Press and Pulpit: The Growth of Religious Magazines in Antebellum America

American https://asr.sagepub.com/ Sociological Review Pulpit and Press: Denominational Dynamics and the Growth of Religious Magazines in Antebellum America Adam Goldstein and Heather A. Haveman American Sociological Review 2013 78: 797 DOI: 10.1177/0003122413500274 The online version of this article can be found at: https://asr.sagepub.com/content/78/5/797 Published by: https://www.sagepublications.com On behalf of: American Sociological Association Additional services and information for American Sociological Review can be found at: Email Alerts: https://asr.sagepub.com/cgi/alerts Subscriptions: https://asr.sagepub.com/subscriptions Reprints: https://www.sagepub.com/journalsReprints.nav Permissions: https://www.sagepub.com/journalsPermissions.nav >> Version of Record - Sep 30, 2013 What is This? Downloaded from asr.sagepub.com at UNIV CALIFORNIA BERKELEY LIB on October 3, 2013 500274 2013 ASRXXX10.1177/0003122413500274American Sociological ReviewGoldstein and Haveman Pulpit and Press: Denominational Dynamics and the Growth of Religious Magazines in Antebellum America American Sociological Review 78(5) 797–827 © American Sociological Association 2013 DOI: 10.1177/0003122413500274 https://asr.sagepub.com Adam Goldsteina and Heather A. Havemana Abstract Religious economies theory, which views religious organizations as akin to single-unit firms competing for adherents in local markets, has three shortcomings that we solve by reconceptualizing religious organizations and developing a new theory of religious mobilization. First, we treat religious organizations as multi-unit entities operating in interdependent markets in a national field. Second, we incorporate insights from social movement theory to challenge the exclusive focus on the impetus to mobilize (competition) by also considering the capacity to do so (resources). Third, we consider competition within organizations as well as between them. To analyze mobilization directly, we study a key religious resource, magazines. We analyze original data covering virtually all faiths and affiliated magazines in antebellum America, a time of great religious ferment. Consistent with our conception of religious organizations, we find that competition played out mostly within a national field. Consistent with resource mobilization theory, we find that the geography of religious mobilization reflected variations in the availability of resources more than variations in the intensity of competitive pressures. Conceiving of religious organizations as translocal movement organizations rather than local firms better accounts for their behavior. Our analysis sheds light on group dynamics in general by revealing how translocal groups in modern societies mobilize and build identity through group media. Keywords religion, competition, organizations, media, mobilization, social movements Preaching of the gospel is a Divine institution – “printing” no less so. . . . They are kindred offices. The PULPIT AND THE PRESS are inseparably connected. . . . The Press, then, is to be regarded with a sacred veneration and supported with religious care. The press must be supported or the pulpit falls. (Editorial in the Christian Herald 1823, quoted in Hatch 1989:142; emphasis in the original) Pluralism has long been a hallmark of American religion (Ahlstrom 1972). The colonies were a University of California-Berkeley Corresponding Authors: Adam Goldstein and Heather A. Haveman, Department of Sociology, 410 Barrows Hall, Berkeley, CA 9472-1980 E-mail: [email protected]; [email protected] 798 home to not just the established Anglican and Congregational Churches but also dozens of dissenting faiths. After the Revolution, the disestablishment of state religions and successive waves of immigration further increased religious diversity (Ahlstrom 1972; Niebuhr 1929). Even more important were two series of religious revivals that swept across the United States from 1790 to 1861 (Butler 1990; Carwardine 1993). Revival leaders clashed with established religious authorities and seceded to found dozens of new religious groups. Persistent religious pluralism has directed research attention to interactions between denominations.1 A particularly prominent debate concerns the relationship between pluralism and the vigorous mobilizing efforts that have long characterized U.S. religious organizations. Religious economies theory (RET) builds on Adam Smith’s writings to argue that the competitive forces unleashed by religious pluralism produce stronger and more vibrant religious organizations (Finke and Stark 1988, 1992; Iannaccone 1994). This argument inverts Durkheimian theories of secularization that view religious heterogeneity as detrimental to religious vitality (e.g., Berger 1967). RET conceptualizes religious suppliers as akin to firms competing for adherents in local product markets. It argues that competitive pressure (measured either as the level of religious pluralism in a faith’s local market or its share of that market) forces churches to work harder to recruit and retain members, appeal to the unchurched, and distinguish themselves from rivals, all of which result in more vigorous mobilization efforts (Finke, Guest, and Stark 1996; Finke and Stark 1988; Stark 1998). This “competitive mobilization” argument has been a lightning rod for contention; critics have challenged both the theoretical precepts and the empirical evidence (e.g., Chaves and Gorski 2001; Land, Deane, and Blau 1991; Olson 1998; Voas, Crockett, and Olson 2002). Unfortunately, these debates have generated more heat than light, and in recent years the discussion has stagnated (Hungerman American Sociological Review 78(5) 2010; Smith 2008). There are several methodological reasons for this. The empirical evidence is limited, as most tests of RET do not measure religious mobilization directly; instead, they assess the argument indirectly by analyzing rates of religious adherence or levels of commitment by members, thereby conflating religious mobilization with its ostensible effects (Hill and Olson 2009; Wilde et al. 2010). Moreover, the most commonly used measures of religious competition and religious mobilization are both based on counts of members, so analyzing them produces artifactual correlations (Olson 1999; Voas et al. 2002). Finally, most work is crosssectional and so cannot pinpoint causality or rule out alternative explanations (Koçak and Carroll 2008). Beyond these methodological issues, the lack of progress in explaining religious mobilization reflects several theoretical shortcomings. First, the basic terms of debate have been structured around RET’s narrow conceptions of religious organizations and the contexts in which they interact. RET treats religious organizations as unitary entities, similar to single-establishment firms competing in local markets.2 Accordingly, scholars have tested RET by examining whether the association between competitive intensity and mobilizing efforts within a given geographic unit is positive or negative (e.g., Stark and McCann 1993). This approach assumes, either by theoretical premise or methodological fiat, that the factors driving religious mobilization are localized, that religious mobilization is locally oriented, and that religious organizations’ activities in different localities are independent of one another (but on the final point, see Land et al. 1991). Such assumptions ignore the history of U.S. religious organizations: they have long had congregations in multiple locations, and as early as the nineteenth century they oriented their actions toward a national field rather than purely local markets (Ahlstrom 1972; Finke and Stark 1992; Goen 1985; Smith 1962). More generally, such assumptions do not reflect the complex, multi-unit nature of religious Goldstein and Haveman organizations, whose members are connected horizontally through worship services and Sunday schools and vertically through religious-authority and agency structures (Chaves 1993; Stout and Cormode 1998). Past studies of religious mobilization thereby conflate the (hotly contested) question of whether religious organizations mobilize in response to competitive pressures with the (usually unexamined) presumption that such mobilization is organized in a manner akin to independent establishments competing in unrelated local markets. Second, debates about religious mobilization have been disconnected from theories of mobilization in the social movements literature, despite calls for such cross-fertilization (e.g., Demerath and Schmitt 1998). Most notably, RET’s proponents and critics both focus attention on the ostensible incentives to mobilize (competition) and ignore what social movement scholars have long known, namely that the capacity to do so (resource availability) is critical (Edwards and McCarthy 2004; McCarthy and Zald 1977). Third, debates have focused on competition between denominations and have given little consideration to competition within them from movements of schismatic subgroups, despite the central role of fragmentation in U.S. religion (Liebman, Sutton, and Wuthnow 1988). (For a notable exception, see Wilde 2007.) To explain the effect of such oppositional movements, social movement scholars point to situations where internal group conflict reflects larger social cleavages and argue that each side in the conflict responds to the other’s mobilizing activities (Kim and Pfaff 2012; Meyer and Staggenborg 1996; Zald and Useem 1987). In this article, we develop an alternative account by incorporating insights from social movement theories of mobilization and coupling them with a more sociologically grounded conception of religious denominations as multi-level, translocal organizations operating within national fields. We accept RET’s premise that religious organizations do compete: they woo people away from other 799 faiths, strive to retain the adherents they already have, and reach out to the unchurched.3 But we move beyond the micro-economic assumptions of RET to generate a more compelling explanation of the patterns by which religious organizations mobilize their adherents. First, we build on historical accounts of American religion (e.g., Goen 1985) and theories of churches as organizations (e.g., Chaves 1993) to consider translocal dynamics of competitive mobilization. Second, we build on resource mobilization theory in the social movements literature (Edwards and McCarthy 2004; McCarthy and Zald 1977) to examine how mobilizing efforts reflect variations in religious organizations’ capacities as well as their leaders’ motivations. Finally, we analyze how patterns of religious mobilization result from fragmentation within denominations as well as competition between them. The result is a series of more nuanced arguments about religious mobilization that go far beyond— and in some cases directly oppose—RET. To study religious mobilization directly, we examine the growth of a key religious resource, denominational magazines, in the United States from 1790 to 1860. We chose this case for two related reasons. First, this era saw much religious ferment: the rise of denominationalism and increased pluralism, the fragmentation of denominations by theological and political contention, the geographic expansion of many faiths, and the development of formal denominational structures. Second, denominational magazines were at the epicenter of American religion during this period. Magazines were a primary platform through which religious groups communicated, energized adherents, wooed the unchurched and members of other faiths, and forged distinctive identities (Hatch 1989; Marty et al. 1963; Nord 2004).4 Denominations of all stripes published magazines because “getting into print became the primary way to prescribe and contest values during the nineteenth century” (Moore 1989:219). As a result, the number of religious magazines exploded from seven in 1800 to 149 in 1830 and 328 in 1860. By the 800 1830s, religious magazines had become “the grand engine of a burgeoning religious culture, the primary means of promotion for, and bond of union within, competing religious groups” (Hatch 1989:125–26). One bibliographer (Albaugh 1994) estimated that as early as 1830, religious periodicals had more than 400,000 subscribers; many more read the even larger number of magazines published in the ensuing decades. Our study also has implications for other kinds of groups that constitute modern societies, including political factions, social movements, ethnic groups, schools of art and literature, and professions. Religious magazines are one type of group media that, in contrast to universal mass media, are affiliated with and oriented toward particular audiences (Blau 1998; Fine and Kleinman 1981), and so are excellent sites for understanding interactions within and between groups (Barnett and Woywode 2004; Olzak and West 1991). To put it simply, group media bind far-flung group members together. Group media are thus key structurating technologies in modern societies, where groups flourish even though their members