Religion without doctrine or clergy: the case of Ancient Greece

Abstract The paper examines doctrinal and political reasons to explain why the Ancient Greek religion did not feature a distinct class of professional priests as suppliers of religious goods. Doctrinal reasons relate to worshiping a multitude of powerful anthropomorphic gods with flawed characters; absence of a founder of religion and of a scripture; lack of religious doctrine and of a code of moral behaviour and piety manifested as mass participation in rituals. These factors denied religious suppliers the opportunity to form a monopoly acting as an autonomous intermediary between humans and gods. Political reasons relate to the supremacy of the demos which watchfully guarded its decision-making powers and prevented other actors like a priestly interest group to challenge its authority.


Introduction
From a consumer economics perspective, religion confers utility to individuals in the form of worship, explanations of the origin of things, comfort against suffering, dispensation of sins and promise of salvation. From an institutional analysis perspective, religion is a fundamental determinant of social interactions and economic performance. By sanctioning some values and banning others it affects individual ethics and social norms shaping actions and outcomes. This is perhaps even more so for ancient societies where, compared to modern ones, formal institutions were weak. The Ancient Greek religion proved to be a formative force of continuity across generations and a major contributor to social cohesion. Even after its disappearance, its myths about gods and heroes have remained an inexhaustible source of flights of fancy to thrill the most demanding imagination and of ethical dilemmas to vex the most rigorous intellect.
The Ancient Greek religion was tightly entwined in the political and social life of the Greek city-state (polis) and the private activities of the individuals. From the family to the city-state, religion was an inexorable part of everyday life. ' [T]here was no sphere of life (or death) in Ancient Greece that was wholly separate or separable from the religious: the family, politics, warfare, sport, knowledge' (Ogden, 2007: 1). The calendar was the order of the religious festivals. 'Cult was … the backbone to much civic life, integrated closely with politics, and taken for granted to such an extent that changes in political life would be reflected in cult practice' (Thomas, 2000: 75). In this respect, religion was embedded in politics. The polis regulated public religious practices, festivals, sacrifices and laws on impiety. 1 The appropriate rituals had to be performed before taking public or private actions.
In the tradition of using economic reasoning to explain the observed record of economic performance and institutions of Ancient Greece, the current study examines a distinct characteristic of the Greek religion, namely, the absence of clerical body, a priestly class as a closed group of specially trained and ordained members with corporate interests and in some religions with hierarchical ranks. The lack of a corporate class of priests draws a clear line of distinction from societies with monotheistic religions, Judaism, Christianity and Islam, as well as the religions of Ancient Egypt and the Near East where the ruler represented the divine or was divine himself. Moreover, it is in sharp contrast to other branches of Indo-European religion such as the Iranian and the Indian ones which came from a common stem with the Greek religion. Building on research which applies the rational choice approach to religion (see Iyer, 2016 for a recent survey), the current study attributes the absence of a priestly class to the lack of religious doctrine. It explains that in the Greek polytheistic religion the lack of a founder of religion and sacred scripture, the absence of a code of moral behaviour and the reliance on local and panhellenic rituals negated the spiritual role of the priests as intermediaries between the gods and the humans. It then explores a second, complementary, argument, namely, the dominance of the demos in the political life of the city-states of Ancient Greece abrogated the potential of a priestly class to emerge as a source of social power.
The inquiry opens with a brief account of salient features of the Ancient Greek religion. Section 3 surveys recent studies using the rational choice framework to explore the connection between religious doctrine and the organizational structure of the clergy, and the nexus between religion and political power. Section 4 introduces a model of the priestly class forming an interest group where the size of group membership is related to the political power of the demos as a sovereign decision maker, the number of gods worshiped, the elasticity of demand for religious goods and the heterogeneity of the group of priests. Based on the predictions of the model, section 5 shows that the absence of monopoly power in the provision of religious goods, and the rise of democratic politics explain why a Greek priestly class failed to emerge. Section 6 concludes.

