The formal rules of organised religion: a framework for empirical research

ABSTRACT

The ideal length of formal rules has been studied as a core preoccupation of firms and states. Shorter rules are a typical firm’s response to performance pressures concerning efficiency; longer rules are a typical polity’s solution to questions of political control and order. Very little is known in this respect about the rules of an institution that has been longer-lived and more influential than most firms and states: organised religion. Are the drafters of church rules more sensitive to performance pressures or to political considerations or to both? This article brings together theories of constitutional politics, church and state, bureaucracy, and economic competition to develop explanations of length variation in the core rules of churches. An empirical exploration proposes ways to test these expectations and produces relevant preliminary evidence. This new direction in the study of institutional religion can update existing understandings of churches as complex institutions that lie somewhere between the ideal-typical firm and the ideal-typical polity.

KEYWORDS:

• Church governance
• statutes
• constitutions
• bureaucracy

Introduction

The length of formal rules is an old preoccupation of human collectivities. This is implied by the vertical monumental depiction of authoritative texts since ancient times and by various contemporary metrics, such as the page count of a country’s tax code and the number of regulatory barriers to starting a business. Scholarly research on the length of formal rules has focused predominantly on rule-making in the context of firms and states (Hall, Haas, and Johnson 1967a; Tsebelis and Nardi 2016; cf. March, Schulz, and Zhou 2000). In the corporate domain, where rules reflect an organisation’s response to “pressures of scale, complexity, markets, resource flows [and] environmental uncertainty” (Hinings and Raynard 2014, 166), longer rules may hinder smooth, lean operations; they can be an obstacle to efficient decision-making. In government, where rules reflect “a bargain among political interests” (Cooter 2000, 6), shorter rules may be too ambiguous and conducive to disorder; they can leave unresolved problems of control over the governing office.

There is little systematic theorising and empirical work on the length of the core rules of an institution that has been longer-lived and more influential than most states and firms: organised religion. This is all the more surprising when considering the enduring scholarly fascination with questions of authority and organisation within churches, questions that are settled to a great extent by written rules. Max Weber is the natural point of departure for any attempt to understand the internal workings of churches as collective entities (e.g. Weber 1978, 54, 251–252, 1164). His work combines insights into the legitimacy of authority within religious communities (politics and conflict) and the historical evolution of these communities towards bureaucratic organisation (administration and efficiency). These two related themes are the basis of the explanatory framework that appears in the following regarding the internal rules governing the operation of modern churches.

The framework adds an overlooked meso-level perspective to the considerable amount of micro- and macro-level research on religion as a politically, socially, and economically relevant phenomenon. The analysis aims to understand the rule-making strategies of organised religions by focusing on the extensiveness of their core internal rules (‘church statutes’ or ‘church constitutions’). Do churches design rules that are adequately long to address conflicting political considerations? Do they design rules that are adequately short to ensure operational agility? To examine these questions, the present study approaches churches from two analytical perspectives. On the one hand, it considers them as organised collectivities reacting to internal and external political pressures regarding the control of institutional resources. On the other hand, it examines them as office hierarchies reacting to performance pressures concerning organisational efficiency. The theoretical contribution of the article is supplemented by primary empirical information from an availability sample of statutes that regulate the internal operation of Christian churches—Eastern Orthodox and Protestant (N = 28).

The novel research direction that the article introduces contributes to the comparative study of institutional religion as an organisational phenomenon—as a “regularly organized and permanent enterprise” (Weber 1978, 425). It adds an ignored aspect of strategic behaviour to the empirical literature on religion and politics: rule-making within clerical hierarchies. It produces indicators of the extensiveness of written rules in the ecclesiastical domain, connecting them to the study of bureaucratic formalisation. Finally, it takes the measurement of ‘institutions’, especially national constitutions, in a new direction: the study of church statutes.

Politics and conflict

Classic notions of the church as societas and civitas are familiar descriptions of the assembly of the faithful (Florovsky 1959, 238). The concept of the church as a regulated political community draws on scriptural references, where the faithful are described as a “holy nation”, a “people”, and an “Assembly”, acceptance of the Christian faith as heavenly “citizenship”, and the church as a city/polity (polis) and a “colony of Heaven” (Florovsky 1959, 233–234). These notions frame the church phenomenon “as a political institution, a body of people governed by constituted authority” (Donahue 1972, 294). The treatment of the church as a political assembly (ecclesia) places emphasis on the distribution and exercise of power across different constituencies and factions (Chaves 1993, 18; Wittman 2014, 725–726).

This perspective casts church statutes as the written instruments that map the entrenched fundamental rules regulating a polity, an established empirical question in political science (Huber and Shipan 2002; Colomer 2006; Elkins, Ginsburg, and Melton 2009; Voigt 2009). From this angle, church statutes, often called ‘constitutions’ by their drafters, are similar to national constitutional documents. They, too, contain “a set of rules for making collective decisions” (Colomer 2006, 217) about the central problem of any association: the coordination of shared life (Weber 1978, 50). They are the defining and binding texts that specify the central structures, decision-making processes, and bylaws of a community.

Alex Rutherford et al. provide a concise definition of the functions served by national constitutions, which overlap with the functions of church statutes:

Constitutions describe the fundamental principles by which the state will be governed, the political and legal state institutions, the powers, procedures, and duties of those institutions, and the rights and responsibilities of individuals. (Rutherford et al. 2018, 592)

Table 1 presents examples of the comparable functions covered by the provisions of national constitutions and church statutes.