are too scattered to meet face-to-face (Calhoun 1998; Thompson 1995). RELIGION AND PRINT MEDIA IN ANTEBELLUM AMERICA Trends in American Religion The period between the Revolution and the Civil War saw tumult and growth in American religion. Religious disestablishment and immigration, revivalism, and the withering of local monopolies increased the number of faiths. Older denominations and new upstart groups aggressively courted adherents. Denominations assumed their modern form as westward migration and energetic recruitment efforts yielded spatially dispersed adherents connected through national organizational structures. Yet denominations also fragmented repeatedly due to schisms spurred by evangelical challengers and the debate American Sociological Review 78(5) over slavery. We will describe these trends, all of which are implicated in the explanation we develop. The rise of pluralism. The gradual disestablishment of state religions, starting with New York in 1777 and ending with Massachusetts in 1833, leveled the playing field in the competition for souls. Waves of immigration from the 1830s onward, notably Irish Catholics and German Lutherans, Anabaptists, and Catholics, also contributed to religious diversity (Ahlstrom 1972; Niebuhr 1929). Even more important were the religious revivals that swept across the United States from 1790 to 1861 (Butler 1990; Carwardine 1993; Smith 1957). Revivals spawned dozens of new faiths—full-fledged churches such as the Disciples of Christ, utopian communities such as the Shakers, and small, unstable sects such as the Plymouth Brethren—and engendered increasing religious pluralism. In New England, Congregationalists dwindled from 62 percent of congregations in 1790 to 26 percent in 1860, while Baptists, Methodists, and Universalists expanded. In the South, Episcopalians were eclipsed by Methodists and Baptists. The West shifted from being a Baptist stronghold to hosting a mix of Methodists, Baptists, and Presbyterians, plus many smaller faiths. Religious competition. Denominations competed ideologically over theological tenets and strategically over members. Many of these battles played out in the pages of religious magazines (Hatch 1989; Marty et al. 1963; Nord 2004). Ample evidence shows that denominational publishing was driven partly by competitive threats from other faiths. For instance, the Presbyterian Christian’s Magazine (1806:ii) inveighed against preachers from upstart denominations with no theological training: The mischiefs arising from these sources are increased by the activity of a “zeal not according to knowledge.” . . . The duty of Christians is to confront and repel, not abet Goldstein and Haveman the enemy, nor admit him into their camp in order to subdue him. . . . The Christian’s Magazine will not be backward in strengthening their hands and stirring up their zeal in this contest. Some magazines fought direct battles against particular rivals. For instance, the aptly named Unitarian Defendant was launched to defend this breakaway faith from slurs by orthodox Congregationalists. In return, Spirit of the Pilgrims (1828:1) sought to defend against Unitarian slanders against Congregationalism, many of which were published in magazines: Misrepresentations, the most palpable and injurious, of the doctrines, preaching, and motives of the orthodox [Congregationalist], have been common for many years; and the continual repetition of them has by no means ceased. The apparent object has been to keep the members of Unitarian congregations from entering the doors of an orthodox church. . . . Unitarians have a magazine published here [Christian Examiner], upon which they spare no labor, and which is constantly employed in promoting their cause. We must have the means of meeting them on this ground. . . . They have found it necessary to make strenuous efforts to keep up the publication and circulation of their magazine; and surely, with our views of truth and duty, we cannot do less than they. Similarly, Connecticut Episcopalians countered losses to Congregationalists and Presbyterians by launching the Watchman (1819:1): It appears that “an association of gentlemen” has been formed, professedly for the purpose of “inculcating the doctrines which have ever prevailed in the great body of the Congregational and Presbyterian churches,”—but really, as one of its members is said to have unwarily voiced, “TO WRITE DOWN THE [EPISCOPAL] CHURCH IN CONNECTICUT!” . . . It cannot be supposed that the friends of the Church will view 801 attacks of this nature with indifference. . . . But it is in their indispensable duty, to defend and explain the principles which they profess, in such a manner, as to repel unfounded imputation, and to turn the weapons of assault back upon their adversaries. This they propose to do in the pages of the Watchman. Starting in the 1830s, the growth of Catholicism increasingly became the animus for Protestant mobilizing. One of the most vehement anti-Catholic periodicals was The Protestant, whose prospectus made clear its singular mission: The sole objects of this publication are, to inculcate Gospel doctrines against Romish corruptions—to maintain the purity and sufficiency of the Holy Scriptures against Monkish traditions. (quoted in Massachusetts Yeoman 1829:1) In their defense, Catholics also turned to magazines, such as United States Catholic Magazine, which defended Catholicism and shielded Catholics from organized attempts to convert them to Protestantism (Gorman 1939). Geographic expansion. As European settlement pushed westward, many denominations spread beyond their old regional strongholds. Following the Methodists’ successful example of circuit riders, Baptist ministerial outreach covered the nation as early as the 1820s; smaller faiths, like the Disciples of Christ, soon followed (Hatch 1989; Smith 1962). As a result, denominations’ activities and denominational leaders’ orientations became increasingly national in scope (Ahlstrom 1972; Goen 1985; Hatch 1989; Newman and Halvorson 2000). Yet basic conditions of life in antebellum America made direct communication among religious leaders and between leaders and their flocks difficult. People were spread thinly, especially along the frontier, and travel was slow and arduous. Moreover, few preachers were available to lead the far-flung faithful. Even in 1850, there 802 were only 23 clergy per 10,000 Americans— nine per 1,000 square miles (U.S. Census Bureau 2006). Religious leaders could not depend solely on sermons to communicate with the faithful, as the publishers of the Churchman’s Repository (1820:1) recognized: The want of a religious publication . . . serviceable to Episcopalians in this section of the Country [Connecticut], has long been acknowledged by all. . . . [Episcopalians] are few in number, are scattered over an extensive territory, and are generally so distant from each other, that some of them are almost exclusively confined to the ministrations of their respective pastors. It is difficult therefore to have those ministerial exchanges which . . . benefit their parishes. From these evils are apt to flow much ignorance . . . and a great want of union and zeal. Magazines were particularly useful for reinforcing adherents’ shared identities and disseminating information widely. For example, the Baptist Latter Day Luminary (1818:iii) noted magazines’ superiority to Bibles and religious tracts: The diffusion of Bibles and the publishing of the ever-lasting Gospel are, without doubt, the grand means which the spirit of the Lord will employ for subduing the nations to the dominion of the Son of God; but there are other means which have been succeeded with his blessing, and have conduced to the moral welfare of thousands. . . . [M]agazines have contributed greatly to the circulation of evangelic truth. . . . Magazines . . . have given rise to a new epoch in the history of intellectual improvement. They come to the purchaser on terms so reasonable, and at periods so regularly distant, as to render the procuring of them a circumstance unattended with inconvenience. . . . They portray and transmit characters and events as they daily occur. . . . They convey information through regions which larger publications [Bibles and books] cannot reach. American Sociological Review 78(5) High levels of geographic dispersion spurred even such committedly decentralized denominations as the Baptists and Disciples of Christ to publish many magazines. One-quarter of religious magazines’ titles made explicit claims about their geographic scope. Of these, 24 percent claimed to serve the entire nation and 35 percent a multi-state region like New England. Only 16 percent claimed to serve a single state and 25 percent a single county or municipality. Some religious magazines had mass followings. For instance, the Disciples’ Milennial Harbinger had 15,000 subscribers throughout the 1850s, and the Methodist Christian Advocate and Journal had 28,000 subscribers in 1828 and 50,000 in 1860. National religious structures. The antebellum era saw denominations evolve from loose affiliations of congregations based on common creed and religious authority into bona fide organizations. As they expanded geographically, denominations developed organizational structures that allowed them to pursue a broad array of translocal activities (Ahlstrom 1972; Hood 1977; Mathews 1969; Nord 2004; Smith 1962; Wright 1984). Missionary organizations converted natives and ministered to whites on the frontier, theological schools provided standardized training to ministers, Sunday school societies inculcated religious principles, and publishing houses printed and distributed Bibles, educational tracts, and periodicals. Even denominations with decentralized authority, such as the Baptists, developed federated structures with nested national, regional, and local operations that could draw resources from dispersed congregations and support activities across the nation (Ahlstrom 1972; Goen 1985). In building these federated structures, American denominations followed a broader pattern of organizational integration in nineteenth-century American civic life (Skocpol, Ganz, and Munson 2000). Magazines were a characteristic product of denominations’ multi-level organizational arrangements. Indeed, denominational publishing efforts represented an early instance Goldstein and Haveman of “a fundamental characteristic of modern denominationalism: the gathering of local and regional efforts into comprehensive organizational unity” (Smith 1962:78). For instance, explaining their rationale for forming another Congregationalist periodical in New England, the Christian Monitor’s (1814:1) founders pointed to the need to direct cultural resources to a state (Maine) where adherents had few churches they could attend regularly: Periodical publications have an extensive influence upon the minds morals and happiness of men. . . . But do any of these publications have an extensive circulation in the District of Maine? . . . The natural consequences of this state are forgetfulness of God and divine things, ignorance, error, profanity, a disregard of the Sabbath and the institutions of religion, immorality, and impiety. The means by which these evils must be arrested are the preaching of the gospel and the circulation of religious periodicals. The first of these can, at present, be but partially enjoyed. But, by the patronage and exertions of the well-disposed, a religious publication may be widely circulated and have a most beneficial effect upon the morals and religious state of this section of the Union. Internal competition. Even as they became increasingly integrated organizationally, denominations fragmented in disputes over theology and politics. Internecine conflicts originated in the revivals that swept the country from 1790 to 1861, amplifying the willingness of Americans in all walks of life to engage in religious debate and question church authorities (Ahlstrom 1972; Carwardine 1993; Finke and Stark 1992). Evangelical challengers opened theological rifts that split the Presbyterian, Congregational, Universalist, Methodist, Lutheran, Quaker, Baptist, Mennonite, and Dutch and German Reformed Churches. Later, debates unleashed by the anti-slavery movement became the primary division within many denominations (Young 2006), causing schisms in the Baptist, 803 Methodist, and Presbyterian Churches (Goen 1985). Magazines were central to the fracturing of denominations, both as vehicles for contention (Hatch 1989; King and Haveman 2008) and as outgrowths of differentiation. For example, there was a sharp increase in the number of religious magazines titled “Southern,” such as the Southern Methodist Quarterly Review, which cast itself as a pro-slavery alternative to the anti-slavery Methodist Quarterly Review. Before the American AntiSlavery Society was founded in 1833, only .8 “Southern” religious magazines were founded per year, on average; the number rose to 4.0 per year in the remainder of the 1830s, then to 5.0 per year in the 1840s, and 6.4 per year in the 1850s. Growth of Denominational Magazines These structural