of Sociology University of Kansas, USA Carlos Arnaldo, PhD Centre for African Studies Eduardo Mondlane University, Mozambique Abstract Polygyny remains widespread in sub-Saharan Africa. In this study, we contribute to a better understanding of the role of organized religion in the resilience of the institution of polygyny. Using a unique combination of data from a census of religious congregation leaders and a household-based survey of women conducted in parallel in a predominantly Christian rural district in southern Mozambique, we examine variations in polygyny attitudes and practices across different types of Christian denominations. The multivariate analyses produced a nuanced picture of how religious leaders and rank-and-file members of different denominations negotiate and reconcile the doctrinal norms with the realities of both the marriage and religious
- markets. Specifically, the analyses detect substantial net denominational variations in
prevalence and acceptance of polygyny, with the largest contrast being between two types of African Initiated Churches (AIC) – Zionist (highest prevalence and acceptance) and Apostolic – further illustrating the complex nature of the AIC phenomenon.
SLIDE 2 1
Background and conceptualization Polygyny remains widespread across sub-Saharan Africa (SSA), including settings where both legal and cultural norms are supposed to discourage it (Corbinian and Kyara 2013). In this study, we contribute to a better understanding of the role of religion in the resilience of the institution of
- polygyny. Using a unique combination of data from a census of religious congregation leaders
and a household-based survey of women conducted at about the same time in a predominantly Christian rural district in southern Mozambique, we examine variations in polygyny attitudes and practices across different types of Christian denominations. Considerable research has addressed the association of organized religion with polygyny in sub-Saharan Africa. Much of that research, however, has focused on Christian-Muslim
- differences. Thus Islam, which doctrinally permits polygyny, is usually believed to be more
conducive to it than Christianity (e.g., Hayase & Liaw, 1997; Klomegah 1997). Indeed, in predominantly Muslim areas of the sub-continent, polygyny is often justified by reference to the Islamic canon (Agadjanian & Ezeh 2000). In Kudo’s (2014) study in Malawi, Muslim women (along with women with no formal religion) were more likely to be in polygynous unions than were Christian women. In their study of five sub-Saharan countries, Ghana, Kenya, Senegal, Uganda and Zambia, Timæus & Reynar (1998) found that Muslims and other non-Christian women were more likely to be married polygynously than Christians, although in Kenya the association of religion and polygynous marriage was not significant after adjusting for other
- characteristics. Recent multi-national analyses of DHS data documented higher rates of polygyny
among Muslims, compared to non-Muslims (Dalton & Leung, 2014; Westoff and Bietsch 2015). Yet, the cross-national evidence is inconclusive (e.g., Barber 2008), and, as Madhavan (2002)
SLIDE 3
2
argued, in Islamic societies attitudes toward polygyny vary across cultural and socioeconomic contexts. Although it is often assumed that Christianity is opposed to polygyny, research has produced evidence of considerable variations in polygyny levels across Christian denominations (e.g., Klomegah 1997). Historically, most Christian missions in Africa adamantly condemned polygyny as incompatible with “true” Christian doctrine and values and even as a source of abuse against women, but the reality has often conflicted with this seemingly unambiguous stance, especially as Christian churches became increasingly indigenized in many parts of the sub-continent (Nmah 2012). Thus in Cameroon, local interpretations of Christianity (often articulated by women), presented polygyny as fully congruent with Christian faith (Notermans 2002). Falen (2008) argues against a generalization that women prefer the monogamous Christian model in Benin, as marriage choices are shaped by considerations of economic rationality and social prestige. In fact, justification for polygyny is often sought in Biblical texts (Phiri 2006). In South Africa, some churches, such as Shembe church (the Nazareth Baptist Church), an African initiated church (AIC) founded by Isaiah Shembe in 1910 that blends Christianity with elements of Zulu traditional religion, took a lenient, if not favorable, attitude toward polygyny (e.g., Hillman 1975). Not surprisingly, in a study conducted in a South African demographic surveillance site, polygyny levels were higher among followers of that church compared to the rest of the population (Hosegood et al. 2009). Baloyi (2013) found greater acceptance of polygyny among AIC and Pentecostal churches, compared to mission-initiated (or “mainline”) ones. In his analyses of correlates of polygyny based on the National Population Census and Demographic and Health Survey data from Mozambique, Arnaldo (2004; 2011) found significantly lower levels of polygyny among Roman Catholics and no significant