Table 1. Constitutionalism.

Country	Features	State constitution		Orthodox Church statute
Romania	Identity	“Romania is a national state … ” (1)		“The Romanian Orthodox Church is the community of … ” (1)
Romania	Enforcement	“ … observance of the Constitution … shall be mandatory” (1.5)		“ … compulsory for the entire Romanian Orthodox Church” (201)
Romania	Symbols	“Romania's flag is tricolor; it consists of three … ” (12.1)		“ … has its own flag described as follows … ” (195)
Romania	Government form	“The form of government … is the Republic” (1.2)		“ … is administrated … through her own representative bodies” (3.2)
Russia	Head selection	“ … not younger than 35 years of age” (81.2)		“ … not younger than 40 years old” (IV.15.e)
Russia	Power separation	“State power … shall be exercised by the President … the Federal Assembly … the Government … and the courts” (11.1)		“The supreme bodies … shall be the Local Council, the Bishops’ Council and the Holy Synod … ” (I.7)
Russia	Judiciary	“Justice in the Russian Federation shall be administered only by court” (118.1)		“Judicial authority … shall be exercised by the Ecclesiastical Courts” (IX.1)
Russia	Amendments	“ … may be submitted by the President of the Russian Federation, the Council of Federation … ” (134)		“The right to introduce amendments to the present Statute shall belong to the Bishops’ Council” (XXIII.3)
Serbia	Sovereignty	“No state body … may usurp the sovereignty from the citizens” (2)		“ … it shall govern … ecclesiastical affairs autonomously” (1)
Serbia	Indivisibility	“The territory … is inseparable and indivisible” (8)		“ … is one and inseparable autocephalous church” (1)

Source: Official translations of current state constitutions from the “Constitute” project web site and of current statutes from church web sites. Section/article number in parentheses. Original spelling preserved.

The church is a political battleground where different internal and external constituencies compete for the control of tangible and symbolic resources, for example, by influencing the rules of the game. The church experiences similar problems to those faced by a typical organised collectivity (March and Simon 1958, ch. 5). A central aspect of the question of control over institutional assets is the canonical and established status of a church. Some of the religious institutional actors that are discussed in this analysis operate under the jurisdiction of mother churches (particularly of the Russian Orthodox Church, which is the largest in the world). Others are or used to be formally established or state churches, such as the (Orthodox) Church of Greece and, until 2000, the (Protestant) Church of Sweden. Incidentally, although the similarity between church and polity allows to borrow insights from empirical research on national constitutions, the church is not necessarily equivalent to a sovereign state. It often finds itself subordinate to more powerful secular or religious actors. The fact that the senior clergy of many churches are vested with a combination of executive and other powers further qualifies the analogy between church and polity.

From a principal-agent perspective, which has been a useful contribution to the study of national constitutions, the mother church or host state (external principals) owns a valuable resource and recruits the clerical hierarchy (internal agents) to manage it. The resource can refer to adherents, religious buildings, donations, and a symbolic religious brand with considerable societal influence. Since agents are closer to the management of the resource, they can use it against principals’ preferences. To guard against this situation, principals can place extensive and restrictive rules on the way agents manage the resource as a type of “pre-commitment device” (Voigt 2009, 291) and “constitutional micromanagement” (Versteeg and Zackin 2016, 658).

In the special case of a state church, the state at some point in history had incorporated the church as a branch of the government bureaucracy, as in the example of the Russian Orthodox Church under Peter the Great (Buss 1989, 248). Such churches typically have their statutes approved by secular authorities (e.g. the Church of Greece today). Here, the state uses the church to serve purposes other than strictly spiritual and pastoral ones (Sarkissian 2015, 90). There is a great degree of state interference in ecclesiastical affairs in this case, including meddling with core institutional rules. Although this top-down meddling can be composition-neutral (interference can be about adding or deleting regulations), the principal-agent framework anticipates that the principal (state or mother church) will guard against mismanagement by its agent (clerical élites) by imposing heavier regulations on the church it controls. This discussion is summarised in the following hypothesis:

H1: Non-autonomous churches (dependent on another entity such as a mother church or host state) will have more extensive core rules.

A different theoretical perspective expects the opposite effect. One strand of the empirical constitutionalism literature in political science suggests that dependence on a more powerful actor might lead to shorter regulatory texts. Specifically, previous research finds that stronger top-down control over the composition of national constitutions, as in the case of autocratic regimes and feeble assemblies, creates weaker incentives for bargaining and compromise (Voigt 2009, 299). Compromise, bargaining, and the elaborate synthesis of diverse interests come with independent deliberation and tend to generate longer constitutional documents. These are uncommon under stronger top-down control (ibid, 292). Following this line of reasoning, a church that lacks independence by being formally attached to a more powerful entity (a mother church or host state) will operate by rules that do not need to reflect compromise across competing factions, but refer instead to the narrow interests of the more powerful entity. This church’s internal rules are bound to be short.

The Belarusian Orthodox Church is a case in point. This organisation is not independent, as it is formally a part (an ‘exarchate’) of the Russian Orthodox Church. It has the shortest statute examined by the present analysis. The charter of the Belarusian church contains references that work as a space-saving device in terms of textual composition. Specifically, Article 3 states that, in addition to its own charter, this church is regulated primarily by rules that appear in the statute of the Russian Orthodox Church. In this ‘autocratic’ context, the Belarusian charter can afford to be short by relying partly on another church’s regulations. From this discussion, this hypothesis follows:

H2: Non-autonomous churches (dependent on another entity such as a mother church or host state) will have less extensive core rules.

