Reconceptualizing Houses of Worship to Advance Comparisons across Religious Traditions

Abstract

The study of religion and politics has struggled to find concepts, methods, and approaches that advance productive comparisons of phenomena across different religious practices and traditions. Consequently, scholars who seek to understand the impact of religious practice on political outcomes across religious traditions encounter challenges in aggregating findings and advancing scholarly inquiry. A reconsideration of the role of houses of worship as an intermediary variable connecting religious practice to political outcomes yields a potentially fruitful avenue for comparative investigation. While social processes that take place within worship spaces are frequently presumed in the mechanisms linking religious variables to political outcomes, these worship spaces are generally undertheorized and overlooked within the study of religion and politics. A body of scholarship has substantiated the significance of congregational variation within the study of Christianity, yet the most commonly cited quantitative literature on religion and political participation omits discussion of this level of variation. Drawing on the shared conceptual space across worship domains from several religious traditions, this article examines houses of worship as an organizational concept that can be employed productively for theoretical and empirical analyses of religion and politics.

1. Introduction

Within the study of religion and politics, most causal and analytical frameworks that seek to understand the impact of religious variables on political outcomes articulate a social process by which doctrinal beliefs and devotional practices become consequential to individual believers’ political attitudes or behaviors, whether through interaction with clergy, family members, co-religionists, or media. While scholars have investigated multiple mechanisms connecting the divine and civic realms, most existing frameworks make a general assumption that these social processes take place at least partly in spaces that are specifically designated for worship.

The social scientific study of religion is a multidisciplinary field, and the discussion of worship spaces varies both across and within disciplines. At the most basic level, there is variation in whether the worship spaces are discussed explicitly, or rather their existence is implied within a broader discussion of religious institutions or networks. When worship spaces are incorporated into analytical frameworks engaging religion and political outcomes, different characteristics might be emphasized by scholars depending on the goal of analysis. For example, some scholars are interested in the spatial relationship between sites of worship and other spaces of political, economic, or social value. This particular focus is common in scholarship examining religiously based conflict. In a valuable contribution to the comparative study of religion and politics, Hassner (2009) conceptualizes “sacred spaces” as phenomena that exist across religious traditions and share the attributes of acting as places where worshippers can communicate with the divine, contain a permanent divine presence, and provide meaning to the faithful. As he describes, many sacred sites, over time, become institutionalized into formalized places of worship that are central to religious devotion. Hassner’s conceptualization of sacred spaces is principally concerned with understanding the process through which sacred sites become the location of intractable sectarian conflict.

For scholars investigating the relationship between religion and the state, the broader institutional structure of embeddedness for houses of worship within the same religious tradition is frequently emphasized. Comparative analysis of houses of worship is rarely the central focus of research in any discipline. As one scholarly review notes, meso-level units, such as houses of worship, are understudied in comparison to macro and micro approaches (Herzog et al. 2020). Most research on the role houses of worship play in mobilizing political action has focused on democracies in predominantly Christian countries (Campbell 2004; Greenberg 2000; Johnston and Figa 1988; McDaniel 2008). More specifically, scholars have extensively studied the relationship between participation in regular religious services and engagement in civic and political life in the United States (Johnson and Tamney 1986; Leege and Kellstedt 1993; Morris 1984; Putnam and Campbell 2010; Wald 1992). Within this scholarship, several findings highlight churches’ roles in fostering political engagement by: (1) providing a forum where individuals can be exposed to political messages; (2) strengthening individual capacity-building by teaching and practicing civic skills; and (3) expanding individuals’ social networks, increasing the likelihood of political mobilization.

Outside of Christianity, however, houses of worship are underexamined in comparison to broader political and social movements or charitable organizations. Within the study of Islam and politics, for example, mosques are usually only discussed within the larger framework of Islamic organizations or the networks of specific political or religious leaders. Despite their general neglect in studies of Islam and politics, an emerging literature aims to consider mosques’ role as mediators between religious practice and political participation. Much of the empirical work is focused on mosques in Muslim-minority countries (Smits and Ultee 2013; Fleischmann et al. 2016; Choi et al. 2011; Jamal 2005; Sobolewska et al. 2015; Dana et al. 2011; Westfall 2019). Within this scholarship, several findings point to the possibility that the functional role played by churches in serving as conduits for civic and political engagement may be present in mosques also. Chouhoud et al. (2019) and Westfall (2019), for example, find that mosque involvement is a positive predictor of Muslim political participation in the United States. In some of the few examinations of mosques in Muslim-majority contexts, Lussier and Ahnaf (2024) find that political messages within individual mosques in Indonesia are often inconsistent from week to week, while Butt’s (2016) study of Friday prayers in Pakistan notes that the sermon can motivate worshippers to participate in political opportunities.