SLIDE 4 3
difference across other Christian denominations or between non-Catholic Christians and Muslims. Following this evidence, we expect to find variations in acceptability and practice of polygyny across denominational types, with denominations that are more lenient on traditional religious and cultural practices in general having higher polygyny rates and greater acceptance of polygynous marriages. Going beyond testing this general hypothesis, we look at various dimensions of polygyny-related attitudes, and we also compare how leaders and rank-and-file members of different churches articulate their views on acceptability of polygyny and reflect on what may account for possible dissonance in the official vs. popular affirmation and interpretation of church position on polygyny. Setting The data for this study come from the district of Chibuto in Mozambique’s southern Gaza
- province. The district’s population numbered around 200,000 at the last (2007) population
census for which data are available at this writing.1 The mainstay of this predominantly rural district’s economy is subsistence farming. Low agriculture yields, made even less predictable by frequents floods and droughts, and the proximity of the area to South Africa, has made male labor migration to the neighboring country an important element of the district’s economic and social fabric. The district is largely monoethnic, dominated by the Changana ethnicity and
- language. The traditional lineage system of southern Mozambique is patrilineal and its culture is
highly patriarchal (Junod 1912; Loforte 2000). As in many patrilineal sub-Saharan settings, marriage in southern Mozambique is traditionally bridewealth-based. However, even in rural
1 The most recent population census was carried out in Mozambique in August 2017.
SLIDE 5 4
areas, the institution of marriage has undergone considerable erosion, with bridewealth payments
- ften delayed or foregone altogether.
The district is predominantly Christian, with considerable denominational diversity. Roman Catholicism was de facto the official church of the Portuguese empire, of which Mozambique was part until its independence in 1975. Yet, the colonial times also saw the arrival and spread of mission-based Protestant denominations, such as Anglican, Methodist, Presbyterian, and other. Starting in the late colonial period and well into independence, the district saw a massive proliferation of Pentecostal-type African Initiated Churches (AIC). Among them, Zionist churches, first brought to the area from South Africa, grew particularly fast. Zionists’ emphasis
- n miracle healing through the power of the Holy Spirit has been a particularly strong attraction
causing massive defections from the Catholic Church and mainline Protestant denominations. Although individual Zionist churches are typically small, together Zionists make up the largest denominational block in southern Mozambique. Another category of AIC that has gained ground in the area is Apostolic churches. Unlike ideologically and organizationally amorphous Zionists, Apostolics are characterized by very rigid and insular ideology and organizational structure. Finally, historically most recent arrivals on southern Mozambique’s religious scene are neo- Pentecostal churches. In contrast to earlier Pentecostals, these new religious actors, often
- riginating outside the African continent, are focused on pursuit of holistic wellness, beyond
cure for or protection from illness (Agadjanian & Yabiku 2015). In the study area, as in other patrilineal societies of SSA, polygyny is traditionally normative and widespread (Arnaldo, 2007; Junod 1912; Sithoe 2009). Polygyny is demographically sustained by high birth rates and gender age difference in age at marriage (Goldman & Pebley 1989) and also by relatively high male adult mortality and shortage of men due to out-migration.
SLIDE 6 5
Persistence of polygyny is also related to very low prevalence of official marriage registration and of religious marriages. Despite a long history of Christian and, more generally, western influence in the region, a minuscule share of marriages are formalized through either a religious ceremony or civil registration. Thus in the representative household survey used in this analysis,
- nly about 3% of married respondents were in a civil or religious union.