Independence is just one type of political consideration, concerning the ‘external principals vs. internal agents’ relationship (controlling entity vs. clerical hierarchy) discussed so far. An alternative focus within the same perspective draws on the ‘internal principals vs. internal agents’ relationship as a source of potential conflict: the laity vs. the clerical hierarchy (Moberg 1962, 267–268; Fama and Jensen 1983, 320; Patrikios 2020, 420–421). In a stylised depiction, Orthodox churches are governed as episcopal polities, where senior clergy are the dominant decision-makers. Power tends to be more diffuse in Protestant churches (although these are a more heterogeneous group), which encourage greater lay involvement in top-level ecclesiastical affairs. These differences should influence the length of church rules in predictable ways. More diffuse, less centralised decision-making of the type encountered in many Protestant ecclesiastical polities requires more rules to maintain order (Wollschleger 2013, 483). This expectation is formalised as follows:

H3: Protestant churches (compared to Orthodox ones) will have more extensive core rules.

In a line of further enquiry, church corruption should affect the length of core rules. The political calculus that connects societal corruption and longer national constitutions (Bjørnskov and Voigt 2014, 93; Tsebelis and Nardi 2016, 472) anticipates a similar phenomenon in the church domain. Specifically, corrupt drafters tend to design longer regulations in order to entrench vested interests. This is a device that aims to discourage future revisions against those interests. An alternative and more optimistic explanation of the same positive link between corruption levels and the length of rules expects that, faced with corruption, non-corrupt drafters will design longer regulations in order to protect their polity from vested interests. The relationship is subject to endogeneity problems (corruption as both a driver and an outcome of regulatory length). The main expectation is formalised as follows:

H4: Churches with greater internal corruption problems will have more extensive core rules.

Efficiency and performance

The modern church is not merely a political domain that encompasses the negotiation and pursuit of collective goals and the resolution of tensions surrounding these goals. In its more quotidian form, and following a general societal trend towards rationalisation, it has evolved over the centuries into an enterprise (Weber 1978, 223, 426). It relies on expert and support personnel to fund, coordinate, and deliver core and ancillary services, such as weekly worship, Sunday school, estates management, news media, and a printing house (Chaves 1993, 9). It has a differentiated structure, continuous operations, and standardised, written procedures. The treatment of the church phenomenon from an administrative perspective is a popular theme in sociological studies (see overviews in Beckford 1973; Hinings and Raynard 2014), which do not ignore, however, the risks that rationalisation poses to the church’s core identity as a spiritual institution (Thung 1976, 322; Thompson 1970, 226).

To illustrate what this perspective entails more concretely and justify further the discussion of performance questions that may arise within this institution, Table 2 lists various departments that compose a church and an insurance firm, respectively. The comparison is not entirely arbitrary in terms of functional equivalence, since economic theories already treat religion as a type of insurance that promises compensation in the hereafter—in the event of death (Stark and Bainbridge 1987). Table 2 suggests that the two organisational structures share a range of units with similar functions, including the offices that are directly involved in top-level management (sections 1–6 in the church column and sections 1–5 in the firm column), IT and digital departments, internal assessment, relations with external constituencies, public affairs and communications, compliance and finance. In other words, “[i]t is the same Church which administers finances and sacraments” (Brodd 2009, 326).

Table 2. Organisational structure: departments (selected).

Greek Orthodox Archdiocese of America	Generali Insurance Company
1. Archdiocese Administration	1. Board of Directors
2. Archdiocesan Council	2. Office of the CEO
3. Office of Internal Assessment & Evaluation	3. Office of the General Manager
4. Office of the Chancellor	4. Corporate Affairs
5. Office of Administration	5. Group Audit
6. Office of the Archbishop	6. Risk Office
7. Archives & Resource Center	7. HR & Organisation
8. Communications	8. Digital Office
9. Finance	9. Marketing & Customer Office
10. Telecommunications	10. Finance Office
11. Internet Ministries	11. Compliance Office
12. Office of Development	12. Communications & Public Affairs
13. Outreach & Evangelism	13. Life & Health Insurance
14. Press Office	14. Investor & Rating Agency Relations
15. Public Affairs	15. Strategy and Business Accelerator

Source: Official organisation charts.

Against this background, the statutes that govern most modern churches are not simple value statements. Instead, they read as the regulations of a Weberian bureaucracy, in which

rules are formulated and recorded in writing … The combination of written documents and a continuous operation by officials constitutes the ‘office’ (Bureau) which is the central focus of all types of modern organized action. (Weber 1978, 219)

Church statutes contain minute provisions that regulate each church’s complex operations. As an illustration, there are 13 paragraphs in the statute of the Orthodox Church of Cyprus (Article 47) that specify with fastidious care the life of parish priests. In addition to hiring requirements such as marital status, educational qualifications, physical health, and a clean criminal record, the rules set out the details of the priest’s required place of residence during tenure, the transfer process to another parish, his line manager (this is a male-only corps), annual leave entitlements, conditions of removal, and superannuation. Similarly, 20 paragraphs in the statute of the Romanian Orthodox Church (Article 26) are dedicated to the description of its leader’s role. Leadership functions include acting as convenor of deliberative bodies within the church, ritual duties, delegation responsibilities to external bodies, oversight of disciplinary procedures and ordainment, and other executive, supervisory, and advisory roles.