changes in American religion coincided with an explosion in the number of magazines with doctrinal or organizational allegiance to a denomination, from just one in 1790 to 328 in 1860. Virtually all denominations embraced magazines, not just large groups like the Presbyterians, Baptists, and Methodists, but also small ones like the Plymouth Brethren, Dunkers, and Christadelphians. Figure 1 plots the number of annual observations of magazines affiliated with each denomination. Denominational magazines were published across the country, as Table 1 shows. Magazines affiliated with older Protestant denominations were concentrated in New England and the Middle States; magazines affiliated with upstart denominations were spread more evenly across the nation; and Jewish and Catholic magazines were mostly in the Middle States and the South. Notwithstanding its broad geographic base, denominational publishing was largely an urban phenomenon: 37 percent of denominational magazines were published in the three largest cities (Philadelphia, Boston, and New York), 45 percent in smaller urban areas, and only 17 percent in rural areas. Magazines affiliated 804 American Sociological Review 78(5) Group Faith (Number of Magazine-Year Observations) Presbyterian Episcopal Congregational Mainline Protestant Lutheran German Reformed Quaker Dutch Reformed Moravian Mennonite Dunker Other Older Protestant Baptist Methodist Universalist Disciples of Christ Unitarian Swedenborgian Adventist United Brethren in Christ Evangelical Association Church of God Shaker Christadelphian Mormon Plymouth Brethren New Upstart NonProtestant Other Catholic Jewish Spiritualist Deist 0 200 400 600 800 1,000 1,200 Figure 1. Denominational Magazines by Faith and Group (Number of Magazine-Year Observations) with older Protestant faiths were slightly more likely than those affiliated with upstart denominations to be published in urban areas, whereas Jewish and Catholic magazines were exclusively urban phenomena. Most denominational publishing efforts were spearheaded by religious leaders. Information on the identities of magazine publishers and editors is spotty, but we were able to use prospectuses of 91 denominational magazines launched before 1820 to identify their founders. In this group, 56 percent of founders were local clergy, 24 percent were national or regional denominational authorities, 18 percent were laity, and 1 percent were unknown. Few later magazines published prospectuses, so to ascertain whether this pattern persisted, we drew a random sample of 30 denominational magazines founded between 1840 and 1860 and searched for data on their founders. Of the 21 magazines for which we could obtain data, 19 were founded by ministers or other denominational authorities; only two were founded by laymen. Publishing denominational magazines required five kinds of resources: printing presses and paper, contents, subscribers, distribution channels, and leaders with the time and money to provide editorial work and financial backing. These resources generally became easier to acquire over time (Haveman 2004). Printing and papermaking technologies advanced greatly in the first decades of the nineteenth century, and faster, cheaper, easier-to-use printing presses proliferated. The availability of written material—not just sermons and dry theological treatises, but also engaging stories, news, and anecdotes—grew as literacy and the eagerness of religious leaders to write for magazines increased. At the same time, population growth and the emergent market economy expanded the potential audience for magazines: there were many more people with more cash to spend on magazines. Finally, national investments in the postal system resulted in faster and more reliable delivery of magazines. Notwithstanding these secular trends, there were vast cross-sectional differences in Table 1. Geographic Distribution of Denominational Magazines, 1790 to 1860 Distribution by Geographic Region New England Faith Group Mainline Protestant Other Older Protestant New/Upstart Faith Jewish + Catholic Total 566 2 1,105 40 1,713 Middle States 1,052 521 1,124 199 2,896 Distribution by Municipality Size South West U.S. Total 280 75 591 137 1,083 344 201 850 45 1,440 2,242 799 3,670 421 7,132 Phil/Bos/ Other Urban NYC (pop >2,500) 1,050 194 1,199 206 2,649 821 452 1,776 215 3,264 Rural U.S. Total 371 153 691 0 1,215 2,242 799 3,666 421 7,128 Note: Faith Group: Mainline Protestant denominations are Presbyterian, Congregationalist, and Episcopalian. Other Older Protestant denominations are non-mainline groups: Dunker, Dutch Reformed, German Reformed, Lutheran, Quaker, Moravian, and Mennonite. New/Upstart faiths are Adventist, Baptist, Christadelphian, Church of God, Disciples of Christ, Evangelical Association, Methodist, Mormon, Plymouth Brethren, Shaker, Swedenborgian, United Brethren in Christ, Unitarian, and Universalist. Geographic Region: New England includes New Hampshire, Maine, Vermont, Massachusetts, Rhode Island, and Connecticut. The Middle States include New York, New Jersey, Pennsylvania, and Delaware. The South includes Maryland, the District of Columbia, Virginia, North Carolina, South Carolina, Georgia, Florida, Alabama, Mississippi, and Louisiana. The West includes Minnesota, Wisconsin, Michigan, Iowa, Indiana, Ohio, Illinois, Tennessee, Kentucky, Missouri, Arkansas, Kansas, Oklahoma, Texas, New Mexico, Arizona, Nevada, California, Oregon, Washington, Utah, and Colorado. 805 806 resource levels within and across locations, notably the adherents on whom denominations relied for subscription fees and the local clergy on whom they relied for content, editorial work, and financial backing. Within locations, more of the resources needed to publish magazines were available to larger denominations (those with more congregations and congregants) than to smaller denominations. Across locations, more resources were available where denominations had more congregations and congregants. EXPLAINING DENOMINATIONAL MAGAZINE GROWTH Virtually all denominations published magazines during our study period, but there was considerable geographic and temporal variation in the growth of this resource. To explain the patterns by which denominations mobilized to build and sustain magazine publications, we begin by drawing on RET because it is at the center of past debates. We then develop new arguments that recognize religious organizations as complex structures with units in many locations, and religious competition as a phenomenon that transcends local markets. We also incorporate ideas about resource mobilization and internal competition from social movement theory. Religious Economies Theory RET claims that competition compels religious suppliers to exert more vigorous efforts to market their faith and mobilize members (e.g., Finke and Stark 1988, 1992). The roots of this competitive mobilization thesis extend back to Adam Smith, who argued in Book V of The Wealth of Nations (1776) that monopolistic religions tend to produce indolent clergy who expend little effort to excite or maintain their adherents’ faith. Smith implied that competition in the market for souls would stimulate more energetic efforts. Weber echoed Smith’s argument in regard to the United States, where he was struck by the vigorous American Sociological Review 78(5) mobilizing efforts he observed among competing denominations (Scaff 2011). This competitive mobilization thesis is embedded in an approach that conceives of religious organizations as similar to for-profit firms. Both compete in market economies: for-profit firms for customers, religious organizations for adherents (Finke and Stark 1988, 1992). RET further assumes that religious organizations are like single-establishment firms operating in local markets because individual religious “consumers” choose which congregation to join within their local community (Finke et al. 1996). Because consumers’ actions are geographically localized, religious suppliers compete for them locally. RET’s basic prediction is clear: mobilization will be greater when and where a denomination experiences more intense local competition. RET conceptualizes local competition in two ways. First, competition is a function of local market structure. Specifically, competition increases with the number of denominations in a local market and with the equality of denominations’ market shares (their numbers of adherents). Competition thus increases as religious markets become less monopolistic and more pluralistic. RET predicts that denominations will sustain more mobilization tools, such as magazines, in more pluralistic religious markets. Moreover, as religious markets become more pluralistic, denominations must work harder to distinguish themselves from other faiths and demarcate their own niches, answering the questions “Who are we?” and “What makes us unique?” Because magazines are ideal instruments to define denominations’ distinctive identities, groups should publish more of them as pluralism increases: Hypothesis 1: As pluralism in a location increases, the number of magazines a denomination publishes there will increase. Second, the intensity of competition a denomination experiences depends on its position within a local market. Whereas dominant churches can afford to be complacent, embattled minority churches must work hard to recruit and retain members, so they must Goldstein and Haveman mobilize their small pools of resources intensively (Finke and Stark 1992; Stark and McCann 1993). RET thus predicts that mobilization will be inversely correlated with local market share. Empirical tests of this prediction have mostly been cross-sectional (e.g., Hill and Olson 2009; Stark 1998; Stark and McCann 1993). Thinking dynamically leads to the conclusion that competition is a function of trends in market share more than levels: weakening competitive positions (declining market shares) will spur leaders of even dominant groups to take action, whereas stable competitive positions will make leaders of even minority faiths less inclined to mobilize aggressively (Wilde 2007; Wilde at al. 2010). For example, in early-nineteenth-century New England, when Congregationalists were losing market share to Baptists and Methodists (although they remained numerically dominant), prominent minister Lyman Beecher exhorted his fellow clergymen to proselytize more vigorously (Finke and Stark 1992). We thus expect that as their share of a local religious market declines, denominations will publish more magazines there: Hypothesis 2: As a denomination’s share of a local market decreases, the number of magazines it publishes there will increase. Beyond RET: Religious Organizations Compete in a National Field Because RET holds that churches compete locally, empirical tests tend to analyze locallevel variations in competitive pressures. Yet religious leaders’ strategic outlook may be less parochial than RET assumes. Historians have shown that the U.S. religious field was nationalized by the early nineteenth century (Ahlstrom 1972; Goen 1985; Smith 1962). Accordingly, we now consider how mobilization reflects denominations’ extralocal orientations and how denominations’ actions in any particular location depend on their actions in other locations. This analysis yields predictions that go beyond those derived from RET, but that are congruent with the notion, central to RET, that competition spurs religious organizations to deploy resources to recruit and retain adherents. 