As Brenneman and Miller (2016) note, “religious buildings are objects that exert unique pressure on those who utilize the space” (p. 84). Yet our tools for productive comparison of these buildings—which are central to the broader social and political processes that appear to connect religion and politics—are limited. The lack of agreed concepts and measures for examining worship spaces across religious traditions limits our ability to compare similarities observed in different contexts, as well as more clearly identify the mechanisms responsible for variation that appears across faith traditions. Ultimately, without a better set of shared tools for comparison, our ability to aggregate findings from different contexts is severely hampered.

The goal of this review is to add to our scholarly toolkit by articulating a thicker conceptualization of houses of worship that can be productively applied in empirical analyses. It will start with a concise overview of the way that worship spaces are characterized within the dominant literature discussing religion and politics, followed by an analysis of the shared conceptual space that exists across denominational boundaries. I will introduce a thicker conceptualization of house of worship and will illustrate how this concept can be applied empirically. While the concept of house of worship developed here is meant to be inclusive of all faith traditions that engage in communal worship, for simplicity’s sake, the forthcoming analysis will focus primarily on examples from Islamic and Christian religious traditions.

2. Explicit, Implicit, and Missing Worship Spaces

How does existing scholarship on religion and politics characterize the spaces in which worship takes place? Within political science, worship spaces are generally examined through the study of congregations, which emerged primarily in examination of American Christianity. In particular, the Little Rock Congregations Study (https://ualr.edu/lrcs/ accessed 26 June 2024) comprises one of the few longitudinal, comprehensive data collection efforts focusing on congregations. Findings from the American experience have inspired scholars to examine religious practice in predominantly Christian communities, particularly in Latin America and Africa (Smith 2019; McClendon and Riedl 2019; Brown and Brown 2003), fostering a broader comparative literature on religion and politics. A particular strength of this comparative and cross-national body of scholarship is its close attention to a range of interactions that take place among worshippers and between clergy and worshippers in religious settings. Scholarship on congregations has helped to problematize our understanding of how religion becomes an everyday experience. It has identified and clarified the mechanisms by which religious values, beliefs, and social expectations are being communicated to believers and how aspects of congregational organization shape worshippers’ experiences outside of the devotional space.

Scholarship in which congregations are either the unit of analysis or are theorized and empirically measured for incorporation into broader studies, however, comprises a smaller proportion of the overall volume of work investigating religious practice as a factor influencing political outcomes. Despite the valuable contributions that research on congregations have made to the study of religion and public life, worship spaces are rarely integrated as a theoretical or empirical variable in broader quantitative analyses connecting religion and politics. One obstacle, undoubtedly, is lack of data. None of the publicly available public opinion surveys include measures of respondents’ houses of worship, thereby limiting the extent to which we can consider them as a possible source of variation impacting individuals’ political outcomes. Consequently, conceptualization of worship spaces within much of the quantitative scholarship on religion and politics is thin. Discussion of characteristics in which houses of worship may vary from each other is absent and the units are largely viewed as interchangeable or homogenous within a single religious tradition.

To better understand how worship spaces are represented within the quantitative literature that has had the greatest influence on the study of religion and political engagement, I looked closely at six heavily cited articles that appear when using the search terms “religion and political participation.”1 All of the articles draw on empirical samples taken from the United States. Three focus on Christians and three focus on Muslims. None of the articles examining Christians discussed houses of worship or worship spaces directly or tried to measure them empirically. Two of the three articles on Muslims deliberately discuss the role of the mosque, but do not measure it. In statistical terms, houses of worship are omitted variables in these empirical models.

Table 1. Summary of Empirical Measures in Quantitative Studies of Religion and Political Participation.