The study of industrial and state bureaucracies recognizes the degree of reliance on and extensiveness of written rules, or formalisation, as a central feature of complex organisations (March and Simon 1958, 62–65; Glaeser et al. 2004, 275; see the related ‘formalism’ concept in Djankov et al. 2003, 455). The typical empirical study of formalisation investigates how heavily an organisation’s administration draws on written regulations, records, and communications (Walton 2005, 574). Specific indicators deal with broad features, such as the degree to which organisational policies are written, and with more routine aspects of workers’ behaviour and roles, such as the existence and extent of written job descriptions (Hall, Haas, and Johnson 1967a, 1967b; Pennings 1973, 689; Atwater 1995, 463).

Often examined in the for-profit context, formalisation is a mechanism of programmed coordination and uniformity (March and Simon 1958, 2). Its intended function is to regulate an organisation in an impersonal way that maximises efficiency, although in practice it often inhibits reform and innovation (Hetherington 1991, 133–134). For empirical purposes, formalisation is defined here as the length and number of internal institutional regulations. These two aspects of textual composition serve as a concrete observable indicator that ensures reliable measurement and have been used widely in the study of corporate bureaucracies.

Formalisation is primarily a function of size (Pugh et al. 1963, 309). An institution that expands by adding personnel, stakeholders, divisions, and hierarchical levels finds it harder to achieve control, coordination, calculability of expected results, and consensus. To ensure that the same procedures are followed invariably, large organisations are likely to rely on detailed written rules rather than on direct personal supervision from the top (Walton 2005, 574). When thinking about size in the church domain, the number of members or adherents is a widely used statistic. The presence of more members and the accompanying need to serve a larger area with more personnel using more resources reflect the type of institutional strain described here. This discussion is summarised in the following hypothesis:

H5: Larger churches will have more extensive core rules.

The bureaucratic formalisation lens is not the only one that applies. A parallel strand within the performance and efficiency theme connects the composition of written rules with the extent of competition that a church faces from other churches. From an economic theoretical viewpoint, competition can lead firms to revise their organisational structures, which includes “rewriting their corporate charter” (Ekelund, Hebert, and Tollison 2004, 703). Greater competition from similar organisations and the prospect of a drop in market share can push firms to adopt a more agile structure.

To elaborate further, there is the case of a monopolistic firm. Facing weak competition and minimal threats to its dominant market status, this firm can afford to survive on more rigid internal structures and operations. Rigidity entails, for instance, not adapting timely to changing consumer preferences. The evolution of a complicated and lengthy internal corpus of rules is a prime example of such rigidity, the product of bureaucratic ‘inertia’: rules becoming ends, which lead to more rules, rather than means to efficient outcomes (Bozeman and Scott 1996, 3). The concept of ‘bad red tape’ has been used in this respect to describe the obstacles placed on organisational efficiency by “endless paperwork” and “excessive, duplicate, and unnecessary procedures” (ibid, 2).

Then there is the opposite case of a firm that is threatened by new market entrants (for instance, when the state breaks up a monopoly). This firm is expected to react by becoming less rigid in a drive to maintain a viable market share. The reaction includes a revision of the complexity and extensiveness of the firm’s core rules; simpler and fewer internal regulations provide greater operational flexibility to a firm that plans to compete successfully. As an illustration of this logic, the use of shorter and fewer rules can expedite the process by which a firm plans for, staffs, and opens a new office or closes and relocates (or liquidates) the assets of an existing one. Empirical evidence from organisational research supports this expectation. Organisations that operate in more competitive environments, namely, business firms, tend to rely less heavily on written rules, policies, and job descriptions compared to organisations in less competitive settings, such as public service or mutual benefit associations (Hall, Haas, and Johnson 1967b, 129−130).

Focusing explicitly on the religious domain, the economics of religion approach (Stark and Bainbridge 1987; Iannaccone 1991) already treats churches as firms and theorises that a church that competes for the same stake against other churches can succeed only as a more efficient and flexible organisation. From this rational choice perspective, it follows that being unbound by too many internal constraints and being responsive to the fickle nature of market pressures are ideal characteristics that should be reflected in the length of core church rules. The opposite outcome, as in heavy formalisation (excessive rules), should prevail among churches that do not have to compete with other churches for resources and members and can afford to have slower (more complex/numerous) decision-making procedures. Therefore, the degree of exposure to competition should be inversely related to the extensiveness of church statutes. This discussion is summarised in the following hypothesis:

H6: Churches under stronger competition pressure will have less extensive core rules.

A different theoretical framework expects the opposite sign in the above relationship (Pfeffer and Leblebici 1973, 278; Vogel 1996). An organisation that competes with other organisations for resources is exposed to a greater degree of uncertainty. Greater competition, therefore, brings greater uncertainty and, by extension, a stronger appetite for control within the organisation. Extensive written rules are one of the basic instruments for achieving control and minimising uncertainty by increasing the predictability of organisational efforts. On the contrary, an organisation that is dominant in its field and does not need to compete faces greater certainty and, therefore, a lower need for controlling internal measures such as extensive rules.

A related pressure that is particular to the church domain is competition not from other churches, but from non-religion. The secularisation of culture and institutional differentiation in advanced modernity expose the religious institution to greater uncertainty regarding its future prospects. A general relationship between secularisation and the marginalisation of organised religion, on the one hand, and further bureaucratisation, on the other hand, have been documented historically in the Protestant European context (Thompson 1970). In summary, this alternative hypothesis views competition as positively associated with the degree of formalisation in church statutes:

H7: Churches under stronger competition pressure will have more extensive core rules.