807 National market share. First and most simply, if competition is national rather than local, resources like magazines will be deployed in response to national, not local, market position. This suggests that denominations will publish more magazines as their national market position becomes more tenuous: Hypothesis 3: As a denomination’s share of the national religious market decreases, the number of magazines it publishes nationally will increase. If we find support for Hypotheses 2 and 3, then any observed effect of national market share likely reflects the aggregation of local competitive pressures. If, however, we find support for Hypothesis 3 but not Hypothesis 2, it would seem that religious mobilization was a function of denominations’ positions at the national, not local, level. Competition across multiple local markets. As denominations expanded across the nation, they came to resemble multi-unit firms with operations in multiple markets (Edwards 1955). As a result, rivalries between denominations played out simultaneously in multiple local markets. In such circumstances, we suspect denominations’ actions in one local market are shaped by their relations with rivals in others. If so, the competitive impetus to mobilize in one local market depends not only on that market’s attributes, but also on its position vis-à-vis other markets in the national field. In particular, we expect denominations will be more likely to mobilize in a given location when their interactions with rivals are more concentrated in that location. Geographic expansion delocalizes religious competition, unmooring it from any single location. The more denominations encounter their rivals across multiple markets, the more geographically dispersed and less localized their rivalries become. This should prompt denominations to consolidate their mobilizing efforts. Therefore, the impetus to engage in separate mobilization efforts in any single market will decline with the extent to which 808 religious organizations encounter rivals in multiple local markets: Hypothesis 4: As a denomination increasingly encounters its local rivals in multiple other markets, the number of magazines it publishes in the focal market will decrease. Countering geographic dispersion. A nationalizing religious field also creates new demand for connective tissue to sustain translocal solidarity. Competition from other faiths in a national field may drive religious organizations to integrate their adherents into a more cohesive community by forging stronger bonds between them. If so, the growth of denominational magazines can be seen as an integrative response to the dispersion of adherents across ever-broader swaths of space. Adherents’ geographic dispersion increases the usefulness of magazines because they, like other media, weave “invisible threads of connection” between readers (Starr 2004:24). Magazines supplement purely ritualistic bases of collective identification, allowing far-flung adherents to interact and reinforcing their shared beliefs (Anderson 1991; Calhoun 1998; Park 1940). This suggests that increasing spatial scale should heighten the importance of translocal technologies, such as magazines, for coordinating and integrating communities. Simply put, operating across more locations necessitates publishing more magazines to bind coreligionists together: Hypothesis 5: As the number of locations in which a denomination operates increases, the number of magazines it publishes nationally will increase. Beyond RET: Availability of Resources for Religious Mobilization RET argues that mobilizing efforts will reflect variations in the amount of competition a religious organization experiences. Within this frictionless micro-economic model, mobilization is assumed to follow unproblematically from competitive incentives. In American Sociological Review 78(5) contrast, social movement theorists abandoned such incentive-based accounts because they ignore an organization’s capacity to mobilize—that is, the resources organizations possess to support mobilizing activities (McCarthy and Zald 1977). Places where an organization has the strongest motivations to mobilize (where it is most embattled) are often precisely the places where it lacks the necessary resources (Edwards and McCarthy 2004). Thus, even if we accept that competitive pressure is a primary motivation for mobilizing efforts, the distribution of resources will be more critical than the distribution of competitive pressure in explaining spatial and temporal patterns by which denominations actually create and sustain mobilizing tools like magazines.5 Publishing magazines requires multiple resources, many of which are tied to location. Denominations rely on local clergy for sermons and educational articles to fill the pages of their magazines, and local congregation members for subscription fees and other content (e.g., letters, poems, and stories).6 Other resources to support magazines, such as funding and publishing infrastructures, are tied to the location of national religious organizations but can be deployed across the nation to support local mobilizing (McCammon 2001). Religious organizations have the greatest capacity to sustain mobilizing devices like magazines in locations where they have the most resources: strongholds where they have the greatest concentration of congregations, congregants, and clergy; these strongholds also tend to be where their national organizations are situated (Edwards and McCarthy 2004). Resource mobilization theory implies that denominations will use slack resources in their strongholds to create organizational infrastructures, such as denominational magazines, that support adherents’ faith in locations where resources are scarce and adherents are socially isolated. In proposing this, we recognize that denominations are complex, multi-unit structures that amass and allocate resources from multiple units to pursue common purposes (Chaves 1993), and we treat Goldstein and Haveman denominational magazines as vehicles for redistributing cultural resources across space. Why might religious organizations in high-market-share strongholds mobilize on behalf of adherents elsewhere? The challenges of social reinforcement are particularly acute for minority denominations because their adherents have fewer day-to-day interactions with coreligionists (Perl and Olson 2000); they are more isolated from the social fabric of their faiths and more vulnerable to secular forces or overtures from proselytizers (Berger 1967). Denominational magazines help mitigate this problem insofar as they represent efforts to extend religious canopies across space. Whether driven by solidaristic or competitive motives, the key point is that religious magazines can compensate for disparities between low-market-share regions where cultural resources are most needed and highmarket-share regions where they are most available. If denominational magazines redistribute cultural resources across locations, then such resources should flow from rich to poor regions. In other words, resource mobilization theory suggests that magazine production should be concentrated in areas where denominations have the most slack resources (where their market share is highest), and magazines should be distributed to areas where they have the greatest need (where their market share is lowest). Hypothesis 6: As a denomination’s share of a local market increases, the number of magazines it publishes there will increase. Hypothesis 6 directly contradicts Hypothesis 2 from RET because RET emphasizes the disciplining effects of local competition, whereas we point to organizational infrastructures that can redirect resources from one area to another. Hypothesis 2 predicts magazine publishing will reflect the efforts of church leaders in embattled low-market-share locations, whereas Hypothesis 6 predicts magazines will be produced in high-market-share strongholds and distributed to low-marketshare locations.7 Our analysis does not oppose 809 the core insight of RET, that competition motivates mobilization; it merely shifts the analytic emphasis away from motivations to mobilize toward resource capacities to support mobilization. In other words, the key distinction between our theory and RET lies in our assumption of which theoretical construct market share captures: competitive weakness and the impetus to mobilize (RET), or capacity to do so (our theory). Resource mobilization theory also suggests that to the extent denominations do mobilize in response to local competition, this effect will be conditional on the availability of resources (McCarthy and Zald 1977). Therefore, denominations will be more likely to mobilize against competition when and where they experience more competition and have more resources to respond to that competition. This suggests that the positive effect of local pluralism will be stronger in places where a denomination has greater local capacity to mobilize—that is, higher market share: Hypothesis 6a: The positive impact of local pluralism on the number of magazines a denomination publishes will be amplified as a denomination’s local market share increases. Beyond RET: Competition within Denominations Debates about religious mobilization have focused on competition between denominations. But the history of American religion reveals that much competition occurred within denominations. Antebellum denominations were embedded in wider fields of cultural and political contention (Niebuhr 1929) and they internalized cultural divisions from society at large. These divisions frequently erupted into conflicts that split denominations into competing factions. Most prominent during the antebellum era were evangelical movements and the North–South divide, both of which sundered many denominations. These two waves of antebellum schisms were similar to subsequent episodes of fragmentation in American religion, in that disaffected subgroups mobilized their distinctive identities to create new 810 sects that more closely accorded with their particular cultural orientations, political positions, and desire for autonomy from central religious authorities (Liebman et al. 1988; Sutton and Chaves 2004). Social movement theory holds that schisms should spur the launch of new magazines because they increase the number of distinctive subgroups and energize existing subgroups (Meyer and Staggenborg 1996; Zald and Useem 1987). Challenger movements typically create new organizational infrastructures to support their differentiating efforts (Carroll and Swaminathan 2000). Competition is most intense between ideologically proximate groups: groups similar enough to occupy the same general resource space but different enough to prevent solidarity and cooperation (Barnett and Woywode 2004). Proximate challenges elicit particularly strong counter-mobilization efforts because they threaten basic identities and domain claims. Media like magazines are especially useful resources to deploy in such conflicts because media are not simply incarnations of alternative moral visions, but also political instruments in struggles over claims to truth, purity, and heritage. Taken together, these ideas imply that the growth of denominational magazines may stem from intradenominational discord and fragmentation: Hypothesis 7: As a denomination’s internal fragmentation increases, the number of magazines it publishes nationally will increase. Table 2 summarizes the predictions derived from RET and each part of our own theory. For each prediction, the table details the level of analysis, the explanatory factors involved, and the direction of the predicted effect on the number of denominational magazines published. RESEARCH DESIGN Sampling Plan We tested these hypotheses by analyzing magazines affiliated with U.S. religious organizations from 1790 to 1860. Our analysis American Sociological Review 78(5) starts in 1790 because that is the first year for which good data are available on many explanatory variables. Only five religious magazines were published before this date. Our study ends in 1860, the year before the Civil War broke out; the war disrupted many denominational activities, including their publishing efforts. Our analysis focuses on denominations, rather than congregations, because denominations are “the fundamental church structure of this country” (Smith 1962:97; see also Ahlstrom 1972; Hall 1998; Niebuhr 1929). We analyzed magazine publishing activity for all 22 denominations founded before 1860 for which we could find good data: Adventist, Baptist, Catholic, Church of God, Congregational, Disciple of Christ, Dunker, Dutch Reformed, Episcopalian, German Reformed, Jewish, Lutheran, Mennonite, Methodist, Moravian, Mormon, Presbyterian, Quaker, Shaker, Swedenborgian, Unitarian, and Universalist.8 Together, these denominations accounted for 94.3 percent of all congregations in the United States in 1776, 98.8 percent in 1850, and 99.6 percent in 1860 (the only years for which reliable national counts are available), so our data quite accurately represent the field of U.S. religion. We conducted analyses at two levels, local and national, because the processes we probe are theorized as occurring at these two levels. Previous research defines the locations where competition occurs as municipalities, counties, or states (Chaves and Gorski 2001). We define locations as states for three reasons. First, as explained earlier, many religious magazines circulated far beyond their sites of production. Second, prior empirical tests show that using a more granular definition of locations makes little difference (Chaves and Gorski 2001). Third, it was extremely difficult to find serially and cross-sectionally reliable state-level data on this time period; it would be impossible to piece together data on smaller geographic units. For the state-level analysis, our data comprise one observation per denomination per year for every state in which the denomination had congregations; for the national-level Table 2. Summary of Predictions and Findings about Religious Magazine Publishing Perspective Religious Economies Theory Religious Organizations Compete in a National Field Availability of Resources for Mobilization Competition within Denominations Level of Hypothesis Analysis Predicted Effect Result Local pluralism (complement of the Herfindahl index of concentration in the local market) Local market share (percentage of local congregations affiliated with the denomination) +  –  Average market share across all locations where the denomination has congregations Contact between denominations in multiple local markets Spatial scale (number of locations where the denomination has congregations) –  –  +  +  +  +  Explanatory Variables 1 Local 2 Local 3 National 