Data and Method We use data from two sources for statistical tests: a census of all religious congregations and a representative household survey, both carried out mainly in 2008 in the district of Chibuto. While the two blocks of data cannot be directly linked, they offer a unique opportunity to compare the bottom-up and top-down perspectives on religion and polygyny. The census used a standardized questionnaire administrated in person to a congregation leader, a man or a woman. In total, leaders of 1125 congregations were interviewed. In the household survey, a standard questionnaire was administered to a randomly selected household female member aged 18-50. In all, 2019 women living both in the district’s headquarters and its rural areas were interviewed. The survey data therefore reflect women’s perspective on religion and polygyny. While non- inclusion of men in the survey is a limitation, we should note that in the study area, as elsewhere in SSA, women typically are much more involved in church activities than men, and arguably, church involvement has more important consequences for women’s wellbeing than it does for men’s (Agadjanian 2015; Agadjanian and Yabiku 2015). We use the denominational classification that reflects the earlier described religious composition of the area and has been tested in numerous prior analyses (e.g., Agadjanian 2013; Agadjanian 2015; Agadjanian & Yabiku 2015). This classification includes five denominational groups – Roman Catholic, Mainline Protestant, Zionist, Apostolic, and Neo-Pentecostal. We
SLIDE 7 6
exclude from the analysis of the census three Islamic communities. Similarly, eleven Muslims are excluded from the analysis of the survey data. We also exclude Jehovah’s Witnesses (one case in the census and three cases in the survey) as this group cannot be easily fitted into the proposed denominational classification. Table 1 displays the distribution of the denominational groups in the census and the household survey samples as used in these analyses. Table 1 here We first use the household survey data to assess the prevalence of polygyny among members of different denominations as well as women not affiliated with organized religion. Next, we look at denominational variations in church acceptability of polygyny among affiliated survey respondents: this analysis is based on respondents’ answers to the question on whether their church allows that a married man takes another wife. Then, we use the congregation census data to examine leaders’ attitudes toward polygyny. This analysis allows for a direct comparison with household survey data, but it also expands beyond that comparison as the census included two additional polygyny-related questions: 1) whether or not a polygynous man can be admitted into the church; and 2) whether or not a polygynous man can hold a leadership position in the
- congregation. We start with descriptive comparisons and then fit a series of multivariate logistic
- regressions. To account for the clustered nature of the household survey sample, we fit multilevel
models that allow the intercept to vary randomly by clusters. These models are fitted using the GLIMMIX routine in SAS. Results Descriptive results
SLIDE 8 7
Table 2 shows the prevalence of polygyny among currently married household survey
- respondents. Among affiliated women, Zionist women had the highest levels of polygyny
(30.2%), followed by Neo-Pentecostals, mainline Protestants, Catholics, and Apostolics. Not surprisingly, non-affiliated women have a higher percentage of those in a polygynous union than do affiliated ones, on average: 32.2% vs. 27.6%. However, it is notable that non-affiliated women are not very different from Zionists. Table 2 here The second section of Table 2 shows the distribution of all church-affiliated survey respondents who answered affirmatively to the question: “Does your church allow that a member of the church who already has one wife takes another wife?” As can be seen, there is a big difference between the share of such women and the share of polygynously married women in most denominational categories: perceived acceptability of entering a polygynous marriage by a church member is much lower than the reported level of polygynous marriages. Zionists are the
- nly exception: in fact, in that segment, acceptability of taking a second wife is slightly higher
than the actual polygyny level. It should be noted here that some of the discrepancies may owe to the formulation of the polygyny acceptability question: that question did not contemplate situations when men who are already polygynously married would ask to join the church. Even so, the difference between the perceived norm and reality in most denominations is quite telling. Table 3 shows descriptive results for three questions asked in the congregation census. There is a noticeable variation in accepting that a married church member may take another wife. Zionists stand out with the highest share of leaders who would accept this possibility. However, even among them, this share is less than 10 percent, in a stark contrast with the corresponding
SLIDE 9 8
share among household survey Zionist respondents. The second section of Table 3 displays the shares of the positive responses to the question on whether the church would admit a polygynous man as a new member. Here, the differences across denominations are rather minor, with denominational percentages hovering around the sample average, 80%. Finally, the last section
- f Table 3 shows the shares of interviewed leaders who said that a polygynous man can occupy a
leadership post in the congregation. Zionists clearly stand out, especially in comparison with Apostolics and Roman Catholics. Again, however, even among Zionists, only one-fifth of respondents gave an affirmative answer. Table 3 here Multivariate analysis Table 4 compares married women in different denominations as well as unaffiliated women on polygyny status of their marriages. Zionists are the reference category. In Model A, no controls are included. In this model, Catholics are Apostolics are significantly less likely to be in polygynous unions than Zionists. At the same time, the likelihood of being married polygynously is higher among non-affiliated women than among the reference group. Model B adds controls. The difference between Zionists and unaffiliated women fully disappears. The difference between Zionists and Catholics is now only marginally significant (p<.07). However, the contrast between the reference group and Apostolics barely changes in the magnitude and remains highly statistically significant: controlling for individual, marital, household, and community characteristics, the odds of being in a polygynous union are 52% lower among Apostolics than among Zionists (OR=exp(-.65)=0.52).