Government intervention, which was discussed in the previous section with specific reference to state churches, features as a broader phenomenon in the economics of religion framework. All churches, business firms, and other non-governmental organisations that operate in a particular country exist under some degree of state supervision. The regulatory weight of the state lies in its ability to facilitate or hinder organisational activity (Iannaccone 1991, 160). There are two stylised extremes in this conceptualisation of the state’s role. One is interventionism, whereby the state monitors closely denominational activity in its jurisdiction. The other extreme is a state with a minarchist role. In this conceptualisation, greater state intervention into religious affairs—conducted typically through excessive legislation—leads to more rules that need to be navigated by religious institutions in their operations, with knock-on effects for the composition of the internal regulations of churches.

As an illustration, the statute of the Russian Orthodox Church (official English translation) contains more than 30 references to ‘government’ and other state-related features that include state registration, state authority, state laws, state territory, and church–state relations. The statute of the Romanian Orthodox Church (official English translation) contains more than 20 separate references to state-related issues. Therefore, the extent to which a government intervenes to restrict the activity of organisations operating in its jurisdiction should be reflected in churches adopting more extensive internal rules, if they want to address government restrictions and avoid attracting unwanted state attention.

A similar effect is anticipated from the theoretical perspective of institutional isomorphism (DiMaggio and Powell 1983, 149; Zucker 1983, 10). This reading diverges from the more explicitly interest-driven considerations described in the previous two paragraphs. Here, the presence of an interventionist state, which typically employs an extensive regulatory corpus to manage organisational activity in its jurisdiction, exerts additional pressure on the social actors that it monitors (churches, NGOs, firms), which is more cultural in nature. These actors are expected to adopt extensive regulatory corpora for their own internal workings because they view lengthy rules as a type of legitimate organisational arrangement in the eyes of the state. Put differently, having more rules is socially defined as ‘appropriate’ in this context. This is an instance of normative organisational behaviour in pursuit of social fitness in the wider environment. This discussion is summarised in the following hypothesis:

H8: Churches that operate in a more heavily regulated environment will have more extensive core rules.

An empirical framework

The empirical framework proposed in the following exploratory analysis is based on the collection of systematic information for most of the conceptual variables discussed above. Regarding the dependent variable, the collection of the documents containing the core rules of each church took place between October 2018 and January 2019 and was based on digital availability. Statutes used in the analysis were valid at the time of collection. Only legal documents that had the following terms in their title were included: statute(s), constitution, charter, ordinance. Table 3 provides the list of churches in the analysis, which should be treated as an availability sample and not as representative of the universe of Eastern Orthodox and Protestant churches operating in Europe or elsewhere (see details in the Appendix, Table A1).

Table 3. Cases.

Church (country, year in which document signed/revised)
Orthodox:	Serbian Orthodox Ch. (Serbia, 1957)
Belarusian Orthodox Ch. (Belarus, 1990)	Romanian Orthodox Ch. (Romania, 2007)
Greek Orthodox Archdiocese of America (USA, 2003)	Polish Orthodox Ch. (Poland, 1995)
Russian Orthodox Ch. / Moscow Patriarchate (Russia, 2017)	Bulgarian Orthodox Ch. (Bulgaria, 2008)
Estonian Orthodox Ch., Moscow Patriarchate (Estonia, 2012)	Protestant:
Orthodox Ch. of the Czech Lands & Slovakia (Slovakia, 1999)	Evangelical Ch., Augsburg Confession (Austria, 2018)
Moldovan Orthodox Ch., Moscow Patriarchate (Moldova, 1993)	United Protestant Ch. (Belgium, 2016)
Orthodox Ch. of Albania (Albania, 2006)	Evangelical Ch. in Croatia (Croatia, 2012)
Montenegrin Orthodox Ch., CPC (Montenegro, 2009)	United Protestant Ch. of France (France, 2014)
Orthodox Ch. of Finland (Finland, 2006)	Reformed Ch. in Hungary (Hungary, 2016)
Georgian Apostolic Autocephalous Orthodox Ch. (Georgia, 1995)	Ch. of Ireland (Ireland, 2003)
Ch. of Cyprus (Cyprus, 2010)	Evangelical Lutheran Ch. in Italy (Italy, 2004)
Ukrainian Orthodox Ch., Kiev Patriarchate (Ukraine, 2016)	Evangelical Lutheran Ch. of Latvia (Latvia, 2016)
Ch. of Greece (Greece, 2018)	Protestant Ch. in the Netherlands (Netherlands, 2016)
Macedonian Orthodox Ch., Ohrid Archbishopric (N. Macedonia, 1994)	Ch. of Sweden (Sweden, 2019)

The dependent variable, the extensiveness of textual composition, is measured in three ways, following the literature on constitutional metrics (Bjørnskov and Voigt 2014, 98). The first indicator captures length—how many words are used in the text of each church’s statute. Because of limitations inherent in comparing texts written in different languages (Voigt 2009, 295), the raw number of words was not used. Instead, I controlled for the different efficiency of each language in terms of word use by applying a weight reflecting language efficiency regarding English. The weight uses comparisons of English and non-English versions of the UN Declaration of Human Rights, following the procedure described by Robert Cooter and Tom Ginsburg (2003, 3). Although the weighted number and the raw number of words are highly correlated, I used the former for comparability.