4 Local 5 National 6 Local 6a Local Local market share (percentage of local congregations affiliated with the denomination) Local market share amplifies the effect of local pluralism 7 National Internal fragmentation (cumulative number of schisms) 811 812 analysis, they comprise one observation per denomination per year. We studied each denomination starting in 1790 (for denominations founded before that date) or the year each was founded. For the state-level analysis, the start of each time series depended on two events: the state must have entered the Union and the denomination must have at least one congregation in the state. Data and Measures Dependent variable. The outcome we studied is the number of magazines affiliated with a given denomination (nationally or in a given state) in a given year. Although much organizational research analyzes foundings and failures separately, we focused on growth in the number of magazines because we are interested in the growth of denominations’ infrastructures, to which each magazine contributes. Data for the dependent variable come from a list encompassing virtually every magazine published in the United States from colonial times to the onset of the Civil War, which was gathered from nine primary and 90 secondary sources (for details, see Haveman 2004). Our dataset explicitly excludes newspapers, pamphlets, and occasional tracts, in accordance with the definition used by historians (Mott 1930, 1938; Tebbel and Zuckerman 1991). We define a “magazine” as a publication containing a variety of written and pictorial material, with more than transient interest, published at regular intervals. Magazines’ contents are more varied than those of pamphlets and newspapers, and they are of longer lasting interest than newspapers. To exclude newspapers and pamphlets from our database, we relied on information in histories of publishing (e.g., Mott 1930) and bibliographies of the magazine and newspaper industries (e.g., Albaugh 1994; Brigham 1962), as well as inspection of archived copies of periodicals. A denominational magazine is one that proclaims a doctrinal or organizational affiliation with a particular denomination. For magazines available in archives, we coded American Sociological Review 78(5) denominational affiliation on the basis of contents and editorial statements; for magazines not available in archives, we relied on magazine titles, industry histories (e.g., Mott 1930), and bibliographies (e.g., Albaugh 1994). Our analysis excluded non- and interdenominational publications, leaving 832 denominational magazines. Independent variables. We constructed our measures of local religious market structure and denominations’ shares of those markets from state-level counts of congregations (for a full description, see King and Haveman 2008). Ideally, we would construct measures using data on both congregations and congregants, but data on congregants simply do not exist for most of our study period. Studying the period 1890 to 1926, Koçak and Carroll (2008) report that both measures yield similar results. Furthermore, we estimated regression models with fixed denomination effects, which obviate biases that might result from systematic differences in congregation size across denominations (Perl and Olson 2000). To capture local competition, we measured local pluralism in each state each year using the complement of the Herfindahl index of market concentration. Although this measure has been criticized for producing artifactual correlations between pluralism and religious participation (Olson 1999; Voas et al. 2002), that does not happen in our analysis because our dependent variable (number of magazines published) is not composed of the same units as the pluralism index (number of congregations). To capture each denomination’s local market position, we measured its local market share in each state each year, based on the number of congregations.9 To capture a denomination’s position in the national market, we calculated its national market share across all states. We measured each denomination’s spatial scale as the number of states where it had congregations. We measured the degree of contact in multiple local markets by counting the number of markets outside the focal market in which the Goldstein and Haveman focal denomination met a local rival, summed across all local rivals. We scaled this count by the number of markets in which the focal denomination operated to yield a proportion. This proportion ranges from zero, when a denomination had no contact with local rivals in any of its other markets, to one, when a denomination met all local rivals in all of its other markets. This measure is identical to one used in previous research on for-profit organizations competing in multiple geographic markets (Haveman and Nonnemaker 2000). We detail the construction of this measure in the online supplement (https://asr. sagepub.com/supplemental). Finally, we counted the cumulative number of schisms in each denomination, based on standard historical reference works (Mead 1980; Melton 2003; Williams 1998). We lagged this measure by two years to capture effects of subgroup mobilization before the schismatic event. This measurement strategy treats denominations that experienced schisms as continuing to constitute a single denomination whose stock of magazines is expected to grow precisely because of its increased internal variety. Model Specification and Estimation Methods State-level analyses. Our dependent variable is a count: the number of religious magazines affiliated with each denomination in each state, each year. Because this variable is over-dispersed, we estimated negative binomial models. Our dependent variable is the number of magazines published, not the number founded, so we modeled a growth process: change over time in the number of denominational magazines in each state. Because past size affects future size, we included the lagged dependent variable in our models (Heckman and Borjas 1980). Each denomination could have congregations in multiple states, and each state could be home to multiple denominations, so we were dealing with cross-classified data, not hierarchically clustered data (Goldstein 1987; Rabe-Hesketh and Skrondal 2008). To accommodate this data structure, we estimated 813 mixed-effects models with crossed random effects for denomination and state. The first effect captures unobserved factors that might affect each denomination’s propensity to publish magazines; the second captures unobserved factors that might affect magazine publishing in each location. The models we estimated took the following form:   list ¼ exp ayist �1 þ b’xist �1 þ zi þ zs ; s2 ½yist � ¼ list t; where list is the fundamental parameter of the negative binomial distribution, yist–1 is the lagged dependent variable, xist–1 is a vector of lagged explanatory and control variables, ζi is the random effect for denomination i, ζs is the random effect for state s, and τ is the scale parameter. We estimated these models using the glmmADMB package in R (Bolker et al. 2012). National-level analyses. Again we modeled a growth process, but because we aggregated data across many states, the average number of magazines published was 5.4 and the range was 0 to 44. Accordingly, we estimated fixed-effects linear models of the following form: yit ¼ a yit�1 þ b’xit�1 þ gi þ eit ; where yit is the dependent variable (the number of magazines published by denomination i across all states at time t), yit–1 is the lagged dependent variable, xit–1 is a vector of lagged explanatory and control variables, gi is the denomination-specific fixed effect, and eit is the error term. Because the lagged dependent variable is correlated with denomination-specific fixed effects, ordinary-least-squares estimates can be biased (Nickell 1981). To circumvent this problem, we estimated fixed-effects instrumentalvariable (FE-IV) models via two-stage least-squares, using the xtivreg2 routine in Stata (Schaffer 2007). This estimation strategy is well-suited to the structure of our data (max t = 70, n = 22). We followed the standard practice of instrumenting yit–1 with yit–2 because the latter is highly correlated with the former but not with the time-demeaned idiosyncratic error. We confirmed our choice of 814 instrument with a Sargan test of the instrument’s validity; we also compared the firstand second-stage R2 to ensure adequate instrument strength. We corrected for heteroskedasticity and for serial autocorrelation. Finally, we estimated robust standard errors. American Sociological Review 78(5) variables renders point estimates and standard errors unstable, however, so their effects should be interpreted cautiously. RESULTS State-Level Analysis Control Variables State-level models. We controlled for denomination size10 (number of congregations in the focal state in the focal year) and denominational growth rate in the focal state (a five-year moving average). We also controlled for state population (in millions) and the percent state urban population (places with more than 2,500 inhabitants). We distinguished between urban and rural areas using historical data on municipal populations (Moffat 1992, 1996; Purvis 1995:253; U.S. Census Bureau 1998). We included nationallevel controls for immigration (U.S. Census Bureau 2006) and an index of industrial production (Davis 2004). Finally, we included several time-varying controls related to the overall growth of literacy and infrastructure: miles of postal roads (in the focal state) and magazine postage rate (in cents), using data from postal histories (John 1995; Kielbowicz 1989; Rich 1924); maximum printing speed (in sheets per hour), based on information from printing-industry histories (Berry and Poole 1966; Moran 1973; Thomas 1874); and number of colleges in the United States, based on data from Marshall (1995). National-level models. We included the same time-varying controls, with all variables calculated for the country as a whole. For example, denomination size is the total number of congregations across all states. Multicollinearity Multicollinearity does not pose a problem for our independent variables. Variance-inflation factors for our independent variables are less than 2.1 in the state-level analyses and less than 4.0 in the national-level analyses. Multicollinearity among some of our control Table 3 presents descriptive statistics on all variables in the state-level analysis, and Table 4 shows results of the negative-binomial regressions. Model 1 in Table 4 contains only control variables. Model 2 adds three variables to test all main-effect hypotheses robustly. Local pluralism had a significant positive effect on the number of denominational magazines published, which supports Hypothesis 1 and is consistent with RET. As denominations met local rivals in more local markets, they published fewer denominational magazines locally, which supports Hypothesis 4. This suggests that as competitive interactions became more geographically dispersed, denominations consolidated their publishing efforts. RET and our theory yield opposing predictions about local market share, due to their different interpretations of this variable. Local market share had a significant positive effect, which confirms Hypothesis 6 (our theory) and disconfirms Hypothesis 2 (RET). Denominations were more likely to publish where their share of the local market was increasing; mobilization thus occurred when and where denominations possessed growing concentrations of resources. This model controls for state population and denominational size in the focal state, so the positive effect of local market share does not reflect the geographic distribution of denominations’ members, but rather indicates more intensive mobilizing efforts in denominations’ strongholds. We found the same result using an alternative measure of local denominational resource concentration: the fraction of a denomination’s total congregations in a given state each year. To save space, we report this analysis in the online supplement. How large are these effects? Figure 2 plots standardized exponentiated coefficients derived from the main effects estimates in Table 3. Descriptive Statistics for Variables Used in the State-Level Analysis Variable # Mean Standard Deviation Minimum Maximum 1 2 3 4 5 6 7 8 9 10 11 12 13 Number of Denom. Magazines in the State Denom. Size (# of congregations in state) Denom. Growth Rate in State State Population/1,000,000 Percent State Urban Population Index of Industrial Production (national) Immigration/1,000 (national) Number of Colleges (national) Miles Postal Roads in the State/1,000 Magazine Postage Rate (cents) Local Pluralism (1 minus Herfindahl index) Local Market Share Contact with Rivals in Multiple Locations 1 2 3 4 5 .121 .124 0 .633 .430 1.15 0 13 88.1 172 1 2341 .173 .477 –.667 24 .756 .666 .047 3.88 .550 –.039 .448 .261 .148 .157 .149 .310 –.040 .169 .169 –.240 –.053 .301 .017 .156 .147 .160 .294 –.033 .004 .594 –.264 –.046 –.082 .040 .045 .038 .037 –.010 –.091 .006 .045 .243 .316 .332 .284 .266 .316 .330 .828 –.031 –.063 –.064 .301 .378 –.159 –.116 –.273 –.116 6 62.7 54.1 4.82 159 .721 .972 .426 –.195 .098 –.095 .241 7 8 136 103 4.28 281 103 66 22 258 .703 .405 .433 –.182 –.188 .065 .095 –.087 –.090 .178 .223 Note: This table is based on 14,389 state-year observations on 22 U.S. denominations in 33 states between 1790 and 1860. 