An institutional approach to the decline of the Ottoman Empire

This paper examines the selected Ottoman institutions during the socalled rise (fourteenth through sixteenth centuries) and identifies the institutional characteristics that may have led to the eventual fall of the Empire in 1918. We propose three criteria based on which the Ottoman institutions are selected. First, there should be nominal accounts of the institution. Second, the institution has to be present during the rise of the Empire. Third, the institution should allow the investigation of whether changes in it led to increased power sharing between the sultan and a larger segment of the society. As a result, the paper identifies three institutions: succession structure, power structure, and the identity of the Ottoman elites and the landownership-military-public finance triangle. Our conclusion is that the weaknesses in the mentioned institutions were fundamental enough to make the Empire vulnerable. Additionally, the examination of these institutions leads to the identification of even more fundamental characteristics of the Ottomans, such as their aversion toward Turkish Muslims and commerce as well as their oblivious attitude toward technological innovations. *Corresponding author: Ayse Y. Evrensel, Department of Economic and Finance, Southern Illinois University Edwardsville, Edwardsville, IL 62025 USA E-mail: aevrens@siue.edu Reviewing editor: Watson Duncan, University of East Anglia, UK Additional information is available at the end of the article

ABOUT THE AUTHORS Ayse Y. Evrensel is an Associate Professor of Economics in the Department of Economics and Finance at the Southern Illinois University Edwardsville. Her teaching focuses on courses such as International Finance, International Trade, Macroeconomics, and Financial Markets & Institutions. While her research focus is currently directed toward culture and institutions, she has published on corruption, exchange rate regimes, the IMF and preferential trade agreements. This paper is part of her research agenda that focuses on institutional quality. Even though her institutional research is often empirical and takes place in a multi-country setting, this paper is an attempt to provide a qualitative examination of institutions in a historical setting.
Tiffany Minx is a senior in the Department of Economics and Finance at the Southern Illinois University Edwardsville. She was the recipient of the Undergraduate Research and Creative Activities (URCA) assistantship during the 2015-2016 academic year.

PUBLIC INTEREST STATEMENT
Examining the rise or the decline of empires may be viewed as an unfruitful enterprise because of the opaque nature of the subject. It seems that the separation of cause and effect is particularly challenging in this line of research, especially when one considers the interactions among empires. Nevertheless, the benefits of focusing on grand subjects such as the rise or the decline of an empire lay in the enrichment of knowledge demonstrated by different viewpoints. This paper adds another perspective to the existing ideas that explain the decline of the Ottoman Empire either based on the marginalization of the Empire by the then rising Western-style World Economy or based on the Ottoman Empire's sheer longevity. Our addition to the existing discussion revolves around the Empire's own dynamics based on the characteristics of its selected institutions and the conflict of interest among various stakeholders.