Salient features of the Ancient Greek religion 2.1 A multitude of gods
Greek speakers entered the Aegean around 2000 BCE. The Greek religion had multiple roots; the pantheon of gods combined the elements of deities from the various civilizations of the Ancient Near East, local pre-Hellenic populations and Indo-European roots (Burkert, 1985, Dowden, 2007. Although Ancient Hellas consisted of numerous independent city-states, common religion and language were the attributes of the common Greek identity. The Greeks were pious towards their gods, as attested by the Greek literature (Harrison, 2007). They worshipped multiple gods, a religion known as polytheism, where 'a god…never disputes the existence of another god' (Burkert, 1985: 216). By the 8th century, the pantheon had reached the form known to us and stayed stable for the next 1,200 years. At the pinnacle stood the Twelve Gods of Olympus. 2 In addition, there was a wide variety of lesser deities, as well as demigods and mythical heroes overseeing all aspects of life, and other divine figures associated with local landscapes, rivers and the sea, whereas the 'chthonic' gods were connected to the underworld. The Greek 2 The Twelve gods of Olympus were (Roman names in brackets): Zeus (Jupiter), king of the gods, god of the sky, weather, fate and justice; Poseidon (Neptune), god of the seas and earthquakes; Hephaestus (Vulcan), god of fire, blacksmiths and volcanoes; Ares (Mars), god of war and violence; Apollo (Apollo), god of light, healing oracles, poetry, music and arts; Hermes (Mercury), god of trade and messenger of the gods; Hera (Juno), wife of Zeus, goddess of marriage, women, childbirth and the family; Athena (Minerva), goddess of wisdom, knowledge, handicraft and warfare; Aphrodite (Venus), goddess of love, beauty and sexuality; Artemis (Diana), goddess of the hunt, nature, the moon and protector of young women; Demeter (Ceres), goddess of fertility, the harvest and sacred law; Hestia (Vesta), goddess of the hearth. The list features six male and six female deities as taught in Greek schools today. However, based on the famous east frieze of the Parthenon, other authors leave Hestia out and include Dionysus (Bacchus), god of wine and theatre. Hades, Zeus' brother, was the god of the underworld. It bears noting that Hades ruled over the souls of the dead but was not conceived as causing death, nor was his kingdom the equivalent of Hell (Felton, 2007: 90). city-states worshiped the same gods and goddesses but not in a uniform way as they practised local rituals. Different epithets characterized each god and goddess in different places and for different functions. 'The Twelve Gods were a group of variable composition, and this potential for variation allowed the Twelve to express either a local or Panhellenic identity depending on the context' (Larson, 2016: 31). When referring to religious matters, the Greeks used the expression 'the things of the gods' rather than the single word 'religion' (Cartledge, 2009). This probably reflected that they conceived religion and the world of the humans as fully integrated (Fustel de Coulanges, 2001[1877; Hedrick, 2007). For them, religion was not a matter of faith; that 'the gods existed and intervened in human affairs was a widely shared inference rather than an article of faith' (Larson, 2016: 6). 3 The poleis did not impose exclusions or other restrictions on their citizens regarding which deities to worship. To use the market analogue, the city-states provided a pluralistic and unregulated religious environment, where religious suppliers faced few constraints to introduce cults and rituals. From the Classical Era (5th and 4th centuries) the Greeks were open to the introduction of additional deities and foreign cults, which were often merged with existing Greek gods and goddesses. 4 The new deities were added to the existing ones instead of supplanting them. The trend of accommodating new deities and partly equating them with existing ones intensified during the Hellenistic times (323-31 BCE). To legitimate their rules, the kings of the Hellenistic states introduced cults which deified the ruling dynasty by claiming mythical ancestry and divine protection. The Greek religion where worshipers supplicate and propitiate several deities, is an early example of what Iannaccone (1995) termed a 'private' religion. In a private religion, worshipers turn to different suppliers, gods in the Greek case, for obtaining religious goods. In contrast, in a 'collective' religion like Judaism, Christianity or Islam, individuals enter exclusive relationships with a single god, and believers pay worship in congregation. At first blush, it may seem that the piety of the Greeks is consistent with the prediction that competition in the religion market increases religiosity by offering believers a greater variety of choices (Iannaccone, 1991;Iannaccone and Bose, 2011). However, it must be borne in mind that contrary to the competitive market, where people may switch between different religious suppliers, the Greeks made simultaneous use of multiple suppliers to ensure that none of the gods would be disregarded and offended. 5

Religion without a founder, sacred text or doctrine
Greek polytheism was not founded by a single charismatic leader; nor did it have an accepted, authoritative, written scripture. According to Herodotus, the 5th century historian, the poets Hesiod and Homer 'are the ones who taught the Greeks the descent of the gods, and gave the gods their names, and determined their spheres and functions, and described their outward forms' (The Histories 2.53.2). The religion was a patchwork of myths told and re-told over several generations. Myths were stories which included supernatural claims about the origins and deeds of gods and humans. According to Dowden (2006: 39) 'We need to bear in mind that Greeks told their myths in whatever way seemed useful for the purpose in hand. There was no Greek bible of myth, and details and precise names do on occasion vary'. Local myths were re-enacted in rituals and offered a 'supposed reason for a current religious practice' (ibid.: 42). They also inspired the plays of the tragedians. The wild and unrealistic claims of the myths explore life through extreme examples (Bowie, 1995: 470). Oral transmission implies that, contrary to a religion written on a text, the myth is easy to change and adjust, which explains why there were different localized versions about the actions and exploits of the 3 However, not all Greek thinkers subscribed to this conceptualization of the gods. Starting with the Pre-Socratic philosophers, who questioned anthropomorphism, theological inquiry culminated with Plato who brought about a revolution in religious language and in piety (Burkert, 1985: 275). 4 For example, in the 5th century the Athenians allowed Cypriots to establish a shrine of Aphrodite and they also accepted worship of the Thracian goddess Bendis who was associated with Artemis. Athens also granted cult status for Pan (from Arcadia) and the healer Asclepius (son of Apollo), whereas the Hellenistic temple of the Egyptian gods near Marathon offers another example from the Roman times. 5 I owe this distinction to an anonymous Referee. same Greek deity or hero. Like the Ancient Iranian and Indian religions, the Greek religion was bequeathed by the ancestors. But unlike the Greek religion, the Zoroastrian Iranians and the Vedic Indians for centuries transmitted orally collections of ritual hymns, not myths, which because of their holiness became fixed already in the oral period (Basuchoudhary et al., 2020;Ferrero, 2021). Lacking a systematic doctrine and a standardized text, the Greek religion was transmitted through what Boyer (2001) calls 'imagistic acquisition of religious concepts' which is based on performance and participation in rituals, as opposed to doctrinal transmission by an organized class of religious specialists (Larson, 2016: 191-194).