Article	Jones-Correa and Leal (2001)	Jamal (2005)	Djupe and Grant (2001)	Driskell et al. (2008)	Ayers and Hofstetter (2008)	Dana et al. (2011)
Google Scholar citations (as of 5 June 2024)	494	331	261	230	113	110
Data Source	Latino National Political Survey (1989); ANES (1990)	Survey of 335 Muslim Americans in NY area	Citizen Participation Survey	Baylor Religion Survey (2005)	Survey of 1846 American Muslims	Muslim-American Public Opinion Survey
Measures used to capture religious practice (IV)	Four measures: (1) religious affiliation; (2) church attendance; (3) centrality of religion in lives; (4) personal religious experience	Ordinal scale of mosque involvement beyond prayer	Sample limited to respondents who attend church at least 1x/month Measures Y/N practice a civic skill in church	Religious participation index comprising 9 different activities; level of engagement in them; religious denomination; and beliefs	Four dimensions of religious context: (1) raised or convert to Islam; (2) Muslim spouse; (3) importance of Islam in life; (4) frequency of mosque attendance, prayer and volunteerism	Two measures: (1) importance of religion in daily life; (2) level of involvement in local mosque
Measures used to capture political participation (DV)	Y/N to voting, being registered to vote and 7-item participation index	Index of 4 non-voting political activities	Y/N to any non-voting political activity	Index of 11 political activities	Index of 8 political activities	Index of four non-voting political activities
Worship space approximated **	Measure of religious denomination (Catholic, Protestant)	Captured within religious practice variable and measure of “foreign born” Muslims	Measure of different religious traditions (Mainline Protestant, Evangelical Protestant, African-American Protestant, Catholic)	Measure of religious denomination (Catholic, black Protestant, evangelical Protestant, mainline Protestant, Jewish and other)	Captured within religious context variables	Measures of ethnic background and religious tradition (Sunni, Arab, Asian, Black, other race, foreign citizen)

See References for full citation information. ** Author’s interpretation based on understanding of the published empirical findings.

provides a concise summary of the primary empirical measures in the six articles. The first row displays the number of Google Scholar citations for each article. The following three lines are information taken from the articles. In the final row of the table, I suggest how worship space likely reveals itself in the empirical models by conjecturing which included variables may be impacted by worship space characteristics that are left unmeasured.

This simple snapshot of six of the most frequently cited articles on religion and political participation reveals several interesting observations. First, while all six studies include broad and comprehensive measures of political participation, their measures of religious practice vary considerably. Djupe and Grant (2001) incorporate a scope condition of only examining individuals who attend church at least once per month, thereby limiting their study to regular adherents, looking narrowly at whether they practice a civic skill. Similarly, Jamal (2005) offers a narrow measure that is a higher bar than typically used by looking at mosque involvement beyond attending prayers. Dana et al. use two measures, one attitudinal and another looking at mosque involvement. In contrast, Driskell et al. (2008) employ a thicker measure of religious practice allowing for more variation of the independent variable by using a multidimensional index of activities and level of engagement. Jones-Correa and Leal (2001) and Ayers and Hofstetter (2008) include similar measures that fall between the leaner and more robust instruments.

Second, while none of the studies explicitly acknowledge or measure the independent effects of worship spaces on political participation, the mechanism of engagement within spaces designed for worship is implicit in all their analyses. Jones-Correa and Leal are specifically interested in understanding the associational role of churches. Djupe and Grant’s focus on civic skills specifies that these skills are in use at a religious congregation. Many of the activities included in the index of activities of Driskell et al., such as religious education, choir practice, and prayer meetings, typically take place in spaces designated for worship. Likewise, the three studies examining mosques all include engagement in a mosque in their measure of religious practice.

The studies of Christians do not assume that all houses of worship are homogenous. Rather, they include variables to separate out distinct religious traditions or denominations, in expectation that there will be variation across how religion is differentially practiced within these different faith systems. Consequently, the statistical models used by Jones-Correa and Leal, Djupe and Grant, and Driskell et al. include variables to distinguish among religious traditions, which is where omitted congregational effects would be likely to appear. These statistical models allow for identification of differences across religious denomination but fail to consider variation across worship spaces within the same religious tradition. Likewise, they cannot account for ways that houses of worship across different denominations may affect worshippers similarly. This possibility is underscored in Jones-Correa and Leal’s finding that religious denomination was not a statistically significant predictor of political engagement. They write, “Churches matter, but they matter because they function much like other associations, and they share associational characteristics across denominations” (Jones-Correa and Leal 2001, p. 763). These models may have the statistical effect of overestimating denominational differences and overlooking the significance of variation that occurs at the level of the worship community. The studies that look exclusively at Muslims lack measures that capture the details about mosque variation that may occur across the samples. Some of the samples include demographic variables that differentiate Muslims’ ethnic, racial, or linguistic identities, which might correlate with how mosques are organized, but, in general, we have no way of knowing how differences across mosques affect the authors’ findings.