SLIDE 10
9
Table 4 here Table 5 presents results of the analysis of respondents’ opinion on whether a married male church member may take another wife. Only women with a religious affiliation are included in this analysis. Model 5.A includes only denomination as a predictor. The contrast between Zionists and women from all other denominations is strong and statistically significant: women in those denominations are significantly less likely to think that their churches accept that a married man takes another wife. Model 5.B adds controls: in addition to the controls used in the analysis of the likelihood of being in a polygynous marriage, it controls for marital status—being in a polygynous, a monogamous union, or not being married. In addition, it controls for respondents’ frequency of church attendance—a proxy for church involvement and familiarity with church rules. Despite the addition of this long battery of controls, the pattern of differences does not change: members of all denominations are significantly less likely to think that their church would accept a second marriage by a married man then are Zionists. Apostolics, however, stand out. When we run the same model with Apostolics as the reference category, Apostolics are significantly less likely to think that polygynous marriage would be acceptable in their churches not only compared to Zionists, but also compared women in all other denominational categories (results are not shown but are available upon request). Table 5 here Next, we examine responses to the same question – if the church would allow a married male member to take another wife – among congregation leaders interviewed in the congregation
SLIDE 11 10
- census. The results of these tests are presented in Section A of Table 6. The model controls for
the gender of the congregation leader, the congregation location (urban vs. rural), congregation size (number of attendees at last full service), and congregation financial health (approximated by the type of congregation’s walls – brick/wooden walls, thatch walls, and no walls). The results are in stark contrast with those of the survey data analyses: no differences at the conventional significance threshold (p<.05) are detected, and only Apostolic leaders are marginally less likely to accept that a married man marries another wife, compared to Zionist leaders (p<.08). No other covariates have significant effects. Table 6 here Table 6 also shows the results of a model predicting an affirmative answer to the question if a polygynous man may become a member of the church (Section 6.B). The models controls for the same congregation characteristics. Here, only Catholic leaders are significantly less likely to accept this possibility than are Zionist leaders. Among the controls, leaders of urban congregations are less inclined to admit a polygynous man to the church, compared to leaders of rural ones. Acceptance of polygynous men as members is also positively associated with the size
Finally, we compare responses of the congregation leaders to the question on whether a polygynous man can assume a leadership post in the congregation. The results, presented in Section 6.C, show a strong difference between Zionists and the rest, the former being significantly more likely to accept that a polygynous man may become a congregation leader.
SLIDE 12 11
Also interestingly, acceptance of a polygynous man as a congregation leader is negatively associated with congregation leader being a woman. Discussion Polygyny remains a pervasive marital arrangement in southern Mozambique, as in many other parts of sub-Saharan Africa, and the prolonged presence and popularity of Christianity, which at least in its pure dogma is incompatible with polygyny, does not seem to have greatly affected its prevalence. In order to understand the persistence of polygyny in largely Christian sub-Saharan rural and semi-rural settings, such as the one considered in our study, one needs to keep in mind the demographic constraints that shape marital dynamics in such settings, namely the gender imbalance owing to a combination of high fertility, age difference at marriage, men’s excessive mortality and absence from the community due to migration. Most churches, even if condemning polygyny as a non-Christian arrangement, have to adjust to this demographic reality. The socioeconomic dimension of polygyny is another important factor in its resilience and de facto
- acceptability. Polygynous men are typically among the wealthiest and most influential
community members. Polygynous marriage is an investment that such men make to enhance the productive and reproductive capacity of their households and to assert and further elevate their
- status. For many religious congregations, the membership of such men nourishes an expectation
that they provide financial support to the church; it also adds to the social standing and visibility
- f the church. Finally, the near-saturated local religious market, with its intense competition for
members is another factor that shapes many church leaders’ positions and actions with regard to polygyny.