The second indicator is an approximate calculation of the number of sentences in the original text of each church’s statute. This is measured as the number of full stops found in each document. In certain cases, however, full stops in a text may denote merely multi-level section numbering (e.g. Section 2.3.4.a.). This example contains four full stops and would, therefore, count as four ‘sentences’ under my coding protocol. Although such multi-level lists are a prime instance of pedantry and formality, which are somewhat related to the underlying variable of interest, this indicator is obviously only an impressionistic measure of length.

The third indicator is the number of rules, measured as the number of articles or specific provisions in the text. Each provision is usually denoted by name (‘article’) and is numbered in the statutes either sequentially from beginning to end or within each chapter/section. When this marker is absent, I coded the first hierarchical level in the text that appears below chapter/section level. Depending on the case, the length of a single article can range from a single dependent clause to a number of pages, although the latter is an uncommon occurrence.

The calculation of the three indicators uses the original language of the text with one exception. The constitution of the Macedonian Orthodox Church was only available in English, but the ‘UN’ weight shows a very close correspondence in length between the two languages (1,778 words in English vs. 1,771 words in the native language for the same UN text). Therefore, the length of the statute in English is used for this case. Preambles, tables of contents, historical introductions, executive summaries, separate ordinances, and appendices were excluded from each text (unless they appear as part of a numbered article). The aim of the coding procedure was for each text to begin with its title, any references to dates and drafting version if available, and then the first article or chapter of the statute. Footnotes with clarifications, notes on revisions, and references that appear in the main text of each statute were retained.

These three indicators measure slightly different aspects of the extensiveness of textual composition, the underlying dependent variable. To produce a more reliable indicator, I combined the three scores into a composite scale. Figure 1 arranges churches along this composite length variable, which reveals considerable variation in statute length across churches (see also the raw data in the Appendix, Table A2). The scores of composite length are regression-based factor values using principal component analysis. This calculation of composite length uses logs of the three textual variables, which produce a single factor solution (variance explained = 85.99%; Cronbach’s α=0.91). The use of natural logarithms for the three raw variables was necessary because distances in low scores, such as between 1,000 and 2,000 words, are considered more important (double) than the same distances in high scores, such as between 20,000 and 21,000 words (see also Tsebelis and Nardi 2016, 463).

Figure 1. Statute length (composite).

The independent variables for which information was available capture characteristics of each church or country in which that church is based. Although some churches operate across national borders (e.g. the Czech Republic and Slovakia), country features used in this analysis refer to the country in which the church is mainly based (Slovakia). To measure church organisational independence (H1–2), I categorised churches into either of two groups: dependent (0) or independent (1). Dependent cases are not canonically sovereign (eparchies and chapters of a larger church, representing the larger church in a different country) or they are state/national churches (the Orthodox Church of Greece, and of Finland). The source of this information comes from International Religious Freedom Reports (United States Department of State 2017) and the Religion and State project (Round 3 data, see Fox 2015). This is a church-level descriptor.

The type of ecclesiastical polity (H3) is a dichotomous variable reflecting whether a church is Protestant (0) or Eastern Orthodox (1). Unfortunately, data availability prevented me from testing the corruption hypothesis (H4). The only comparative measure of corruption in religious bodies presently available, the Global Corruption Barometer by Transparency International, gauges mass perceptions of corruption for some of the countries examined in this article. These perceptions are not organisation- (church-)specific, but refer to all the religious bodies active in a given country. This feature makes the indicator invalid as a measure of church-level corruption.

A church’s absolute size in membership terms (H5) and the degree of denominational competition (H6–7) were measured using the same sources of population data. The competition indicator gauges a church’s exposure to pressures from other churches and from non-religion. This is a dichotomous indicator of market position, which reflects whether a church holds majority status among the host country’s population—whether its adherents are more than 50% of the population. Indirectly, majority status also reveals important information about the heterogeneity of the religious market, with the presence of a majority church implying a more homogeneous (and less secular), hence less competitive, religious economy. These population statistics come from the most recent entry provided by the World Religion Project (WRP, as quoted in Maoz and Henderson 2013).

This commonly used source is problematic for my purposes. Specifically, WRP data are country-level estimates of adherents by faith tradition rather than by specific church. This is not ideal for countries where more than one church of the same faith tradition operates (especially Estonia, Montenegro, Ukraine, and the United States). To correct this limitation of the WRP data, information for the selected churches in Estonia, Montenegro, and Ukraine is based on recent International Religious Freedom Reports (United States Department of State 2017). Church adherent estimates for the US case, the Greek Orthodox Archdiocese of America, come from the Church’s own web site (www.goarch.org/about, accessed 11 January 2019).

To measure the extent of government regulation (H8) I relied on the International Religious Freedom data programme (Grim and Finke 2006, aggregate file 2003–2008). The programme codifies the International Religious Freedom Reports produced by the US Department of State, covering almost 200 countries and territories. The reports reflect evaluations by US government experts regarding a state’s stance towards institutional religion. The International Religious Freedom dataset that quantifies these reports has been used extensively as a source of comparative information on religious regulation (e.g. Buckley and Mantilla 2013). The government regulation index (GRI) that was used here reflects the sections of the reports that discuss government-imposed restrictions on religious activity such as missionary activity, proselytising, and public expressions of religiosity. The index captures the degree of government intervention in denominational operations. High scores show heavier regulation by the state. An alternative indicator was also used, which measures the overall degree of interventionism by the state in society (not simply in religious affairs). This relied on the logged number of words in the host country’s constitution (2013 or most recent measurement, Comparative Constitutions Project 2014). Table 4 presents descriptive statistics for all the variables in the analysis.