The Greek gods and goddesses were human-like, but fundamental differences remained: the gods were born but they were immortal; they ate and drank but their food and drink were ambrosia and nectar, not human; they were not visible to humans because they operated unseen or in disguise. They could be offended, feel disgust and get angry as humans did, but they were powerful and grand. They were as flawed as the humans driven by passions; they were jealous, spiteful, revengeful, sometimes scheming and quarrelling and stayed silent on how humans should treat each other. Importantly, the Olympian gods had not created the world; the world existed before them, but each one had a role in keeping it in order. 'For most Greeks, then, acceptance of some version of the religious system involved a belief in interaction with the gods, including an acceptance of some sort of ordering role for them. The gods were seen to act in the visible world, and to that extent religion was inevitably concerned with basic, "secular" needs' (Kearns, 1995: 517). Supplications to the gods related to earthly matters rather than salvation, since the world of the dead as described by Homer was a dreary, joyless place of ghosts, as there was no notion of reward or punishment for one's deeds in life.
Like all religions, an exchange relationship between humans and the gods permeated the Greek religion. Humans asked for favours and in exchange made sacrifices. 6 The relationship between the gods and the humans was reciprocal: the gods owed their status and prestige to mankind almost as much as men and the city owed their existence and prosperity to the gods (Cartledge, 2009: 15-16), but it was one of unequal powers. Humans who wished to insure against losses and receive good fortune had to appeal to several deities, because the jurisdictions of the gods overlapped considerably. Homer's Iliad makes clear that in situations of conflict, different gods support different sides. Thence, a worshipper did not choose to respect some gods and ignore others, but had to venerate all of them, although to a different degree. Making offerings to several gods is consistent with portfolio diversification under uncertainty. Considering the worship of several different gods like shopping from several firms offering differentiated products, Ferrero and Tridimas (2018) show that the overlap of the competencies of the gods was costly for the individual worshiper and generated inefficiency. Consumers of religious services would have gained by worshipping a single deity (something which took place after paganism ended).
The gods and goddesses did not decree norms of social behaviour; nor were they interested in the moral choices of individuals, whether people should be charitable, kind or pugnacious. Equally, the gods stayed silent on stealing, adultery or homicide. Furthermore, with earthly life not affecting afterlife, there was no moral code for earning salvation. Nor did the gods act as models of behaviour. They only demanded respect and ritual sacrifices from humans. Lacking a code of moral behaviour, the Ancient Greeks could not turn to the gods for moral guidance. In the absence of a moral code 'religion had far less of a reach into daily life [and] was less constraining on behaviour' (Basuchoudhary et al., 2020: 197). 7 6 See Stark and Bainbridge (1980) and Iannaccone (1995). The hypothesis that individuals confronted with uncertainty enter a reciprocal exchange with the source of uncertainty, like a deity, by offering sacrifices, receives empirical support by Frijters and Baron (2012), who using data from experiments find people are willing to sacrifice large portions of their income in the hope of a return favour, even when this has no discernible effect on outcomes. 7 The lack of moral code does not imply that the Greeks lacked moral compass. They did accept that the gods, although not exemplary in their moral behaviour, punished wrongdoing against parents, guests, supplicants and the dead, as well as oath-

Rituals
Economic analysis of religion emphasizes that public rituals create common knowledge and a shared experience for the participants signalling that they are members of the same community (Gill, 2020;McBride, 2016). The Greek religion was founded on ritual tradition upholding custom (rather than faith, love and hope as in Christianity). Rituals were closely and attentively connected with the myths of the gods and the heroes. Important family occasions, like marriage, childbirth, stages of farming and death, were also marked by religious rites. Rituals of animal sacrifice, bloody and messy businesses as they were, required altars, rather temples. Temples were houses for gods rather than places for worship. Sacrifices took place outside not inside the temple. Since people gathered outside the temples, it was their exteriors which were brilliantly decorated and still acclaimed for their beauty; unlike Christian churches, their interiors were not decorated. The inside of the temple housed the statue of a god and dedications by worshipers.
For the Greeks piety was to recognize and respect the gods by performing the relevant rituals. Worshippers, paraded, prayed, sang hymns, danced, watched or participated in athletic contests and theatre plays, presented offerings to the sanctuary of the god, sacrificed animals, offered fruits, vegetables and cakes, and poured wine and milk. The most central religious function was the sacrifice, which sometimes could be performed by any male person. Piety required libation, sacrifice, fruit offerings and prayer, but there were no formulas for liturgical prayer to be recited and handed down from generation to generation (Burkert, 1985). The ritual was attended in the expectation that it would produce certain beneficial effects, but it was understood that the gods could refuse without having to justify their decisions (Harrison, 2007).