SLIDE 13 12
Nonetheless, the variations across denominations in stated attitudes toward polygyny that our analyses detected are quite instructive. On one extreme, there are Zionist churches, where polygyny seems more acceptable than in other denominations. These African-initiated churches, whose doctrine and ritual, while ostensibly rejecting pre-Christian traditional beliefs and practices, in reality absorbed and appropriated them, have the highest level of polygyny acceptance, even though among Zionist leaders less than a third deemed it acceptable for a married church member to take another wife. On the other extreme are Apostolics, also an African-initiated group of churches, but with a much more hostile stance on non/pre-Christian forms and expressions. Apostolics indeed have the lowest prevalence of polygyny net of other
- factors. Yet, in official attitudes, the difference between Apostolics and Zionists is noticeable
- nly in accepting the possibility that a polygynous man assumes a church leadership post. And as
far as admitting a polygynous man to church, it is Roman Catholics, with their long-standing, even if generally flaccid, opposition to polygyny, who stand out. Yet, even among Catholic congregation leaders, fully three-quarters would allow a polygynously married man to become a member of the church. Our study points to another important aspect of the religion-polygyny association – differences in interpretations of and attitudes toward polygyny between church leadership and congregation rank-and-file members, the latter being generally more tolerant of polygyny (with the exception of Zionists). That religious dogma, prescriptions, and proscriptions are differently interpreted and acted upon by church members and leaders (and oftentimes, by leaders at different levels) is part of real-life dynamics of many religious organizations. This apparent disjunction is, in fact, a reflection of negotiability and conditionality of religious tenets, norms, and practices, which, we argue, help ensure the organizational viability of the church.
SLIDE 14 13
Finally, it is also important to note that denominational and leader-vs.-members attitudes toward polygyny in that sub-Saharan setting are not framed in terms of relative gender equitability of monogamous vs. polygynous marriage models. The rejection and condemnation
- f polygyny is justified entirely on the basis of Christian norms or the church organizational and
financial needs, and not on the ground that it might be oppressive for women. Yet, the significant net gender difference in acceptance of polygynous men as congregation leaders in the census data analysis suggests the potential importance of the gendered angle in church views and actions regarding polygyny, echoing earlier research on how women’s leadership may transform religious congregations (Agadjanian 2015). Although we cannot capture these gendered variations with the survey data gathered from women only, they are important destinations of our future inquiries.
SLIDE 15 14
References Agadjanian, V. 2013. “Religious denomination, religious involvement and modern contraceptive use in southern Mozambique” Studies in Family Planning 44(3): 259-274. Agadjanian, V. 2015. “Women’s religious authority in sub-Saharan Africa: Dialectics of empowerment and dependency” Gender & Society 29 (6): 982–1008. Agadjanian, V., & A.C. Ezeh. 2000. “Polygyny, gender relations, and reproduction in Ghana.” Journal of Comparative Family Studies, 31 (4): 427-441. Agadjanian, V., & S. T. Yabiku. 2015. “Religious belonging, religious agency, and women’s autonomy in Mozambique.” Journal for the Scientific Study of Religion 53(3): 461-476. Arnaldo, C. 2004. “Regional fertility trends in Mozambique” Journal of Population Research 21(2): 177-197. Arnaldo, C. 2007. Fecundidade e seus Determinantes Próximos em Moçambique: uma análise dos níveis, tendências, diferenciais e variação regional. Maputo: Texto Editores. Arnaldo, C. 2011. “Tendências e factores associados à poligamia em Moçambique.” Pp. 193-
- 211. in N. Teles, B. Muianga, & E. Brás (eds.) Mosaico Sociológico. Maputo: Universidade
Eduardo Mondlane. Baloyi, E. M. 2013. “Critical reflections on polygamy in the African Christian context.” Missionaria 41(2): 164-181. Barber, N. 2008. “Explaining cross-national differences in polygyny intensity: Resource-defense, sex ratio, and infectious diseases” Cross-Cultural Research 42 (2): 103-117 Corbinian V. J., & S.J. Kyara. 2013. “New Evangelization and the incidence of polygamy.” Hekima Review 48: 41-48.