The six articles examined in Table 1 are all based on samples taken from the United States, which is consistently overrepresented in the multidisciplinary social scientific study of religion. Even though much of the existing literature on the relationship between religious and political or civic engagement is drawn from empirical studies of the United States and is organized around concepts that presume both Christianity and democracy, relevant findings from this work should not be reflexively dismissed as irrelevant to the study of non-Western contexts. As Herzog et al. (2020) argue, much of the micro- and meso-level work on religious communities in the United States is well-designed and contextually appropriate for what the scholars seek to accomplish. The challenge is to consider the findings of the existing body of work and what elements can be applied to different contexts rather than assuming the universality of the American experience, which is characterized by significant variation across Protestant denominations, as well as overlapping identity markers between race, immigrant status, and religious affiliation.

Sociological studies of religion, which emphasize the salience of local context in considering how individuals perceive religious content and act on it, problematize overly simplistic models that treat houses of worship within the same religious tradition as interchangeable. Edgell (2012) encourages the study of empirical variation in practices and the institutions that facilitate them. She advocates “a practice-oriented and contextual approach to religious identity and experience that recognizes their social embeddedness” (p. 257). Guhin challenges scholars to “stop thinking of religion as a category and start thinking of it as a site through which to propose and test theories of social life” (Guhin 2014, p. 582). He notes, in particular, that while “The centrality of practices and their complicated interactions with beliefs are certainly clear in the study of religions…there is nothing essentially religious about these processes or their interactions” (Guhin 2014, p. 585) that should prevent the comparative study of practices.

Brenneman and Miller (2016) offer one of the few sociological studies to take seriously the potential role of religious buildings in shaping social interactions in religious settings. They point out the distinct social role played by worship spaces, writing that:

in worship, a group of people meets together at a certain time to interact in a very particular way, often claiming to interact with the transcendent or divine. Whereas a home provides a place for an individual or a small group to seek rest, shelter, and privacy, and to interact in patterned ways, in a church, temple, mosque, or synagogue, a group of people, most of whom are typically of no blood relation, gather to do something that can only be accomplished together.

(p. 84)

Among their arguments for focusing greater attention on physical structures, they point out that buildings play a communicative role, drawing those who are part of the community in while also providing legible signs to outsiders about which community worships in the space (p. 84). A similar point is made by Wittenberg (2006), who notes that in communist Hungary the ornate church buildings that remained intact as religious practice and instruction were being actively repressed by the state served two valuable purposes, as a symbol of local resistance and as a physical alternative to state-controlled spaces. Wittenberg’s observation supports Brenneman and Miller’s (2016) reminder that space is both a product and producer of social relations (p. 96).

3. Shared Conceptual Space across Denominational Boundaries

An extensive body of scholarship has detailed churches’ role in catalyzing the relationship between religious practice and political outcomes. As Verba et al. (1995) demonstrate, houses of worship can provide any number of resources for political participation. Scholars have documented the institutional role of churches in inculcating civic skills (Djupe and Gilbert 2006; Brown and Brown 2003; Jones-Correa and Leal 2001) and building social networks (Greenberg 2000; Johnston and Figa 1988; McDaniel 2008). As Lim and MacGregor (2012) and others describe, congregations bring people together socially, and the specific experience of shared social ties with co-religionists can help generate trust (p. 749). Djupe and Grant (2001) find support for several mechanisms that can draw worshippers into political engagement, including recruitment by a co-religionist, the presence of political meetings in church spaces, and perceptions of partisan leanings within a congregation. Any of these mechanisms suggest that worship spaces can foster political participation through forms of social diffusion.

Worship spaces play varying roles within local communities depending on the functions they are meant to serve in a specific religious tradition, as well as within the broader sociopolitical ecosystem in which they operate. For example, in his longitudinal analysis of local church networks following the development of a communist-party regime in Hungary, Wittenberg (2006) documents how religious instruction shifted over time from taking place in schools to occurring in churches as a response to party-level repression. Much of the work that compares what happens in worship spaces tends to focus on variation within a limited range of religious traditions, such as denominations within Christianity.