Participating in the rituals, performing sacrifices and taking on their financial costs was a most significant factor in one's life. It attested the commitment of individuals to the polis and its religion. 8 For the individual participation was a signal of obedience to the norms of the society. For the community it was a screen showing that the individual accepted such norms. Participation in the ritual enabled the formation of a group, created solidarity among its members and buttressed prosocial behaviour. 'For Aristotle there is a philosophical certainty that there are gods: "the rest"by which he obviously means the whole of mythology and ritualwas introduced to persuade the multitude and with a view to practical use for the laws and expediency' (Burkert, 1985: 247). Ritual and invocation of the gods was the basis for mutual understanding and trust. Sacrificing together and attending festivals together were the most important contexts for the expression of social intimacy (Parker, 2005). 9 2.4 Religion without a class of organized priesthood Priests are the producers of 'religious goods'; typically, they perform rituals and are expert preceptors able to inform, teach and provide moral guidance to the rest of the society. But, the Greek religion did not feature a class of priests recruited, trained and sanctioned by a special religious institution, which would supervise them and guarantee their competence in supplying the religious goods. 10 Rituals could be conducted by men or women provided some conditions of body wholesomeness or purity were satisfied which were most probably rooted in magical beliefs (Larson, 2016: 222). Rather than being members of a separate clergy, an organized hierarchical organization of priests with corporate interests, the Greek priests were laymen, officials of the state like any magistrates rotating in office, breaking and hubris, that is, boasting against the gods (Larson, 2016: 127-132). The punishment could be severe in the form of disease, famine or death. 8 'The religious idea was, among the ancients, the inspiring breath and organizer of society' (Fustel de Coulanges, 2001: 109). 'Just as a domestic altar held the members of a family grouped around it, so the city was the collective group of those who had the same protecting deities, and who performed the religious ceremony at the same altar' (ibid.: 121). 9 This is also mirrored in that certain offences were punished by excluding the guilty person from religious rites (Bowie, 1995). 10 It bears noting that the institutional landscape of Ancient Greece lacked not only a clerical body, but also other governance organizations commonly found in contemporary societies, like professional judges, legal experts and political parties; Tridimas (2019) explores the absence of the latter in the context of direct democracy. administering rites (Burkert 1985;Knox, 2008). There was no head of church or hierarchy, like the Christian ranks of bishop, priest and deacon. Nor did the priests have to signal commitment to religious service by special lifestyles like celibacy or dress. The priests were local specialists performing rituals in specific temples of specific gods. 'As there was no hierarchical "clergy", each cult had its own priest, and, whether at the deme or polis level, priestly powers were limited to performing the customary rites endorsed by the polis' (Bonnechere, 2013: 368).
It is important to emphasize that the priests were limited local specialists. Contrary to Christianity 'there was no such thing as a priest of Apollo in general, but only priests of Apollo's temple at Delos' (Ferrero and Tridimas, 2018: 145). 11 A visitor passing through different poleis would find different ceremonies in honour of the same god. The reason was that despite using the same name for the gods, people in different poleis did not worship identical deities (Fustel de Coulanges, 2001: 123-126).
In general, priests and priestesses received a share of the items offered for sacrifice and, in some cases in democratic Athens, an annual payment for their services. 12 Nevertheless, priesthood was not a life-time professional which could earn a living but a part-time activity. Priesthoods for some old established cults were hereditary, indicating that some aristocratic families retained long-held privileges. 13 However, priesthoods for newer cults, like 5th century Athena Nike of Athens, were not hereditary and were allocated by lot, the hallmark of classical democracy. 14 3. Religion, priests and rulers 3.1 Theology and priests Research applying rational choice theory to the study of religion has shown that there is a connection between religious doctrine and the organization of the priesthood as a self-appointed group claiming to speak and act on the behalf of the gods.
Looking at Christian denominations Allen (1995) argues that the doctrine regarding moral behaviour in earthly life and how to earn salvation constrains the organization of the Churches because doctrine creates opportunities for some members to exploit their fellow members. Specifically, doctrine creates opportunities for wealth transfers from lay members to the clergy, and even within the ranks of the clergy. To diminish this risk, when doctrinal demands imposed to the members increase so should the constraints on the clergy. Thus, if different doctrines create different 'cheating' opportunities, churches will have different organizational structures to mitigate these opportunities for exploitation. For example, the Roman Catholic Church obliges its members to follow the teaching of the Vatican (so that they are not free to interpret doctrinal issues) and is organized as a hierarchy, where governance decisions are made by the Pope and the church bureaucracy. At the opposite extreme, in the 'Society of Friends' informally called the Quakers, rather than submission to any authority members may reach knowledge of God by first-hand experience, whereas on the organization side, decisions are made by voting; all members are encouraged to take part, and there is no professional clergy, which means that in practice it is almost impossible to exploit other members for financial gain. Ferrero (2008) examines the relationship between the size of membership of the early Christian Church and doctrinal rigidity. When the Christian Church was persecuted by the Roman authorities 11 'As a general rule, if a priest could speak with authority in the assembly, it was only on the affairs of his own sanctuary; to offer advice on those of another would very likely have been perceived as an intrusion' (Parker, 2005: 94). During the Hellenistic times, in many places of Asia Minor, priesthoods were auctioned (Burkert, 1985: 96;Lane Fox, 1988: 60-61). 14 The only true religious professionals were the oracle-mongers (manteis) and the seers or soothsayers, private agents not sanctioned by any formal institution. Their professional success depended on whether they had enough followers, effectively customers, asking for their services and acting upon their divinations. Bonnechere (2013)  it was organized as an exclusive sect strictly controlling membership and distancing itself from outside competitors but broadly tolerant of alternative interpretations of the theology. After its triumph in the 4th century, the Church opened its membership granting free access and becoming inclusive, but it hardened its doctrinal stance as refined and clarified in various ecumenical councils. The reason for hardening its stance was that 'If anyone could join in and undertake an ecclesiastical career simply by subscribing to a broad, vague set of unconstraining and undemanding principles, then the benefits of monopoly accruing to the professional clergy and especially its leadership would be diluted away by excessive entry … So, while universal admission lowered the price of membership, doctrinal radicalization raised it, protecting the hierarchy's rents' (Ferrero, 2008: 79). Ferrero and Tridimas (2018) show that in the Ancient Greco-Roman religion a substantial, and over time increasing, overlap between the jurisdictions of different gods took place; this was coupled with the absence of a class of professional priests, as priesthood was not a full-time profession. The jurisdictional overlap was inefficient for it meant that gods valued and rewarded the size of offerings of supplicants who because of budget constraints could only partially satisfy each god. Supplicants would have benefited from an integration of the jurisdictions of the gods. The latter, however, did not materialize because there was no professional religious class which could act as an interest group and profit from the unification of supply of the religious services.