Table 4. Summary statistics.

	Mean	Median	Std Dev	Min	Max
Length	0.00	0.06	1.00	−1.54	2.21
Independent	0.79	1.00	0.42	0.00	1.00
Orthodox	0.64	1.00	0.49	0.00	1.00
Size(lg)	13.94	14.05	1.97	9.62	18.45
Majority	0.36	0.00	0.49	0.00	1.00
Regulation	2.53	1.53	2.32	0.00	7.69
Constitution(lg)	9.57	9.51	0.43	8.50	10.63

N = 28

Table 5 reports case scores for the five churches with the longest statutes (top quintile) and the five churches with the shortest statutes (bottom quintile) in the sample. The comparison of scores between the two groups reveals at least three patterns that are theoretically meaningful. Firstly, churches that are formally independent religious actors tend to have longer statutes. These actors are free to deliberate on their own without external intervention, a situation that usually requires lengthier rules in order to solve governance problems by compromise and bargaining. Secondly, churches that are not majority religious actors tend to have shorter statutes. These actors face relatively greater pressures from competing denominations or from secularism, compared to majority religions that face weaker pressures of this kind. Thirdly, churches that operate in countries with shorter national constitutions appear to have shorter rules. These actors, along with other non-governmental organisations, have to address fewer governmental regulations when drafting their own statutes and may also model the length of their rules on a normative benchmark: the shorter length of the national constitution.

Table 5. A comparison of selected churches.

Church in:	Independent	Size(lg)	Majority religion	Regulation	Constitution(lg)	Statute length
Sweden	yes	15.67	yes	0.28	9.52	2.21
Ireland	yes	11.81	no	0.00	9.68	2.08
Austria	yes	13.13	no	1.11	10.63	1.28
Bulgaria^#	yes	15.62	yes	6.67	9.52	1.06
Russia^#	yes	18.45	yes	5.19	9.47	0.89
…	…	…	…	…	…	…
Estonia^#	no	12.04	no	0.00	9.34	−1.00
USA^#	no	14.22	no	0.00	8.96	−1.14
Croatia	yes	9.62	no	0.93	9.30	−1.16
Netherlands	yes	14.85	no	0.00	9.08	−1.34
Belarus^#	no	15.34	no	7.68	9.49	−1.54

Note: The table refers to the five longest (+) and the five shortest (−) statutes, respectively.

# Tradition = Orthodox

Using all 28 cases in the sample, Table 6 presents bivariate correlation coefficients (Pearson’s r) for the relationship between composite length and each of the explanatory variables for which measures were available. Bivariate correlations were used in the main text due to the small sample size, which renders it problematic to conduct multivariate statistical analysis (see subject/variable ratio problem, Miller and Kunce 1973, 160). According to Table 6, longer rules are more prevalent among independent churches (H2), larger churches (H5), majority churches (H6), and churches hosted by countries with interventionist states (states with longer constitutional documents) (H8). Significant correlations persist after applying basic country controls (see regressions in the Appendix, Table A3). Taken at face value, these correlations imply that the length of church regulations is indeed a function of various types of considerations, both politics- and performance-oriented.

Table 6. Bivariate correlations (N = 28).

	Statute length
Independent	0.36*
Orthodox	−0.17
Size(lg)	0.36*
Majority religion	0.48***
Regulation	0.16
Constitution(lg)	0.31*

***p ≤ 0.01 **p ≤ 0.05 *p ≤ 0.10

Discussion and outlook

Despite the intensity of research programmes on the determinants and consequences of regulatory length in other domains, there has been no similar effort to understand the phenomenon in the domain of organised religion. This is peculiar, considering the enduring scholarly fascination with questions of authority and organisation within churches. Introducing a new perspective to the comparative study of organised religion at the meso-level of analysis, the article surveyed literatures from various fields to develop conjectures regarding determinants of length in the core rules of modern churches. Specifically, institutional religion served as a hybrid terrain that allowed the application of theoretical frameworks originally developed in the study of national constitutions, organisational (bureaucratic) performance, and the economics of religion. An empirical supplement to this theoretical work used a small availability sample to establish that church rules differed dramatically in terms of length (number of words, number of sentences, number of provisions) and to explore some of these conjectures.

Several next steps come to mind. In addition to expanding the small and non-representative sample of cases and to measuring variables that could not be included in this article (e.g. corruption), the next logical question to ask concerns the post-adoption effects of length variation on other processes. Can the introduction of longer and more detailed statutes help a church to address principal-agent conflicts surrounding ecclesiastical governance? Can the introduction of shorter and simpler statutes help a church to gain operational flexibility and increase its adherents?

There are endogeneity problems that are inherent in these questions. Specifically, rules can at the same time be a reaction to pressures and relieve or exacerbate these pressures. To overcome these problems, future work in this direction needs to pay attention to the timing of amendments to statute length and how this relates to performance and political dynamics. Such an undertaking faces the considerable task of locating previous versions of the statutes for an adequate number of churches, even different churches from the same country, drawn from an adequate number of time points.

In another line of enquiry, the separate discussion of political and performance considerations that the article followed, although analytically neat, might not be describing reality. A less ideal-typical possibility is that performance considerations are embedded strategically in political considerations and vice versa. In this setting, drafters of ecclesiastical rules are responsive to the interplay of different pressures, rather than to these pressures separately. For example, perhaps longer rules are used by churches that are both independent and large in size or have some other combination of features. This more sophisticated depiction can be evaluated with larger datasets that facilitate the specification of moderating effects between the main explanatory variables.