Some might argue that comparing houses of worship across a broader range of faith traditions, such as Islam, Judaism, and Christianity, is unproductive, assuming that differences across devotional beliefs, practice and ritual render mosques, synagogues, and churches too distinctive to yield meaningful insights when examined together. Yet, an assumption that the known differences in the religious functions of worship spaces inherently yields differentiation in all potential (and possibly unknown) roles that houses of worship might play can cause us to overlook their similarities. Looking specifically within the Abrahamic religious traditions, at the most fundamental level of comparison, mosques, synagogues, and churches share the primary role of providing a space for worship. Friday prayers at mosques, Saturday morning Shabbat services in synagogues, and Sunday services in churches typically include a sermon by a religious leader. Mosques, synagogues, and churches regularly organize other activities related to ritual belief and practice, including religious education, celebrations of major religious holidays, and coordinating charitable programs connected to religious values. Accordingly, worshippers may become involved in auxiliary functions of houses of worship and undertake tasks that are separate from their devotional responsibilities as adherents to their faith. As such, houses of worship have the potential to serve as social settings where co-religionists engage in community building that can mediate how their religious values shape other aspects of their social, political, and economic identities and practices.

Additionally, mosques, synagogues, and churches share a distinction from other types of mass religious associations, including the political, social, and charitable organizations that are affiliated with faith traditions. For example, the Muslim Brotherhood, Hillel International, and the Christian Democratic Union are associations comprised of individuals who consciously choose to join the group for any number of reasons. Individuals attending a house of worship—mosque, synagogue, or church—to fulfill devotional obligations are not necessarily signing up for active volunteerism as members of mass religious associations essentially do. While the populations of religious associations and houses of worship may overlap, they are distinct.

As Westfall (2019, p. 683) contends, the robust literature on Christianity demonstrates a broad consensus that attending church has at least some indirect effect on civic engagement and political participation, and that the mechanism behind this relationship comes primarily from group dynamics rather than specific denominational principles. Consequently, the theoretical connection between religious practice and political participation is independent of the content of religious beliefs, although research that accepts this premise and examines the relationship across religious traditions is limited. In general, there is little reason to expect that mosques, synagogues, and churches should look fundamentally different regarding the role they play in creating a local community of co-religionists who may encounter opportunities to build civic skills, encounter political messages, and be recruited into broader civic and political causes. While the structure of worship is different in Christian, Jewish, and Islamic traditions, all practices emphasize communal worship and fellowship among co-religionists; all prize sacred texts as a source of guidance; and all include a sermon by a religious leader at a weekly worship service.

While houses of worship perform similar functions irrespective of religious tradition, there are important differences in worship rituals across all faiths, including the Abrahamic belief systems that share similarities in the ways worship spaces are utilized. Jewish tradition emphasizes domestic worship practices with public worship at synagogue an extension of these rituals. Christianity calls on adherents to attend a weekly Sunday service. Islam prioritizes the performance of five daily prayers that are conducted at specific time intervals each day. While observers of Judaism and Christianity tend to visit the same house of worship on a consistent basis, attachment to a specific mosque may be weaker among Muslims, particularly in a Muslim-majority context. Observant Muslims who do not spend their days at home will typically offer their prayers at different locations over the course of the day, including at home, at a mosque, a prayer hall in a public building, or in another private location. Observant Muslims in a Muslim-majority context may affiliate primarily with the mosque that is closest to their house as this is the location where their family will typically fulfill obligations during Ramadan and celebrate religious feasts. Some mosques, however, will have a secondary and separate group of adherents comprised primarily of observant Muslim men employed or studying in the area who attend Friday midday prayers. These different levels of affiliation with houses of worship may contribute to variation in community norms and mechanisms of social diffusion among Christians, Jews, and Muslims.

In sum, while houses of worship vary in the significance that they may play within devotional practices and the full range of activities that take place within them, they share considerable functional similarities irrespective of denominational differences. When we look closely at the mechanisms that underpin most findings of interest in the study of religion and political engagement, they have little to do with specific rituals or doctrinal beliefs. Rather, the findings emphasize social processes and interactions that, while related to localized expectations about religious practice, also vary across time, place, and individual. There is little reason to reject findings from a predominantly American Christian experience as irrelevant to non-Christian contexts a priori simply because of the empirical origins from which they came to light. Social scientists, by definition, engage in comparison, and it is a matter of scholarly debate as to how to apply findings from one context to examine phenomena in another.