Such notions are further explored by Basuchoudhary et al. (2020) who attribute the resilience of polytheistic Hinduism as opposed to the demise of the Greco-Roman religion to the theological structure and the role and organization of the priesthood. Contrary to the Greco-Roman conception of the divine, in post-Vedic, sectarian Hinduism, the gods are seen as valuing and rewarding the devotion of supplicants rather than the size of offerings they receive. The resulting concentration of supplicants' resources on their god of choice, incentivized the priests to cooperate and implement their monopoly of ritual services maintaining their caste privileges. The different conceptions of the demand by the gods for sacrifice imply different payoffs for the priests of the two religions in a prisoner's dilemma game. The Hindu priests in effect play the game repeatedly and as a result they developed into a professional body of specialists in the service of different gods. On the contrary, the Greco-Roman priests having had no certainty about the future rounds of the game remained fragmented and did not see themselves as a professional class.
Similarly, Ferrero (2021) concludes that theology and an organized priesthood account for the radical transformation of the ancient Iranian polytheism to Zoroastrianism, founded by the prophet Zoroaster around the middle of the second millennium BCE, which became the dominant religion of Persia for more than a millennium (until the arrival of Islam in the 7th century).

Religion and rulers
'That religion is a means to maintain authority and domination was stated by ancient authors from the fifth century onwards as a self-evident state of affairs. The position of ruler always entails priestly functions; status is dramatized and thus confirmed by ritual' (Burkert, 1985: 257). More generally, anthropologists have suggested that the origin of kingship is found in the performance of rituals, with those chosen as kings performing special functions for the society (Atack, 2020 and the literature therein). It suits a rational, self-interested, ruler to claim endorsement by gods. Religious legitimation of the ruler makes his subjects and co-religionists less eager to defy his rule so they will acquiesce to paying their taxes. The implied fall in law-and-order and tax collection expenditures increases the surplus available to the ruler as a state proprietor for other types of spending including his personal consumption. Moreover, the religion may strengthen the fighting spirit of the ruler's army strengthening his position against his enemies, internal and external. The ruler may privilege the religion of his choice by providing cash transfers or tax exemptions to the faithful, employment rules favouring co-religionists, building religious monuments, restricting other faiths, expropriating their properties and even banning them. In addition to its legitimating role, a ruler as a believer may wish to promote his religion.
Analysing the relationship between religion and the government, Gill (2005) shows that the selfinterest of politicians guides decisions to regulate religious freedom and the rights and privileges (or lack thereof) of religious organizations; he notes that, relatively to coercion or patronage, ideological legitimacy is the cheapest way for the ruler to secure compliance of the population, especially if priests are trusted. Comparing religious developments in the USA and Latin America he concludes that weak political leaders tend to favour hegemonic religions (which command the loyalty of the majority); however, secure politicians tend to put in place strict regulations while governments facing fiscal crises even expropriate the assets of the church.
Coşgel and Miceli (2009) assume a state which aims to maximize tax revenue minus expenses for religious goods, where the latter provide utility to the citizens, and priests legitimate the ruler by granting him divine status, or a divine right to rule; in turn, legitimacy of the ruler lowers the costs of tax collection. They show that the strength of the legitimating effect, the degree of concentration in the religion market and the degree of democratic governance determine the decision of the ruler to fund or repress religion or grant independence to the religion market. According to their empirical results, religious loyalty to the state and concentration in the religion market, have positive and statistically significant effects on the state control of religion, whereas democracy has a negative and insignificant effect. Furthermore, examining religion as a legitimating force for political leaders, Coşgel et al. (2018) construct a cross-national time-series dataset for the period 1000-2000 CE and estimate the effects of the degree of concentration in the religion market, historical inertia of a state (which determines the cost of monitoring religious or secular legitimating agents), ethnic and religious differences between rulers and the general population, and the prevailing political regime on the adoption of a secular state.
Examining 31 cases of rulers introducing new state religions, Vaubel (2017) finds that autocratic rulers tend to introduce state religions to enhance their legitimacy when they are weak in the face of internal or external opposition; aim to unify heterogenous groups; promote religions with mass following; and support monotheistic state religions when they are powerful. Skaperdas and Vaidya (2020) also study a ruler seeking to maximize the difference between tax revenue and the cost of paying for guards to protect domestic producers from looters, where producers pursue material payoffs and religious rewards. They assume that the ruler is legitimated by adopting a 'Big-God' religion (Norenzayan, 2013). An all-seeing 'Big-God' demands a certain conduct of humans and is able and willing to punish transgressors even if undetected by other humans, an aspect which increases prosocial behaviour and facilitates large-scale cooperation. 15 The model predicts that the ruler enjoys higher office rents by making investments to increase the number of believers and the intensity of belief (in the form of more temples, priests and rituals), especially when the ruler too is a believer. Furthermore, when the size of the believers is very high relative to total population, the surplus received by the ruler is highest and the incentives for various investments are strong; these generate 'a drive towards consolidation and institutionalization of Big-God religions so that they become essentially monopoly ("established") state religion ' (2020: 19).