9 4.84 3.99 0 20.4 –.097 .145 –.124 –.146 10 11 12 13 10.1 31.0 2 300 .742 .098 .237 .871 .099 .140 0 .868 .739 .131 0 1 .000 .017 –.254 –.043 –.373 –.135 815 816 American Sociological Review 78(5) Table 4. Mixed-Effects Negative-Binomial Models (with Crossed Unit Effects) of the Number of Magazines Published by Each Denomination in Each State in Each Year (1) Lagged Number of Denominational Magazines in the State Denomination Size (number of congregations in the state/100) Denominational Growth Rate in the State State Population/1,000,000 Percent State Urban Population Index of Industrial Production (constant $1860/100) U.S. Immigration (national) Magazine Postage Rate (cents/100) Number of Colleges (national) Maximum Printing Speed/10,000 (national) Miles of Postal Road in State/1,000 .578*** (.018) –.029** (.010) –.098 (.055) –.387*** (.078) –.485 (.538) –.406* (.186) 1.43*** (.272) –.065 (.082) .005*** (.001) .008 (.073) .111*** (.013) Local Pluralism (complement of the Herfindahl index) Local Market Share Contact between Denominations in Multiple Local Markets Local Pluralism × Local Market Share Constant Standard Deviation of the Latent DenominationSpecific Parameter Standard Deviation of the Latent State-Specific Parameter Log-likelihood Number of Observations (2) .508*** (.017) –.133*** (.011) –.065 (.053) –.191* .494*** (.017) –.148*** (.012) –.056 (.052) –.175* (.076) –1.74** (.539) –.297 (.178) 1.84*** (.264) –.075 (.082) .006*** (.001) .017 (.069) .112*** (.013) 2.84*** (.076) –.136* (.541) –.309 (.177) 1.79*** (.263) –.074 (.082) .006*** (.001) .013 (.069) .117*** (.013) 1.39** (.445) 3.92*** (.208) –1.023** (.327) (.533) .598 (.756) –.924** (.327) 5.356*** (1.160) –4.421*** –3.477*** (.251) –5.342*** (.516) .571 .464 1.11 –7,510 13,990 (3) 1.19 –7,336 13,975 (.545) .465 1.19 –7,325 13,975 Note: This table presents results of multi-level mixed-effects negative-binomial regressions of the number of magazines published by a denomination in each state and each year for 22 U.S. denominations from 1790 to 1860. These models include crossed latent effects for state and denomination. Standard errors are in parentheses below parameter estimates. *p < .05; **p < .01; ***p < .001 (two-tailed t-tests). Goldstein and Haveman Model 2. (Effects of control variables are omitted to save space.) The height of each bar is calculated as exp[βσ(x)], which corresponds to the factor by which the number of magazines published is expected to increase, based on a one-standard-deviation increase in the level of the focal variable, holding all other variables constant. A one-standarddeviation increase in local pluralism increased the expected number of denominational magazines published by 43 percent, and a onestandard-deviation increase in local market share increased the expected number of magazines by double that amount, 88 percent. The negative effect of contact with rivals across multiple locations was much smaller: a onestandard-deviation increase in the extent to which denominations met local rivals across multiple other markets reduced the expected number of magazines by 10 percent. Model 3 in Table 4 adds an interaction between local pluralism and local market share to test whether mobilization in response to local pluralism was contingent on local resource availability, as we predicted. The interaction term is positive, which supports Hypothesis 6a. This indicates that the effect of increasing local pluralism was stronger when and where denominations had growing concentrations of resources to support mobilization. The contingent effect of local pluralism can be seen by comparing predicted counts. When local market share is low (2 percent), increasing local pluralism from one-standarddeviation below the mean to one-standarddeviation above the mean yields a 35 percent increase in the expected number of magazines published, holding all other variables and random effects at their means. When local market share is high (30 percent), the same-magnitude increase in local pluralism yields a 91 percent increase in the expected number of magazines published. Thus, a denomination’s capacity to mobilize in more pluralistic environments depended on it possessing a large local market share. In contrast, increasing a denomination’s local market share had a big impact on local magazine publishing, even at low levels of pluralism. When pluralism was one-standarddeviation below the mean (.644), increasing a 817 denomination’s local market share from 2 to 30 percent (approximately two standard deviations) yields a 230 percent increase in the expected number of magazines published, holding all other variables and random effects at their means. The bottom of Table 4 shows estimated standard deviations for denomination- and statespecific random effects. The denominationspecific effect varied much less than the state-specific effect, which indicates that differences across states outweighed differences across denominations. That is, variations in local context shaped the growth of religious media more than did underlying variations in theology, formal authority, membership criteria, and practices. National-Level Analysis Table 5 presents descriptive statistics for all variables in our national-level analysis, and Table 6 shows results of this analysis. Model 1 in Table 6 includes just the control variables. Model 2 adds all theoretical variables to test all main-effect hypotheses robustly. National market share had a significant negative effect, which supports Hypothesis 3. This finding suggests that weakening national competitive positions mobilized denominations to publish more magazines. This result is robust to an alternative measure, the focal denomination’s share across only those states where it had congregations (instead of all states in the Union). Given the positive effect of local market share in the state-level analysis, the negative effect of national market share suggests that antebellum religious leaders were less locally oriented than RET assumes. Consistent with Hypothesis 4, the effect of spatial scale (number of states where the denomination had congregations) was positive and significant. This supports the claim that denominational magazines grew in response to the challenges of organizing the faithful across space. The effect of spatial scale is independent of the effect of denominational size (number of congregations), which suggests that the former variable taps into geographic expansion in particular, not 818 American Sociological Review 78(5) Percent Change in Number of Denominational Magazines 100% 80% 60% 40% 20% Contact across Multiple Markets 0% Local Pluralism Market Share -20% Figure 2. State-Level Denominational Magazine Growth: Standardized Factor Coefficients for Theoretical Variables overall denominational growth. Finally, consistent with Hypothesis 7, increasing internal denominational differentiation, as measured by the cumulative number of schisms, had a significant positive effect on the number of denominational magazines. This suggests that denominations experiencing more internal discord published more magazines, as both established and splinter groups sought to distinguish themselves and mobilize supporters. Figure 3 charts standardized coefficient estimates for all statistically significant parameters in Model 2. The height of each bar represents the expected change in the number of magazines published by a denomination nationally, given a one-standard-deviation increase in the corresponding variable, holding all other variables constant. Not surprisingly, expansion of the postal system (β = .36) and increasing denominational size (β = .83) both yielded substantial growth of denominations’ publishing infrastructures. The standardized effects of geographic expansion (β = .62) and market share (β = –.40) were also quite large. Increasing internal fragmentation had a smaller standardized effect (β = .22), due in part to this variable’s low variance. Disambiguating the Effect of Denominational Market Share These findings invite further questions. One notable ambiguity concerns the opposite effects exerted by denominational market share at the two levels of analysis: positive at the state level and negative at the national level. This pattern indicates that, although denominations published more magazines overall as their share of the increasingly crowded national market declined, mobilization was concentrated in states where denominations were growing relative to local rivals. The second result is anomalous from the perspective of RET, but taken together, the two results are consistent with our model of religious organizations operating in multiple, interdependent markets and sharing resources across space: denominational actors in resource-rich areas mobilized to address overarching challenges in the national field. One possible objection to this interpretation is that because the state-level models present fixed-effects estimates, results show that denominations’ magazine-publishing efforts expanded where their local market Table 5. Descriptive Statistics for Variables Used in the National-Level Analysis Variable # 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 Mean Standard Deviation Minimum Maximum 4.97 7.95 0 44 10.1 23.0 .015 192 .045 .104 –.286 2 45.9 50.2 4.17 158 .066 .076 .002 .2 1.08 .728 .038 .251 .229 .551 .020 3.00 1.14 1.01 .096 2.82 82.4 63.2 16 258 .052 .081 .000 .375 Number of Denominational Magazines Denomination Size (# congregations) Denominational Growth Rate Index of Industrial Production Maximum Printing Speed (pages/hour) Postage Rate for Magazines ($) Postal Roads (millions of miles) Immigration Number of Colleges National Market Share Spatial Scale (number of states) Internal Fragmentation (cumulative schisms) .830 –.024 .466 .460 .496 –.163 .453 .470 .567 .770 .654 –.028 .268 .267 .273 –.079 .246 .270 .796 .648 .604 –.003 –.016 –.006 .004 .008 –.010 –.038 –.023 –.059 .980 .934 –.252 .802 .973 –.052 .371 .315 .919 –.236 .824 .940 –.050 .374 .316 –.329 .853 .961 –.060 .371 –.120 –.290 –.269 .026 –.104 .356 .788 –.056 .361 .350 –.055 .361 .327 .627 .408 Note: This table is based on 1,314 annual observations of 22 U.S. religious denominations between 1790 and 1860. 9 10 11 13.3 9.09 2 35 .452 12 .381 .786 0 4 819 820 American Sociological Review 78(5) Table 6. Two-Stage Least-Squares Fixed-Effects Instrumental-Variable (2SLS-FE-IV) Models of the Number of Magazines Published by Each Denomination Each Year (1) Lagged Number of Denominational Magazines (instrumented) Denomination Size (# congregations/100) Denominational Growth Rate Index of U.S. Industrial Production (constant $1860/100) Maximum Printing Speed (# pages per hour/10,000) Postal Roads/100,000 Magazine Postage Rate ($/100) Immigration/1,000,000 Number of Colleges/100 .925*** (.012) .024*** (.004) .216 (.225) –.035 (.049) .152 (.214) .594** (.189) –2.06 (5.93) .410 (.597) –.275 (.303) National Market Share Spatial Scale (number of states) .866*** (.018) .038*** (.006) .210 (.231) –.110* (.049) .398 (.218) .504* (.227) –4.28 (6.01) –.417 (.683) –.159 (.367) –4.908** (1.639) .068*** (.014) .254** (.091) Internal Fragmentation (cumulative schisms) Number of Observations (2) 1,346 1,314 Note: This table presents regressions of the number of magazines published by a denomination across the nation in each year for 22 U.S. denominations from 1790 to 1860. Both models are corrected for serial autocorrelation and heteroskedasticity. Standard errors are in parentheses below parameter estimates. *p < .05; **p < .01; ***p < .001 (two-tailed t-tests). share was growing, not where it was greatest. To assess this possibility, we reexamined the pooled cross-sectional relationship between local market share and magazine foundings. Figure 4 plots magazine foundings by the local market-share rank of the denomination that founded each magazine. It shows that the founding denomination was usually one of the largest in the focal state. Figure 5 plots magazine foundings by rank of the state where a magazine was founded in terms of the state’s share of the denomination’s total congregations. It shows that denominations usually founded magazines in the state where they had the most congregations. Thus, contrary to RET’s claim that denominations mobilize more vigorously where they are small, the largest denominations in a location were disproportionately active in publishing magazines they could use to support their congregations elsewhere. The convergent results of the longitudinal and cross-sectional analyses lend further support to our theory. DISCUSSION AND CONCLUSIONS This article began by lamenting that the study of religious mobilization has been structured around simplistic conceptions of religious organizations and markets, a narrow focus on the impetus to mobilize and neglect of the Goldstein and Haveman 821 1.00 