At the same time, the empirical foundations of congregational study as centered on American Protestantism raises several questions about how to apply relevant concepts and hypotheses effectively outside of the original field of study. For example, one of the most frequent—and extremely relevant—critiques of applying insights from the study of Christianity to non-Christian contexts is the common use of weekly attendance at worship services as a measure of piety in public opinion surveys (see Herzog et al. 2020). Indeed, no other faith tradition prizes weekly attendance at public worship to the same degree as Christianity. Similarly, the most used questions about religious practice that appear in prominent cross-national surveys, such as the World Values Survey, are frequently worded in a way that centers Christian practices, assuming monotheism and other beliefs commonly related to Christianity, such as a focus on the afterlife. For effective comparison between worshippers of different faith traditions, we need concepts that are sufficiently flexible to ensure neutrality across different devotional beliefs, together with measures that accurately capture details deemed significant within specified religious practice. If we can identify measures that are internally valid for the phenomena under investigation, these measures can serve as functional equivalents across different religious traditions to facilitate externally valid comparisons.

4. Conceptualizing Houses of Worship

To advance methodological tools for cross-denominational analysis, a thicker, more robust conceptualization of house of worship is warranted. I define house of worship as a permanent physical space where individuals of the same religious denomination come together with the primary purpose of fulfilling devotional duties related to their religious practice. This physical space has a consistent group of common believers who worship in it, as well as religious leaders who tend to the space and the worshippers who affiliate with it. Mosques, church buildings, temples, and synagogues are all examples of “houses of worship”, even though the significance of worship space may vary across different religious denominations. I have intentionally chosen the language “house of worship”, over “place of worship”, to anchor the concept to the related semantic themes of residence, dwelling, and family, which connote a sense of intimacy and community that is absent from the more neutral sounding “place”. The conceptualization of houses of worship proposed here is narrower in scope than Hassner’s (2009) concept of “sacred spaces”. Not all sacred spaces are houses of worship, and, depending on context and tradition, not all houses of worship may meet the criteria to be considered sacred spaces. Thus, while there may be some overlap between these two concepts, for analytical and empirical purposes they are distinct.

Houses of worship are also distinct from organized communities of believers, who may—or may not—be affiliated with a house of worship. Organized religious communities that lack a permanent physical space whose primary purpose is worship, such as followers whose circumstances necessitate worshipping in ad hoc locations, virtual communities that congregate online, or groups that gather in spaces primarily used for other purposes (such as domiciles), are not houses of worship. My proposed definition of house of worship does not deny the existence or significance of these communities but sees them as a separate concept. Additionally, not all “places” of worship may be “houses” of worship. For example, hospital and airport prayer rooms and chapels are typically not houses of worship even if worship can take place in them. Similarly, places of worship situated within communities of people who have devoted their lives to religious practice and leadership and are not open to common believers, such as spaces within monasteries, are also typically not houses of worship. The conceptualization of house of worship advanced here conjoins a physical space where worship takes place, a consistent group of common believers who use it, and the leaders who tend to them. All are necessary attributes.

This conceptualization of house of worship is not very different from Chaves’ (2004) definition of “congregation”,2 which is sufficiently neutral to apply to non-Christian contexts. In contrast to Chaves, the definition proposed here places fewer requirements on frequency and scheduling of activities, as these aspects can vary across religious traditions. Of greater significance, though, is a deliberate choice to move away from the term “congregation”, which has etymological roots in Christianity. The Latin root of “congregation”, greg, meaning “flock” took on a religious meaning in the 16th century when Protestant reformers began to use it as distinct from the term for “church”, yielding to the modern understanding of local believers (vocabulary.com and etymonlin.com, 5 May 2023). The term is still commonly associated with Protestantism, and thus a shift to “house of worship” is an opportunity to reorient the study of religious practice away from a default Christian center.

In determining the conceptual space occupied by houses of worship, Sartori’s (1970) ladder of abstraction is a helpful tool. Figure 1 offers a visual depiction of the ladder of abstraction for houses of worship. Figure 1 is illustrative of a general relationship between conceptual categories and is not meant to be comprehensive or exhaustive of all examples of houses of worship. Below the overarching concept of house of worship is Level 1, comprising categories that generally overlap with conceptual distinctions between faith traditions or branches of religion. We can compare houses of worship at this level, considering differences, for example, between mosques (Islam) and temples (Hinduism). Level 2 denotes houses of worship divided across denominations within faith traditions, such as distinctions between different houses of worship within branches of Christianity, Islam, Judaism, etc. Level 3 on the ladder specifies individual houses of worship. Houses of worship can be compared at any of these three levels, and the choice of level will depend on the research goal. There is no reason, in principle, why specific Catholic parishes cannot be compared with specific mosques, for example, if the goal of analysis is to understand variation in experience across faith traditions.