However, extending arguments traced to Adam Smith, Finke (1990) highlighted that state regulation of the religious market stifles its growth and diversity. Examining the withdrawal of state support for established churches from the late colonial USA to the 19th century, he concluded that deregulation of the market and separation of religious organizations from the secular state increased the level of religious mobilization of the population (see Finke and Iannaccone, 1993 for further discussion).

Priests as an interest group 4.1 Priests, monopoly in supply and rent seeking
The studies reviewed show that a priestly class becomes powerful because it may possess monopoly power in the supply of religious goods; in this case it develops power and obtains rents independently of the ruler's support but may nevertheless need it to retain the monopoly. But, even without monopoly, a priestly class may be granted privileges by the state if it advantages the ruler. 15 See also Norenzayan and Shariff (2008). However, Johnson (2015) emphasizes that what is important for pro-social behaviour is fear of supernatural punishment in general rather than a watching, moralizing gods.
In the religious market, a monopoly arises when only a single religious firm, 'church' and the associated 'clergy', supplies religious goods and services. 16 Without government restrictions, a monopoly supplier may emerge from doctrinal causes. If the consumers of religious goods consider a particular supplier as providing a unique package of services without close substitutes, that supplier may crowd out producers of alternative religious goods. 17 The package includes the benefits conferred by following the doctrine and moral code, as well as social networking with co-religionists, and other services like education and welfare. In addition, the likelihood of establishing a monopoly increases if there are economies of scale in the production of the religious good (e.g. delivering the same sermon to an expanding audience).
But, a religious market monopoly may not last forever since rival suppliers of religious goods including 'heretics' will challenge its dominance and try to win members. A religious supplier may then turn to the secular ruler and ask for support to fend off competition. Favours from the ruler benefit the priestly class by increasing its wealth, power and prestige (Gill, 2005(Gill, , 2020. In exchange, the favoured priestly guild proclaims supernatural justification for the ruler equivalent to kingship by 'the grace of god' which as discussed above confers great benefits to the ruler. This exchange assumes that by virtue of their occupational status priests have been able to form an effective interest group, so that they 'use political influence to enhance the well-being of their members' (Becker, 1983: 372).
In common with any interest group, the priestly class may pursue economic objectives, like gaining monopoly rents, or ideological objectives, like promoting a particular lifestyle. The bulk of the literature on interest groups focuses on competition between interest groups for rent seeking and the ensuing contributions to political parties, and the effects on election outcomes, public policy and the size of government. 18 However, our preoccupation here is the prior question of whether priests form an interest group to pursue their interests. Olson (1965) showed that the formation of an interest group is subject to the 'collective action problem'. A policy favourable to the interest group is a public good for each member of the group, non-rival and non-excludable, and therefore, subject to the free-rider problem. Hence, 'only groups that are relatively efficient at limiting free riding become politically powerful' (Becker, 1983: 392). Regarding the determinants of interest groups as an empirical matter, the literature has tested several hypotheses: a stable political environment offers opportunities for groups to organize; economic development by expanding existing activities and generating new ones adds new interest groups to the existing ones; democracy is conducive to interest group formation, since an open political system allows more participation; large-size countries feature large numbers of industries which in turn increases the number of groups; higher import competition stimulates demands for protection by domestic producers; larger shares of government spending attract more interest groups wishing to affect public policy and social diversity, linguistic ethnic and religious, creates more divisions and thence spawns more interest groups. Addressing earlier often contradictory findings of empirical work on these questions, Coates et al. (2007) report estimates from an international sample of countries which show that the political stability of a country, its level of socioeconomic development, political system, size and diversity all appear to contribute to interest group formation.

A formal model of a priestly interest group
This subsection investigates whether priests join an interest group drawing on the works of Murrell (1984) and more recently Ferrero (2021) on the optimal size of interest group membership. Like their work, it compares the benefit and costs of increased number of members, but unlike them, since it is not a priori clear whether a priest joins or not the group, it considers the benefit from membership as probabilistic rather than certain. 16 Ekelund et al. (1996) use notions of monopoly, rent seeking and transaction costs to explain several practices, behavioural restrictions and organizational procedures of medieval Catholicism. 17 On the other hand, Raskovich (1996) explores protection of intellectual property rights in theology as a potential efficiency rationale for a monopoly in worship. Let M denote the number of priests attempting to form a rent seeking interest group. Each member gets a benefit B if the quest for rents is successful with P probability of success. The benefit each priest derives depends negatively on the number of gods T stipulated by the religion because 'propitiating gods is a costly business in a world with scarce resources [implying that] propitiating one god reduces resources (and therefore rents to priests) available to propitiate another' (Basuchoudhary et al., 2020: 203). The benefit also depends negatively on the price elasticity of demand for religious services E since the higher the price elasticity of demand for religious services, the lower the profit maximizing price charged by a monopoly. 19 We then have B = B(T, E) with ∂B/ ∂T ≡ B T < 0, ∂B/∂E ≡ B E < 0.