SLIDE 16 15
Dalton, J. T. & T. C. Leung. 2014. “Why is polygyny more prevalent in Western Africa? An African slave trade perspective.” Economic Development and Cultural Change 62(4): 599- 632. Falen, D. J. 2008. “Polygyny and Christian marriage in Africa: The case of Benin” African Studies Review 51(2): 51-74. Goldman, N. & A. Pebley. 1989. “The demography of polygyny in sub-Saharan Africa.” Pp. 212-237 in R. J. Lesthaeghe (ed.) Reproduction and Social Organization in Sub-Saharan
- Africa. Berkeley: University of California Press.
Hayase, Y. & K.-L. Liaw. 1997. “Factors on Polygamy in Sub-Saharan Africa: Findings Based
- n the Demographic and Health Surveys.” The Developing Economies 35(3): 293–327.
Hillman, E. 1975. Polygamy reconsidered. African Plural Marriage and the Christian African
- Churches. Maryknoll: Orbis Books.
Hosegood, V., N. McGrath, & T. Moultrie. 2006. “Dispensing with marriage: Marital and partnership trends in rural KwaZulu-Natal, South Africa 2000-2006” Demographic Research 20: 279–312. Junod, H. A. (1912). The life of a South African tribe (Vol. 1). Attinger frères. Klomegah, R. 1997. “Socio-economic characteristics of Ghanaian women in polygynous marriages.” Journal of Comparative Family Studies 28(1): 73-88. Kudo, Y. 2014. “Religion and Polygamy: Evidence from the Livingstonia Mission in Malawi.” Institute of Developing Economies Discussion Paper No. 477. Loforte, A.M. (2000). Género e Poder entre os Tsonga de Moçambique. Maputo: Promédia. Madhavan, M. 2002. “Best of friends and worst of enemies: Competition and collaboration in polygyny.” Ethnology 41(1): 69-84.
SLIDE 17 16
Nmah, P. E. 2012. “Christianity, polygyny and homosexuality in Nigeria: A Theopeotics culture
- f acceptance and rejection.” AFRREV An International Journal of Arts and Humanities
1(2): 41-56. Notermans, Catrien. 2002. “True Christianity without dialogue. Women and the polygyny debate in Cameroon” Anthropos Bd. 97, H. 2: 341-353. Phiri, I. 2006. “The Bible and polygamy.” In T. Adeyemo (ed.) Africa Bible Commentary. London: Piquant. Sithoe, Y. 2009. “Poligamia: tudo em nome da ‘tradiçao’” Outras Vozes 26: 12-13. Timæus, I. M. & A. Reynar. 1998. “Polygynists and their wives in sub-Saharan Africa: an analysis of five Demographic and Health Surveys.” Population Studies 52(2): 145-162. Westoff, C. F. & K. Bietsch. 2015. “Religion and reproductive behavior in Sub-Saharan Africa.” DHS Analytical Studies No. 48. Rockville, Maryland, USA: ICF International.