The ladder of abstraction articulated in Figure 1 exists separately from hierarchical structures that may be present within religious traditions. For example, many branches of Christianity, most notably the Catholic Church, have their own ecclesiastical structure in which houses of worship are embedded. This reality does not negate the conceptual space shared by these houses of worship and those of other religious traditions. Rather, by clarifying the role played by houses of worship in each analysis, scholars can sharpen which aspect of religious institutions or practice is most relevant to the research question.

A thicker conceptualization of houses of worship that is not attached to specific religious beliefs advances two methodological goals. First, it allows scholars to more deliberately build theoretical and empirical models that take seriously the role that houses of worship play in catalyzing religious beliefs into politically consequential outcomes. In other words, we cannot overcome the omitted variable bias of much of the extant quantitative literature on religion and political outcomes without having a clearer articulation of what variable is missing and why its omission is problematic. Second, a denominationally neutral conceptualization of houses of worship facilitates easier comparisons of individuals and communities belonging to different religious traditions. Cross-denominational comparisons are necessary if scholars seek to understand the power of religious concepts in general, and specific denominational differences, in particular.

5. Tools for Productive Comparison

The process of moving from abstract concept to empirical measurement necessitates consideration of a concept’s dimensions.

Table 2. Dimensions for Comparing Houses of Worship.

Dimension	Potential Sources of Variation
Local Context	How pluralist is religious landscape in which HW operates? Does HW belong to a majority or minority religion in local context? In national context? What is the broader religious ecosystem in which the HW operates? Are there other religious organizations (charitable, educational, etc.) providing services to co-religionists? What is the geographic location of the HW? Does location create particular constraints or opportunities for HW?
Regulation	Who has jurisdiction to decide where HW can be built? How difficult is it to build a new HW? Who owns HW? To what extent are the activities that take place inside HW regulated by: (a) state actors; (b) religious hierarchies outside of the HW; (c) decision-makers within HW? To what extent are a HW’s financial resources regulated by an external actor?
Level of Autonomy	Who has primary decision-making authority for HW over (a) worship/spiritual matters; (b) finances and budget; (c) personnel; (d) auxiliary activities? Is decision-making vested primarily in worshippers, in religious leaders within the HW, religious leaders outside the HW, or state authorities? How much autonomy does HW have vis-à-vis the religious denomination with which HW is affiliated? To what degree is a HW financially independent? Who preaches? Who decides who will preach? Which topics are forbidden from being preached?
Level of formalized organizational structure	Does the HW have a formalized organizational structure for making decisions and executing functions? Which functions are incorporated into a formal organizational structure? How broad is the constituency incorporated in the formal organizational structure?
Access to Resources	Does the HW have access to financial, property, or other in-kind resources from (a) the state; (b) an interdependent religious institution; (c) worshippers; (d) independent, autonomous revenue streams? What are the volume and variety of events (worship, educational, social, etc.) that take place within the HW?
Worshipper Profile	What is the size of the community using HW? What are the demographic characteristics of the worship community (age, ethno-linguistic composition, income level, socioeconomic status)?

outlines six distinct dimensions along which houses of worship could potentially vary: local context, regulation, level of autonomy, level of formalized organizational structure, access to resources, and worshipper profile. The boundaries between some of these dimensions may be fuzzy, but they reflect distinct attributes of houses of worship that likely impact how individual houses of worship function within a local and national context, as well as worshippers’ experiences within them. Some sources of variation, such as level of autonomy, are partly endogenous to religious denomination. Other sources of variation are a product of a combination of factors that reflect the broader social and political structures within which houses of worship exist. For example, two different mosques, one situated in the United Kingdom and another in Saudi Arabia would both vary on multiple dimensions while still serving as worship spaces for individuals professing adherence to Islam. Alternately, a synagogue and a congregational church both operating in New York city would be situated in the same local context yet might be differentially impacted by that context.