Change in Number of Magazines Published by Denomination .80 .60 .40 .20 National Market Share .00 Postal Roads Denomination Size -.20 Internal Fragmentation Spatial Expansion -.40 -.60 Figure 3. National Denominational Magazine Growth: Standardized Beta Coefficients for All Statistically Significant Effects resource capacity to do so, a general disregard for intra-denominational processes, a dearth of direct empirical tests of mobilization, and a paucity of dynamic models. Our analysis sought to place the debate on firmer theoretical and methodological ground. We developed predictions based on a conception of denominations as complex multi-unit organizations (Chaves 1993) that operate in multiple, interdependent markets within a national field. Building on social movement theories of mobilization, we assessed the resource capacity to mobilize as well as the impetus to do so (Edwards and McCarthy 2004; McCarthy and Zald 1977). And we explicitly analyzed the effects of intra-denominational discord alongside inter-denominational competition. Methodologically, we captured processes of mobilization more directly than most previous studies by analyzing the growth of an organizational resource—denominational magazines. Finally, we applied dynamic techniques to longitudinal data, which is an advance on previous cross-sectional research. The last column in Table 2 summarizes our findings. At the local (state) level, we found partial support for the original (locally focused) version of RET. Increasing local pluralism increased the number of denominational magazines published. But counter to RET, denominations also increased their publishing as their share of the local market increased. This finding supports our argument, derived from resource mobilization theory, that religious mobilization reflects the geographic distribution of resource availability more than the geographic distribution of competitive pressures. This interpretation is bolstered by our finding that the positive effect of local pluralism was amplified when and where a denomination’s stock of resources (as captured by local market share) was growing. Although mobilizing actions tended to concentrate in denominations’ strongholds, our findings suggest that competition and mobilization were also structured by processes at the national level. First, denominations published more magazines overall as their shares of the increasingly crowded national market declined. Second, local magazine-publishing efforts diminished when a denomination’s interactions with rivals became more geographically dispersed and thus delocalized; that is, as a denomination 822 American Sociological Review 78(5) 160 Number of Magazine Founding Events 140 120 100 80 60 40 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Market Share Rank of Denomination in the State where Founding Occurred Figure 4. Magazine Foundings by Denominational Market-Share Rank in the Founding State Note: This figure plots the number of magazine foundings across all denominations and years. This outcome differs from the outcome analyzed in the regressions: growth in the total number of magazines published by a given denomination in a given year (in a given state). Number of Magazine Founding Events 250 200 150 100 50 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 State's Rank in Terms of its Share of the Founding Denomination's Congregrations Figure 5. Magazine Foundings by State Rank (of the State’s Share of the Founding Denomination’s Congregations) Note: This figure plots the number of magazine foundings across all denominations and years. This outcome differs from the outcome analyzed in the regressions: growth in the total number of magazines published by a given denomination in a given year (in a given state). Goldstein and Haveman 823 met its local rivals in a larger number of other markets. Third, national magazine-publishing efforts expanded as denominations dispersed geographically and had to work harder to bind their adherents together. Fourth, the internalization of national cultural and political fractures, as proxied by schisms, prompted more publishing efforts, as upstart and established subgroups sought to craft distinctive identities and mobilize supporters. Together, these results imply that competition matters for religious mobilization, but not in the manner that RET suggests. The growth of publishing did not reflect a process whereby local entrepreneurs responded to competitive pressures they faced in their immediate environments. Rather, the evidence is more consistent with a dynamic whereby denominational actors in resource-rich strongholds mobilized to address overarching concerns they faced in the national field. We assessed the sensitivity of our results to the idiosyncratic social positions of the denominations we studied and to alternative estimation and measurement strategies. Results of these robustness checks, discussed in the online supplement, were virtually identical to those shown here. about the mobilizing effects of religious competition are muddled because they fail to account for the basic facts that mobilization depends on resources, and that religious organizations, like all modern groups, are structured translocally. Our analysis also opens several further lines of inquiry. By incorporating insights from resource mobilization theory, we drew new linkages between social movement theory and the sociology of religion. This complements studies of how secular social movements piggyback on religious organizations’ resources (e.g., Morris 1984). Future research could continue in this vein, perhaps by analyzing political opportunity structures for religion or the use of religious resources as framing devices in secular debates. Future research might also investigate the different ways that religious organizations are translocally structured, perhaps by contrasting moreand less-centralized denominations, and thus the different ways they are affected by processes in local versus national fields. Our analysis controlled away many of these denomination-level differences by estimating models with fixed denomination effects, but they certainly merit investigation. Theoretical implications for the sociology of religion. What should we take away from this analysis? In our view, the underlying theoretical problems with RET stem not from its emphasis on competition per se, but rather from its presumption that competition can be modeled in a manner akin to rivalries among single-unit firms in purely local markets. RET begins from a conception of religious groups as aggregations of local actors and seeks explanatory leverage from variations in the strength of competitive pressures to mobilize. By proceeding from a more nuanced understanding of denominations as multi-level organizations and attending to the distribution of denominations’ endowments, we developed a better explanation of the dynamics by which antebellum denominations expanded their media infrastructure. Our findings suggest that previous debates Implications for the study of other kinds of modern social groups. Our analysis also carries broader implications for research on other modern, translocal groups. Researchers must closely scrutinize the geographic assumptions embedded in the theories they test, as well as the implicit assumptions that research designs make about the spatial structuring of social processes. In particular, our finding of a spatial disconnect between the processes that catalyze competitive mobilization and the locations where mobilizing responses occur highlights limitations of the local ecological study designs that dominate research on intergroup relations, including the literatures on ethnic and racial conflict (e.g., Olzak and West 1991). This issue extends beyond the well-known fact that clustering and spatial diffusion processes may create interdependence between local 824 American Sociological Review 78(5) units (e.g., Cunningham and Phillips 2007). Rather, researchers must be attentive to two facts: (1) modern groups are complexly structured as translocal communities with nested units and (2) group members’ concerns and actions may be oriented beyond the bounds of their immediate localities and organizational units. Our analysis focused on the antebellum era, but these facts have only become more pronounced as groups’ ability to transfer monetary, symbolic, and organizational resources across space has increased. Acknowledgments The authors would like to thank Steve Vaisey for helpful suggestions concerning model specification. Funding This research was funded by an NSF Graduate Fellowship to the first author and by grants to the second author from the NSF (SES-0727502 and SES-0096016) and the Institute for Research on Labor and Employment at the University of California-Berkeley. Notes 1. 2. 3. 4. 5. For simplicity in exposition, we use the word denomination to refer to both long-established, stable groups like the Presbyterians and newer, often unstable fringe groups like the antebellum Adventists, which are often labeled sects or movements. Recent theoretical restatements by RET proponents recognize denominations as multi-level entities with core–periphery structures (Stark and Finke 2000). But the analytic strategies developed to test RET still treat religious organizations as if they were atomized entities. We leave aside the issue of commitment displayed by adherents, which is theoretically and empirically distinct from efforts to recruit and retain them. Magazines were not the only instrument religious organizations deployed during this period: circulating preachers, camp meetings, sermons reprinted as pamphlets, Bibles, and missionary tracts were also powerful mobilizing resources. Magazines, however, were more widespread and prolific than these other resources, and they were published by religious groups large and small. They are thus particularly useful for analysis of religious mobilization in this era. RET proponents occasionally reference resource mobilization theory in passing. For instance, Finke and colleagues (1996) cite McCarthy and Zald’s (1977) classic statement to motivate a control variable for urban population density. Our point, 6. 7. 8. 9. 10. however, is that RET proponents have failed to grapple with the fundamental implications of resource mobilization theory. Denominational magazines also rely on the federal government for distribution through the postal system. Our analysis controls for the expansion of the postal system. Our test of this argument is limited because our dependent variable captures only the production side. A definitive test would require explicit data on resource flows, such as the distribution of magazines across states. Unfortunately, such data are not available for the antebellum era. We followed Koçak and Carroll (2008) and distinguished among denominations rather than between groups within denominations, such as branches of the Baptists. Market share can also be calculated across locations (rather than across denominations) as the percentage of a denomination’s congregations in the focal location. We report results using the standard measure but find identical results using the alternative measure. Denomination size may be endogenous. If magazines did help denominations grow, then the causal dynamics may be nonrecursive. To assess this possibility, we re-estimated the 2SLS model, treating denomination size as endogenous. The c-statistic test revealed, at most, marginal evidence of endogeneity ( p = .11). References Ahlstrom, Sydney E. 1972. A Religious History of the American People. New Haven, CT: Yale University Press. Albaugh, Gaylord P. 1994. History and Annotated Bibliography of American Religious Periodicals and Newspapers. Worcester, MA: American Antiquarian Society. Anderson, Benedict. 1991. Imagined Communities. London: Verso. Barnett, William P. and Michael Woywode. 2004. “From Red Vienna to the Anschluss: Ideological Competition among Viennese Newspapers during the Rise of National Socialism.” American Journal of Sociology 109:1452–1500. Berger, Peter L. 1967. The Sacred Canopy. New York: Anchor Books. Berry, William T. and Herbert E. Poole. 1966. Annals of Printing. London: Blandford Publishing. Blau, Judith R. 1998. “Group Enmity and Accord: The Commercial Press in Three American Cities.” Social Science History 24:395–413. Bolker, Ben, Hans Skaug, Arni Magnusson, and Anders Nielsen. 2012. “Getting Started with the glmmADMB Package.” Retrieved February 21, 2013 (https:// glmmadmb.r-forge.r-project.org/glmmADMB.pdf). Brigham, Clarence. 1962. History and Bibliography of American Newspapers, 1690–1820, Including Additions and Corrections, 1961, 2 Vols. Hamden, CT: Archon Books. Goldstein and Haveman Butler, Jon. 1990. Awash in a Sea of Faith: Christianizing the American People. Cambridge, MA: Harvard University Press. Calhoun, Craig. 1998. “Community without Propinquity Revisited: Communication Technology and the Transformation of the Urban Public Sphere.” Sociological Inquiry 68:373–97. Carroll, Glenn R. and Anand Swaminathan. 