Borrowing a leaf from conflict economics it is posited that a larger number of group members is expected to increase the probability of successful rent extraction, but the contribution of each additional member is diminishing. Furthermore, in a democracy there is no autocratic ruler maximizing tax surplus; instead, voters maximize their utilities. Inverting the reciprocal relationship between the ruler and the priestly class this implies that in a democratic society the probability of success of the priestly interest group is negatively related to the power of the demos, δ. Thus, P = P(M, δ), with ∂P/∂M ≡ P M > 0, ∂ 2 P/∂M 2 ≡ P MM < 0 and ∂P/∂d ; P d , 0. Furthermore, we assume that ∂ 2 P/ (∂M∂δ) ≡ P Mδ < 0, that is, an increase in the power of demos decreases the marginal probability of membership (the effect of the size of membership on the probability of successful rent seeking).
Although a priest joins an interest group to pursue common interests with other priests, the members of the interest group are heterogenous. They have different characteristics, and depending on their individual circumstances, they may bid for different policies. The latter implies that for each priest the cost of interest group formation increases with the increase in the size of membership. Moreover, in the context of Greek religion, the cost of joining increases with the number of gods, because the latter increases the heterogeneity of priests. With C denoting cost, we then have C = C(M, T ), with ∂C/∂M ≡ C M > 0, ∂C/∂T ≡ C T > 0, ∂ 2 C/(∂M∂T ) ≡ C MT > 0. The latter inequality reflects that the marginal cost (MC) of membership is increasing in the number of gods. Finally, the interest group is unsuccessful with probability 1 − P in which case each member gets a zero benefit. The expected benefit from membership of the priestly interest group is given by the expression: Maximizing the latter with respect to the size of membership M and denoting dU/dM ≡ U M , d 2 U/ dM 2 ≡ U MM we have Equation (2) determines M* the equilibrium size the priestly group membership as a function of δ, E and T. For M in equation (2) to be a maximum, inequality (3) must hold which will with C MM being positive, zero and even negative provided it is 'not too strongly negative'. The expression P M B shows the expected marginal benefit (EMB) from an extra priest joining the interest group and is drawn as the downward sloping curve in Figure 1, where both curves are drawn as linear for convenience. The For concreteness we abstract from demand factors which may increase the benefit of supplying religious goods. equilibrium size of membership is zero. Intuitively, when the individual benefit from joining a priestly interest group is smaller than the cost from the very start, the individual priest is better off by remaining a single unorganized supplier. The position of the curves, and therefore their intersection, depends on the effect of the number of gods, elasticity of demand and power of the demos. Formally this is shown by the comparative static properties of the equilibrium M*. Using the implicit function theorem, we identify how M* varies with its determinants: Analytically, an increase in the number of gods worshiped increases the diversity and heterogeneity of the priestly group because Greek priests are specialists rather than generalists. The expected benefit falls while the MC increases reducing the equilibrium size of the group: Since a higher price elasticity of demand decreases the monopoly power of the supplier, the associated rent falls bringing down the expected benefit from joining the interest group and so it diminishes the incentive to join: An increase in the power of the demos reduces the probability of successful rent extraction by the priestly class. When the demos rather than a single person control the state, divine legitimation of the government is superfluous and the priests may no longer receive preferential treatment. With the expected benefit from interest group membership falling the equilibrium size of the priestly group falls too.
In light of inequalities (4)-(6) the reasons why a Greek priestly class failed to emerge were the absence of priestly monopoly power and the rise of democratic politics. These are discussed in what follows. 5. Explanation of the absence of a priestly class with monopoly power 5.1 Market fragmentation: a multitude of gods and lack of doctrine With the religion featuring a multitude of gods, and a very large number of priests tending specifically local temples and sanctuaries dedicated to specific deities lacking the ability to serve all gods, the supply side of Greek religion was highly heterogenous and widely dispersed. This meant that the basic ingredients of a concentrated industry and economies of scale were missing. Moreover, the priests could not claim social power as agents intermediating between the gods and the humans. Our previous analysis revealed that several factors contributed to this absence of principal-agent relationship. The religion focused on ritual participation and upholding customs with respect to gods with actions known to every participant. There were neither a scripture nor rules of behaviour needing interpretation. An additional important factor comes into play here. A sacred text and the accompanying scholarship allude to literacy as an additional advantage held by an educated priestly class (Boyer, 2001). Priests were a literate elite in comparison with the rest of the society in the Middle East, Egypt, India or China, but not in Greece. Being phonetic, the Greek alphabet was easier to learn and use, compared to the cuneiform script or hieroglyphs. Although accurate quantitative information is not available, historians suggest that in Greece, rather than being the preserve of a priestly class, literacy was widespread. 20 Such factors imply that priests could not have claimed superior knowledge relative to the rest of the worshipers; they could not have been discretionary agents intermediating between humans and gods. Nor was it their responsibility to offer spiritual guidance or pastoral support. To put it another way, the supply of religious goods was routine without the need for expert knowhow. On this account, the reverse of Allen's (1995) argument applied: the lack of doctrine of the Greek religion translated into the lack of opportunities for the occupants of priestly offices to exploit the rest of the worshipers.

The political supremacy of the demos
Privilege for the priestly class in exchange for religious legitimation of the ruler was a feature of the Ancient Egyptian and Eastern monarchies, but not of the city-states of Archaic and Classical Greece. Kingship was of course present during the Archaic and Classical times, 750-300 BCE, and a close relationship between kingship and religion remained. For example, in oligarchic Sparta, which was headed by two kings reigning concurrently, the kings were responsible for religious matters. In democratic Athens, the so-called archon 'basileus' (king) was responsible for religious issues, which indicates that in the distant past kings were performing religious duties. But Greek kings were not divine figures. Ancient Greek thinkers like Plato, Xenophon and Isocrates sympathetic to kingship spoke of a virtuous rather than a divine king. They sought to identify a 'kingly craft' which would bestow on its possessor the authority to rule and lead in peace and war and deliver the good life for their citizens (Atack, 2020).