Depending on the research question of interest, a scholar might choose to focus on one or more dimensions for analysis. Each dimension itself requires further elaboration to be empirically useful, and there are multiple ways each dimension could be operationalized.

Table 3. Ordinal Scale Measuring Level of Formalized Organizational Structure.

	Level of Formalized Organizational Structure	Scores for Houses of Worship in Sample
Lower (0)	No formal organizational structure for determining and delegating responsibilities.
	Formal organizational structure for organizing worship events. Informal avenues or ad hoc committees determine remaining functions.	Mosque 2, Mosque 3
	Formal organizational structure delineates how most functions are carried out, but determination of most responsibilities occurs informally or through leader directives.	Mosque 1
	Formal organizational structure delineates how most functions are carried out, but some lines of responsibility remain unclear.	Pentecostal Church
Higher (4)	Formal organizational structure delineates how all functions are carried out. Regular meetings are held for various bodies with clear accountability mechanisms.	Catholic parish 1, Catholic parish 2, Presbyterian Church

and

Table 4. Aggregate Dichotomous Measures for Local Context.

Question	M1	M2	M3	CP1	CP2	Pent	Pres
Are there other equally accessible HW in the city for worshippers of same religious tradition?	Y	Y	Y	Y	Y	Y	Y
Does HW belong to local majority religion?	Y	Y	Y	N	N	N	N
Does HW belong to national majority religion?	Y	Y	Y	N	N	N	N
Are other religious organizations providing services to co-religionists?	Y	Y	Y	Y	Y	Y	Y
Is HW easily accessible to worshippers?	Y	Y	Y	Y	Y	Y	Y

offer examples of how one might apply these dimensions to specific empirical cases. For illustrative purposes, I have used information I have gathered from three mosques and four churches in one Indonesian city to demonstrate how these dimensions may be applied in comparative analysis. Table 2 offers an example of how the level of formalized organizational structure could be developed into an ordinal scale that allows for empirical comparison across houses of worship based on evaluation of qualitative data gathered from each case.

While an ordinal scale is logical for a dimension that can be articulated in ordered categories, not all the dimensions described in Table 2 meet this criterion. For example, local context encompasses several characteristics that are typically measured on a nominal scale. Table 4 displays one approach to examine local context by using a series of yes/no questions asked of each house of worship. The cells in Table 4 reveal that the local context in which these houses of worship are situated is largely pluralist within an ecosystem in which other religious organizations provide services to co-religionists of both majority and minority religions.

6. Conclusions

While social processes taking place within worship spaces are frequently assumed within studies that connect religious causal variables to political outcomes, the specific role houses of worship play in analytical frameworks is often underspecified. Consequently, houses of worship are regularly treated as thin concepts that are interchangeable, homogenous units within the faith tradition to which they belong. Indeed, several of the most cited scholarly publications examining religion and political participation do not mention houses of worship and do not account for their potential independent effects on worshippers in their empirical models. Yet, scholarship that looks closely at the interactions taking place within houses of worship has consistently shown that individual houses of worship vary in meaningful ways that cannot simply be reduced to the individual-level characteristics of worshippers or clergy members.

This review has sought to expand our methodological toolkit for critically analyzing the role of worship spaces in our scholarship by articulating a thicker conceptualization of house of worship. The conceptualization advanced here is neutral with regard to religious denomination, allowing for a reorientation of discussion of worship spaces and the community of people who use them away from language and approaches that normalize Christian institutions and practices as a default. By clarifying the conceptual space occupied by houses of worship and the relationship between houses, faith traditions, and denominations, we see that houses of worship can be compared productively along six different dimensions at three different levels of analysis. Individual researchers can focus their comparisons on the levels and dimensions most relevant to their research question.

The conceptualization of house of worship advanced in this review initiates a conversation that will be enriched by refinements from scholars working on religious traditions not touched on or fully elaborated here, such as Hinduism, Buddhism, Sikhism, Jainism, Judaism, and numerous others practiced across the globe. The tools suggested in this essay offer concrete suggestions for meaningful comparison across religious traditions, yet they do nothing to remedy the data limitations that have been a primary driver of omission of houses of worship from most quantitative studies of religion and politics. The operationalization recommended in Table 2 necessitates fine-grained, qualitative data about houses of worship that is absent from any publicly available survey of worshippers. In many respects, however, gathering some basic data about individual houses of worship within a given geographic area is a feasible data collection task and something that could be undertaken effectively and productively by a group of committed scholars.