2000. “Why the Microbrewery Movement?” American Journal of Sociology 106:715–62. Carwardine, Richard R. 1993. Evangelicals and Politics in Antebellum America. New Haven, CT: Yale University Press. Chaves, Mark. 1993. “Denominations as Dual Structures: An Organizational Analysis.” Sociology of Religion 54:147–69. Chaves, Mark and Philip S. Gorski. 2001. “Religious Pluralism and Religious Participation.” Annual Review of Sociology 27:261–81. The Christian Monitor. 1814. “Preface.” Vol. 1, p.1 (retrieved from American Periodicals Service on May 13, 2003). The Christian’s Magazine. 1806. “Introduction” Vol.1, pp. ii–xvi (retrieved from American Periodicals Service on May 14, 2003). The Churchman’s Repository for the Eastern Diocese. 1820. “Introductory Essay.” Vol. 1, p.1 (retrieved from American Periodicals Service on May 14, 2003). Cunningham, David and Benjamin T. Phillips. 2007. “Contexts for Mobilization: Spatial Settings and Klan Presence in North Carolina, 1964–1966.” American Journal of Sociology 113:781–814. Davis, James H. 2004. “An Annual Index of U.S. Industrial Production, 1790–1915.” Quarterly Journal of Economics 109:1177–1215. Demerath, N. J. and Terry Schmitt. 1998. “Transcending Sacred and Secular.” Pp. 381–400 in Sacred Canopies, edited by N. J. Demerath, P. Dobkin Hall, T. Schmitt, and R. H. Williams. New York: Oxford University Press. Edwards, Bob and John D. McCarthy. 2004. “Resources and Social Movement Mobilization.” Pp. 116–52 in Blackwell Companion to Social Movements, edited by D. Snow, S. Soule, and H. Kriesi. Oxford: Blackwell. Edwards, Corwin D. 1955. “Conglomerate Bigness as a Source of Power.” Pp. 331–59 in Business Concentration and Price Policy, edited by G. Stigler. Princeton, NJ: Princeton University Press. Fine, Gary Alan and Sherryl Kleinman. 1981. “Mass and Specialized Media.” American Journal of Sociology 87:173–77. Finke, Roger, Avery Guest, and Rodney Stark. 1996. “Mobilizing Local Religious Markets: Religious Pluralism in the Empire State, 1855 to 1865.” American Sociological Review 61:203–218. Finke, Roger and Rodney Stark. 1988. “Religious Economies and Sacred Canopies: Religious Mobilization in American Cities, 1906.” American Sociological Review 53:41–49. 825 Finke, Roger and Rodney Stark. 1992. The Churching of America 1776–1990. New Brunswick, NJ: Rutgers University Press. Goen, C. C. 1985. Broken Churches, Broken Nation. Macon, GA: Mercer University Press. Goldstein, Harvey. 1987. “Multilevel Covariance Components Analysis.” Biometrika 74:430–31. Gorman, Robert. 1939. Catholic Apologetical Literature in the United States, 1784–1858. Washington, DC: Catholic University of America Press. Hall, Peter. 1998. “Religion and the Organizational Revolution in the United States.” Pp. 99–115 in Sacred Canopies, edited by N. J. Demerath, P. Dobkin Hall, T. Schmitt, and R. H. Williams. New York: Oxford University Press. Hatch, Nathan O. 1989. The Democratization of American Christianity. New Haven, CT: Yale University Press. Haveman, Heather A. 2004. “Antebellum Literary Culture and the Evolution of American Magazines.” Poetics 32:5–28. Haveman, Heather A. and Lynn Nonnemaker. 2000. “Competition in Multiple Geographic Markets: The Impact on Market Entry and Growth.” Administrative Science Quarterly 44:232–67. Heckman, James J. and George J. Borjas. 1980. “Does Unemployment Cause Future Unemployment?” Economica 47:247–83. Hill, Jonathan P. and Daniel V. Olson 2009. “Market Share and Religious Competition: Do Small Market Share Congregations and Their Leaders Try Harder?” Journal for the Scientific Study of Religion 48:629–49. Hood, Fred J. 1977. “Evolution of the Denomination among the Reformed of the Middle and Southern States, 1780–1840.” Pp. 139–60 in Denominationalism, edited by R. E. Richey. Nashville, TN: Abingdon. Hungerman, Daniel. 2010. “Rethinking the Study of Religious Markets.” Pp. 257–75 in The Oxford Handbook of the Economics of Religion, edited by R. McCleary. Oxford: Oxford University Press. Iannaccone, Laurence R. 1994. “Why Strict Churches Are Strong.” American Journal of Sociology 99:1180–21. John, Richard R. 1995. Spreading the News. Cambridge, MA: Harvard University Press. Kielbowicz, Richard B. 1989. News in the Mail. New York: Greenwood Press. Kim, Hyojoung and Steven Pfaff. 2012. “Structure and Dynamics of Religious Insurgency: Students and the Spread of the Reformation.” American Sociological Review 77:188–215. King, Marissa D. and Heather A. Haveman. 2008. “Antislavery in America: The Press, the Pulpit, and the Rise of Anti-slavery Societies.” Administrative Science Quarterly 53:492–528. Koçak, Özgecan and Glenn R. Carroll. 2008. “Growing Church Organizations in Diverse U.S. Communities, 1890–1926.” American Journal of Sociology 113:1272–1315. Land, Kenneth C., Glenn Deane, and Judith R. Blau. 1991. “Religious Pluralism and Church Membership: 826 A Spatial Diffusion Model.” American Sociological Review 56:237–49. Latter Day Luminary. 1818. “Introduction.” Vol. 1, p.iii (retrieved from American Periodicals Service on May 14, 2003). Liebman, Robert C., John R. Sutton, and Robert Wuthnow. 1988. “Exploring the Social Sources of Denominationalism: Schisms in American Protestant Denominations, 1890–1980.” American Sociological Review 53:343–52. Marshall, Gloria J. 1995. The Survival of Colleges in America. PhD dissertation, Stanford University School of Education, Stanford, CA. Marty, Martin E., John G. Deedy Jr., David W. Silverman, and Robert Lekachman. 1963. The Religious Press in America. New York: Holt, Rinehart and Winston. Massachusetts Yeoman. 1829. “Prostestantism vs. Popery.” December 19, Vol. vii(18) p.1. Mathews, Donald G. 1969. “The Second Great Awakening as an Organizing Process, 1780–1830: An Hypothesis.” American Quarterly 21:23–43. McCammon, Holly J. 2001. “Stirring Up Suffrage Sentiment: The Formation of the State Woman Suffrage Organizations, 1866–1914.” Social Forces 80:449–80. McCarthy, John D. and Mayer N. Zald. 1977. “Resource Mobilization and Social Movements: A Partial Theory.” American Journal of Sociology 82:1212–41. Mead, Frank S. 1980. Handbook of Denominations in the United States. Nashville, TN: Abington Press. Melton, Gordon J. 2003. Encyclopedia of American Religions, 7th ed. Detroit, MI: Gale Research. Meyer, David S. and Suzanne Staggenborg. 1996. “Movements, Countermovements, and the Structure of Political Opportunity.” American Journal of Sociology 101:1628–60. Moffat, Riley. 1992. Population History of Eastern U.S. Cities and Towns, 1790–1870. Metuchen, NJ: Scarecrow Press. Moffat, Riley. 1996. Population History of Western U.S. Cities and Towns, 1850–1990. Lanham, MD: Scarecrow Press. Moore, Laurence. 1989. “Religion, Secularization, and the Shaping of the Culture Industry in Antebellum America.” American Quarterly 41:216–42. Moran, James. 1973. Printing Presses. Berkeley: University of California Press. Morris, Aldon. 1984. The Origins of the Civil Rights Movement. New York: Free Press. Mott, Frank L. 1930. A History of American Magazines. Vol. 1, 1741–1850. Cambridge, MA: Harvard University Press. Mott, Frank L. 1938. A History of American Magazines. Vol. 2, 1850–1865. Cambridge, MA: Harvard University Press. Newman, Richard S. and Peter L. Halvorson. 2000. Atlas of American Religion. Lanham, MD: Rowan and Littlefield. Nickell, Stephen. 1981. “Biases in Dynamic Models with Fixed Effects.” Econometrica 49:1417–26. American Sociological Review 78(5) Niebuhr, Richard. 1929. The Social Sources of Denominationalism. New York: Henry Holt. Nord, David P. 2004. Faith in Reading. New York: Oxford University Press. Olson, Daniel V. A. 1998. “Religious Pluralism in Contemporary U.S. Counties.” American Sociological Review 63:759–61. Olson, Daniel V. A. 1999. “Religious Pluralism and US Church Membership: A Reassessment.” Sociology of Religion 60:149–73. Olzak, Susan and Elizabeth West. 1991. “Ethnic Conflict and the Rise and Fall of Ethnic Newspapers.” American Sociological Review 56:458–74. Park, Robert E. 1940. “News as a Form of Knowledge.” American Journal of Sociology 45:675–77. Perl, Paul and Daniel V. A. Olson. 2000. “Religious Market Share and Intensity of Church Involvement in Five Denominations.” Journal for the Scientific Study of Religion 39:12–31. Purvis, Thomas. 1995. Revolutionary America, 1763 to 1800. New York: Facts on File. Rabe-Hesketh, Sophia and Anders Skrondal. 2008. Multilevel and Longitudinal Modeling Using Stata, 2nd ed. College Station, TX: Stata Press. Rich, Wesley Everett. 1924. The History of the United States Post Office to the Year 1829. Cambridge, MA: Harvard University Press. Scaff, Lawrence A. 2011. Max Weber in America. Princeton, NJ: Princeton University Press. Schaffer, Mark E. 2007. “XTIVREG2: STATA Module to Perform Extended IV/2SLS, GMM and AC/ HAC, LIML and K-class Regression for Panel Data Models.” Retrieved November 10, 2009 (https://ideas .repec.org/c/boc/bocode/s456501.html). Skocpol, Theda, Marshall Ganz, and Ziad Munson. 2000. “A Nation of Organizers.” American Political Science Review 94:527–46. Smith, Adam. 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. London: Strahan and Caddell. Smith, Christian. 2008. “New Directions in the Sociology of Religion.” Social Forces 86:1561–89. Smith, Elwyn A. 1962. “The Forming of a Modern American Denomination.” Church History 31:74–99. Smith, Timothy. 1957. Revivalism and Social Reform in Mid-Nineteenth Century America. Nashville, TN: Abington. The Spirit of the Pilgrims. 1828. “Front Matter.” Vol. 1, p.1 (retrieved from American Periodicals Service Online on September 5, 2006). Stark, Rodney. 1998. “Catholic Contexts: Competition, Commitment, and Innovation.” Review of Religious Research 39:197–208. Stark, Rodney and Roger Finke. 2000. Acts of Faith. Berkeley: University of California Press. Stark, Rodney and James C. McCann. 1993. “Market Forces and Catholic Commitment: Exploring the New Paradigm.” Journal for the Scientific Study of Religion 32:111–24. Goldstein and Haveman Starr, Paul. 2004. The Creation of the Media. New York: Basic Books. Stout, Harry and Scott Cormode. 1998. “Institutions and the Story of American Religion.” Pp. 62–78 in Sacred Canopies, edited by N. J. Demerath, P. Dobkin Hall, T. Schmitt, and R. H. Williams. New York: Oxford University Press. Sutton, John R. and Mark Chaves. 2004. “Explaining Schisms in American Protestant Denominations, 1890–1990.” Journal for the Scientific Study of Religion 43:171–90. Tebbel, John W. and Mary E. Zuckerman. 1991. The Magazine in America. New York: Oxford University Press. Thomas, Isaiah. 1874. The History of Printing in America, 2nd ed. New York: Weathervane Books. Thompson, John. 1995. The Media and Modernity. Stanford, CA: Stanford University Press. U.S. Census Bureau. 1998. Population of the 100 Largest Cities and Other Urban Places in the United States. Washington, DC: U.S. Census Bureau. U.S. Census Bureau. 2006. Historical Statistics of the United States, Earliest Times to the Present: Millennial Edition Online. Washington, DC and New York: U.S. Census Bureau and Cambridge University Press. Voas, David, Alasdair Crockett, and Daniel Olson. 2002. “Religious Pluralism and Participation: Why Previous Research Is Wrong.” American Sociological Review 67:212–30. The Watchman. 1819. “Preface to the Watchman.” Vol. 1, p.1 (retrieved from American Periodicals Service Online on April 17, 2006). Wilde, Melissa J. 2007. Vatican II: A Sociological Analysis of Religious Change. Princeton, NJ: Princeton University Press. Wilde, Melissa J., Kristin Geraty, Shelley L. Nelson, and Emily A. Bowman. 2010. “Religious Economy or 827 Organizational Field? Predicting Bishops’ Votes at the Second Vatican Council.” American Sociological Review 75:586–606. Williams, Peter W. 1998. America’s Religions: From Their Origins to the Twenty-first Century. Urbana: University of Illinois Press. Wright, Conrad. 1984. “The Growth of Denominational Bureaucracies.” Harvard Theological Review 77:177–94. Young, Michael. 2006. Bearing Witness against Sin. Chicago: University of Chicago Press. Zald, Meyer N. and Bert Useem. 1987. “Movement and Countermovement Interaction: Mobilization, Tactics, and State Involvement.” Pp. 247–72 in Social Movements in an Organizational Society, edited by M. Zald and J. McCarthy. New Brunswick, NJ: Transaction. Adam Goldstein is a doctoral candidate in the Department of Sociology at the University of CaliforniaBerkeley. His research focuses primarily on the economic sociology of financial capitalism in the contemporary United States. Heather A. Haveman is Professor of Sociology and Business at the University of California-Berkeley. She received a BA in history and an MBA from the University of Toronto, and a PhD in organizational behavior and industrial relations from UC-Berkeley. She studies how organizations, industries, and employees’ careers evolve. Her work has appeared in Administrative Science Quarterly, American Sociological Review, American Journal of Sociology, Poetics, Organization Science, Journal of Business Venturing, and Academy of Management Journal. Her current research involves the evolution of antebellum American magazines, and corporate governance in twentyfirst-century Chinese corporations.