The research into the different logics that determine the length of church statutes has implications for our understanding of churches as complex institutions. Contrary to simplistic depictions, the analysis developed in this article implies that, in order to serve their core aim of salvation, churches are called to play two worldly games simultaneously: the political game of representation and control and the performance game of efficient administration. Criticism of the ills of ‘church politics’ or, conversely, laments about the infusion of church administration with ‘private-sector’ practices may be misunderstanding this dual nature of religious institutions.

Acknowledgements

Earlier versions of the article were improved with support from the Visiting Scholar Programme of the Quality of Government Institute, University of Gothenburg, Gothenburg, Sweden.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Stratos Patrikios

Stratos Patrikios is a Senior Lecturer at the University of Strathclyde in Glasgow, UK. His research interests are in the area of religion and politics with a focus on organisational processes, church–state relations, and mass attitudes. CORRESPONDENCE: Department of Government and Public Policy, University of Strathclyde, 16 Richmond Street, Glasgow G1 1XQ, UK.

Appendix

Table A1. Document titles.

Church (document title)
Orthodox	Serbian Orthodox Ch. (УСТАВ)
Belarusian Orthodox Ch. (УСТАВ)	Romanian Orthodox Ch. (STATUTUL)
Greek Orthodox Archdiocese of America (Official Charter)	Polish Orthodox Ch. (Statut)
Russian Orthodox Ch. / Moscow Patriarchate (УСТАВ)	Bulgarian Orthodox Ch. (УСТАВ)
Estonian Orthodox Ch., Moscow Patriarchate (УСТАВ)	Protestant
Orthodox Ch. of the Czech Lands & Slovakia (Ústava)	Evangelical Ch. of Augsburg Confession (Verfassung)
Moldovan Orthodox Ch., Moscow Patriarchate (STATUTUL)	United Protestant Ch. (Constitution)
Orthodox Ch. of Albania (STATUTI)	Evangelical Ch. in Croatia (USTAV)
Montenegrin Orthodox Ch., CPC (УСТАВ)	United Protestant Ch. of France (Constitution)
Orthodox Ch. of Finland (Kirkkojärjestys)	Reformed Ch. in Hungary (TÖRVÉNY)
Georgian Apostolic Autocephalous Orthodox Ch. (დებულება)	Ch. of Ireland (Constitution)
Ch. of Cyprus (Καταστατικός Χάρτης)	Evangelical Lutheran Ch. in Italy (Statuto)
Ukrainian Orthodox Ch., Kiev Patriarchate (СТАТУТ)	Evangelical Lutheran Ch. of Latvia (Satversme)
Ch. of Greece (Καταστατικός Χάρτης)	Protestant Ch. in the Netherlands (KERKORDE)
Macedonian Orthodox Ch., Ohrid Archbishopric (Constitution)	Ch. of Sweden (Kyrkoordning)

Table A2. Length measures.

Church in	Words (count)	Sentences (full stops)	Articles (numbered)	Statute Length
Sweden	54817	3177	1425	2.21
Ireland	82294	3217	582	2.08
Austria	31976	3378	127	1.28
Bulgaria	21906	1779	274	1.06
Russia	15601	1079	435	0.89
Romania	24472	1136	205	0.82
Serbia	18034	940	269	0.78
N. Macedonia	16544	825	205	0.55
Hungary	11144	1158	173	0.52
Greece	15942	1539	75	0.45
Ukraine	9636	669	271	0.36
Latvia	7525	814	156	0.23
Cyprus	13620	806	92	0.19
Georgia	5404	841	187	0.15
Finland	6855	468	143	−0.03
France	16982	580	36	−0.14
Montenegro	6466	404	68	−0.41
Italy	5598	447	43	−0.63
Albania	5869	250	70	−0.71
Moldova	3953	249	105	−0.75
Belgium	5223	345	41	−0.80
Slovakia	3786	205	29	−0.94
Poland	2880	297	48	−0.98
Estonia	5182	629	7	−1.00
USA	4497	227	25	−1.14
Croatia	2686	199	47	−1.16
Netherlands	3855	225	19	−1.34
Belarus	2288	108	41	−1.54

Note: Statute length provides regression-based factor values via principal component analysis, using logs of the three textual variables.

Table A3. Linear regressions of statute length, with country-level controls.

	(1)	(2)	(3)	(4)	(5)	(6)
Independent	0.36*	- -	- -	- -	- -	- -
Orthodox	- -	−0.17	- -	- -	- -	- -
Size(lg)	- -	- -	0.56**	- -	- -	- -
Majority religion	- -	- -	- -	0.64***	- -	- -
Regulation	- -	- -	- -	- -	0.31	- -
Constitution(lg)	- -	- -	- -	- -	- -	0.33*
GDPpc(lg)	0.11	0.04	0.34*	0.39**	0.29	0.13
Surface	0.07	0.08	−0.22	−0.09	−0.01	0.10
[Obs]	28	28	28	28	28	28
[R^2]	0.15	0.04	0.24	0.36	0.09	0.13

Note: Main cell entries are standardised regression coefficients. GDPpc(lg): Logged GDP per capita, PPP, of host country (World Bank 2017). Surface: Land area of host country in km² (World Bank 2017). Constant included. All Variance Inflation Factors are below 1.5.

***p ≤ 0.01 **p ≤ 0.05 *p ≤ 0.10