SLIDE 18
17
Table 1. Denominational composition of the Chibuto Congregation Census and Household Survey, Chibuto Denominational category Census Survey Roman Catholic 8.3 12.6 Mainline Protestant 15.7 9.6 Apostolic 9.6 12.0 Zionist 48.2 43.1 Neo-Pentecostal 18.2 11.0 Unaffiliated n/a 11.6 Number of cases 1121 2005 Table 2. Prevalence of polygyny among currently married respondents and perceived acceptability of polygyny among all respondents by denominational affiliation, Chibuto household survey (percent) Denominational category Prevalence of polygyny Church accepts that a married member takes another wife Roman Catholic 19.5 13.0 Mainline Protestant 25.5 13.5 Apostolic 18.0 7.1 Zionist 30.2 31.7 Neo-Pentecostal 26.2 14.0 All denominations 27.6 21.5 Unaffiliated 32.2 n/a Table 3. Descriptive statistics, Chibuto Congregation census (percent) Denominational category
Church accepts that a married member takes another wife A polygynous man can be admitted to church Polygynous man can occupy a leadership post
Roman Catholic 4.3 75.3 3.2 Mainline Protestant 5.7 78.0 5.1 Apostolic 5.8 78.5 1.9 Zionist 9.2 81.7 20.0 Neo-Pentecostal 6.9 79.4 9.8 All denominations 7.6 79.9 12.9
SLIDE 19 18
Table 4. Denominational variations in the likelihood of being in a polygynous union, currently married women, Chibuto household survey, 2008 (multi-level logistic regression, parameter estimates and standard errors) Covariates A B Estimate SE Estimate SE [Zionist] Roman Catholic
0.2193 + Mainline Protestant
0.2280 Apostolic
0.2214 ** Neo-Pentecostal
0.2105 Unaffiliated 0.3419 0.1761 * 0.1933 0.1872 Age 0.2066 0.0590 ** Age squared
0.0009 ** [No education] 1 to 4 years of education
0.1437 5 of more years
0.1872 [Rural residence] Urban residence
0.1858 * [Did not have a previous marriage] Had a previous marriage 1.0276 0.1543 ** Household material possession scale 0.1484 0.0693 * [No bridewealth paid] At least some bridewealth paid 0.1293 0.1444 Constant
0.9744 ** Number of cases 1549 1549 Notes: reference categories in brackets; significance levels: ** p<.01, * p<.05; + p<.1.
SLIDE 20 19
Table 5. Denominational variations in perception of church’s acceptance of polygyny, all affiliated women, Chibuto household survey (multi-level logistic regression, parameter estimates and standard errors) Covariates A B Estimate SE Estimate SE Denominational category [Zionist] Roman Catholic
- 1.0581 0.2119 **
- 1.0672 0.2212 **
Mainline Protestant
- 1.0617 0.2357 **
- 1.0565 0.2433 **
Apostolic
- 1.8191 0.2694 **
- 1.7916 0.2728 **
Neo-Pentecostal
- 1.0662 0.2159 **
- 1.0841 0.2203 **
Age
Age squared 0.0013 0.0008 + [No education] 1 to 4 years of education
5 of more years
[Rural residence] Urban residence 0.1152 0.2231 Household material possession scale 0.1350 0.06977 * [In polygynous union] In monogamous union
Not married
[Did not have a previous marriage] Had a previous marriage 0.2055 0.1608 [Did not go to church in past 2 weeks] Went to church 1-2 times in past 2 weeks
Went to church 3 or more times in past 2 weeks
Constant
1.7297 0.8919 + Number of cases 1773 1770 Notes: Reference categories in brackets; significance levels: ** p<.01, * p<.05; + p<.1.
SLIDE 21 20
Table 6. Acceptability of polygyny and of leadership post for polygynous congregation members, Chibuto congregation census (logistic regression, parameter estimates and standard errors) Covariates A Married member may take another wife
may be admitted to church
may assume a leadership post Estimate SE Estimate SE Estimate SE Denomination category [Zionist] Roman Catholic
- 0.8055 0.5417
- 0.6355 0.2769 *
- 1.9471
0.6016
**
Mainline Protestant
- 0.4985 0.3613
- 0.3353 0.2180
- 1.4993
0.3610
**
Apostolic
0.4921 +
0.7263
**
Neo-Pentecostal
0.3165
0.2615
**
Congregation leader is a woman [men]
0.2557 0.1700
0.2134
*
Urban location [rural]
- 0.3235 0.3193
- 0.4956 0.1899 ** -0.3605
0.2567 Number of attendees at last regular cult 0.0855 0.0793 0.2674 0.0588 ** -0.0509 0.0686 Type of congregation building [no building] Reed walls 0.0504 0.2516
0.2311 0.2021 Wooden or brick walls
0.3864 0.0832 0.2434
0.3197 Constant
1.1808 0.1762 ** -1.1473 0.2000
**
Number of cases 1119 1119 1119 Notes: reference categories in brackets; significance levels: ** p<.01, * p<.05; + p<.1.