At the conceptual level, the multitude of the Olympian gods ran counter to the idea of a monarch by the grace of god. It is difficult to argue that the person sitting on the throne had been chosen by god or represents god if there are more than one god, and no god can undo the deeds of another god. Monotheism is arguably better placed to support royal rule than polytheism; a single god supports a single ruler. 21 In a similar vein, Economou and Kyriazis argue that ideologically the religion contributed to the emergence of ancient Greek democracy. They interpret the deliberations between the Olympians gods, described in the poems of Homer, as participatory decision making in a 'parliament of gods'.
The political implication of conflict between gods is illustrated by the changing fortune of the temple of Zeus in Athens. Peisistratus who in the years 564-527 ruled Athens as a tyrant (meaning coming to power extra-constitutionally) began the construction of a temple of Zeus to intimate the protection 20 Cartledge (2016)  of the ruler by Zeus. 22 But the tyranny was overthrown in 510 before the temple was completed and the democracy which succeeded it built a temple to Athena, the famous Parthenon of the Acropolis of Athens. The temple of Zeus was eventually completed early in the 2nd century CE by the Roman Emperor Hadrian long after the democracy had disappeared.
The transmission of cultural traditions including religion depends on the acceptance of authority. In Greece, the polis as a political institution developed after the demise of the Mycenaean kingdoms, starting from small, isolated settlements around 1,000 BCE and evolving over the course of four centuries. In the polis authority passed from the 'well-born', the landed aristocrats dominating the Mycenaean polities, to the demos of the ordinary (adult male) citizens. Pivotal to this political transformation was the technological change where cheaper iron weapons replaced expensive bronze weapons, which were the preserve of the aristocrats (Economou and Kyriazis, 2019: 12-30;Lyttkens, 2013: 336-338). In the polis, the man who carried arms and possessed full civic and political rights was able to extrude those who might challenge his authority, including not only priests but also women, those born outside the polis, and slaves. The emergence of the polis of free citizens as an independent selfgoverning political and economic unit with an assembly, council, magistrates, written laws and full sovereignty in its geographical territory, meant that power rested with the male citizens and no professional group could challenge it. The demos controlled cultic matters, and priests could have authority only by the grace of the demos. Parker (2005: 90) writes 'priests do not give orders to the assembly, but the assembly to priests' and calls priests 'politically disengaged' (ibid.: 95). Hence, priests were like any other layman, and vice-versa ordinary male citizens could perform sacrifices. Commenting on lessons that modern democracy can take from Ancient Greece, Walter (2013: 517) makes a similar claim: 'In contrast to many states in the Ancient Near East, religious functions offered no starting point for long-term power-building. The most central religious function was the sacrifice, which generally could be performed by any male person. No social group thus managed to monopolize religious activities and resourcesprayer, sacrifice, oraclesand conditions did not allow for any influential let alone dominant priesthood to evolve'.
Looking specifically at Athens, Garland (1990) describes three examples of the power of the state in cultic matters, namely, the vote of the assembly of the demos was required for the introduction of new public cults, the demos had 'overriding authority' in all matters involving the finances of state cults, and it was the state which prosecuted religious crimes. Gunther (2016) further illustrates the nexus of religion and politics in the political discourse of Athens. The assembly took for granted the existence of powerful gods and opened its meetings with sacrifice and prayers so that gods would remain on its side; but then as a sovereign body it proceeded to regulate cult practices, manage sanctuaries and control their finances. There was no room for a corporate priestly class in that environment.

Conclusions
The Ancient Greek religion was a localized religion which postulated the existence of a multitude of deities and perceived humans and gods as entering an unequal and asymmetric exchange relationship where humans had to perform rituals and offer sacrifices to placate the gods. It did not feature a formal founder, scripture or doctrine, and it lacked an organized priesthood. Instead of doctrine, the religion featured myths about gods and goddesses with human faults told and retold by poets rather than prophets or preachers. Without commandments how to behave on earthly life, participation in the ritual was the hallmark of religion and confirmed individual commitment to the community of the polis. The absence of doctrine also meant that priests as suppliers of religious services administered rituals but lacked sanctity. Priests were laymen, neither professionally trained nor ordained by a formal central organization.
Since the power and rents of a priestly class emanate from monopoly in the supply of religious services and privileges granted by the state, the current study argued that both these factors were missing 22 Dowden (2006) contends that kings ruled with an authority coming from Zeus, himself the king of the gods. However, he adds that Zeus was also the source of justice, and since before the codification of laws it was kings who heard cases and meted out justice, Zeus monitored and audited the kings, which effectively limited their power. from Greece. The worshipers could not see priests as agents between the gods and the humans. Hence, priests could neither form a monopoly as intermediaries between gods and humans, nor could they get state protection in exchange for its legitimacy. The political supremacy of the free citizens of the polis left no room for rival centres of hard or soft social power. As a result, priests remained civilian rather than sanctified officers.
But a warning is also in order. As with any observational study, we cannot exclude the possibility that the absence of a priestly class, is not causally related to polytheism, lack of doctrine and power of the demos, and that they all are the common result of other omitted factors like, for example, geography. What has become clear however is that the features of the religion discussed were inextricably linked and they mutually reinforced each other.