Introduction
The Ottoman presence was known in the world from the thirteenth century to the end of WWI in 1918. At its height of power in the seventeenth century, the Empire's territories reached from the Balkans to the Black Sea and to North Africa, Middle East and the Arabian Peninsula. Its army marched as far west as Vienna and laid siege to the city twice in 1529 and 1683 (Kasaba, 1988;Quataert, 2000). The contemporary empires and kingdoms of the Ottomans in the West survived, albeit eventually losing the majority of their overseas territories. Still, none of them disintegrated as the Ottoman Empire. Among the most frequently mentioned accounts of the fall of this vast empire are the problems associated with a large, multi-ethnic (the Turks, the Arabs, the Greeks, the Armenians, the Jews, the Bulgarians, the Hungarians, the Albanians, etc.), and centrally governed empire with extraordinarily weak public finances (Shaw, 1976). In addition to the mentioned internal reasons, the marginalization of the Ottomans by the world economy in the eighteenth and nineteenth century is also suggested as an important external reason for the Empire's fall (Ergil & Rhodes, 1975;Kasaba, 1988).
The aim of this paper is to apply the institutional perspective to the selected areas of social, political, and economic decision-making in the Ottoman Empire and reinterpret the existing nominal accounts of events between the rise and the fall of the Ottoman Empire. The key idea is that any approach that examines the fall of the Ottomans should go back to the period of the rise and the expansion. Examination is expected to show that the Ottoman institutions, in other words, the explicit and implicit setups or rules, which governed the Empire during its rise may have been influential for bringing about the end as well. The methodology that we employ in our review of the selected Ottoman institutions relies on secondary historical sources that are reviewed qualitatively. We provide a new interpretation for these historical sources, while keeping the internal consistency of our argument in tact over the selected institutions.
Before we provide more information on the Ottoman institutions that we consider in this paper, we want to briefly talk about two subjects associated with our approach to the fall of the Ottoman Empire: the selection of the type of the institutional approach and the issues surrounding the choice of a "grand narrative" as the fall of an empire. First, because there are competing institutional approaches to economic history, we want to discuss the characteristics of the institutional approach that we apply to our subject. While Polanyi (1957), Wallerstein (1974Wallerstein ( , 1979a and North (1981North ( , 1990North ( , 1993 and North and Thomas (1973) investigate the emergence of fundamental structures throughout history, they do so quite differently. According to Polanyi (1957), the modern nation state initiated fundamental changes in the social and economic system as well as in human nature. Because a capitalist economy was thought to bring legitimacy to the nation state, the market economy was introduced, in which the price of inputs as well as of goods is determined in self-regulating markets. The fundamental nature of this transformation lies in the accompanying changes not only in institutions that are supposed to support the capitalist system, but also changes in human nature and its economic mentality. While people lived in economic spheres of reciprocity and redistribution prior to the great transformation, they became utility-maximizing rational agents afterward (Polanyi, 1957). Wallerstein (1974Wallerstein ( , 1979a world system emerges during what he calls the long sixteenth century in Europe when the continent was in dire circumstances starting the late thirteenth century. With the decline in agricultural production due to epidemics as well as climatic changes, feudalism went into crisis during the late thirteenth century from which the preliminary signs of capitalism emerged in the sixteenth century with its core and periphery. The first core was the Netherlands that was later replaced by England. Even though later on the core consisted of a number of powerful empires/ countries, the core-periphery relationship with unequal exchange was forming in its essence where the dominant core in all economic spheres (production, trade, finance, etc.) started imposing its rules on the periphery. Therefore, according to Wallerstein (1974Wallerstein ( , 1979a, the European success in getting out of feudalism has been detrimental to most of the world, because it set the unequal world development in motion. While there are differences between Polanyi (1957) and Wallerstein (1974Wallerstein ( , 1979a regarding the emergence of capitalism, they both point out to capitalism as a force that not only transformed individual states but also their relationships with other states, which eventually organized states in a hierarchical manner. At this point, we may want to consider how an approach to the fall of the Ottoman Empire based on Polanyi and Wallerstein would look like. In fact, Wallerstein (1979b) and Wallerstein et al. (1987) takes on this subject and states that even though the Ottomans built a world empire, they were not part of the core of the emerging capitalist system in the late sixteenth century, which eventually steered the Empire toward the periphery. Following Wallerstein (1974Wallerstein ( , 1979aWallerstein ( , 1979b and Wallerstein et al. (1987), Ergil and Rhodes (1975) as well as Kasaba (1988) note that the marginalization of the Ottoman Empire started in the sixteenth century and was completed by the end of the nineteenth century. According to this view, the Empire was victimized by the emergence of the capitalist system. North's (1981North's ( , 1990North's ( , 1993 and North and Thomas (1973) approach is similar to that of Wallerstein's except for the fact that the former sees nothing wrong with feudal Europe's struggle to improve its economic conditions starting in the late thirteenth century. In fact, North and Thomas (1973) contrasts the European kingdoms of that time based on their success in replacing feudalism with a new social and economic system and discusses the possible reasons for their success and failures, i.e. the rise of the Low Countries (the Netherlands) and England as super powers and the eventual decline of Spain and Portugal. Our paper follows North's approach, because we aim to observe the decisions of the Ottomans when they were considering setting up or changing an institution, taking the world order as given.
Second, the subject of this paper implies a "grand narrative," because its aim is to reexamine the selected institutions of the Empire that may have led to its fall. Especially in historical studies, such narratives have been abandoned in favor of insights into the authentic nature of the subject without using the terms rise, fall, success, or failure. Moreover, the subject implies a comparison of the Ottoman Empire with other (Western) empires, which raises the issue of ethnocentrism (Hofstede, 2001). The post WWII-era introduced new sensitivities in social sciences following the end of colonialism and at the dawn of the developing country phenomenon. In the 1950s, dozens of independent countries that were former colonies appeared on the world stage having different economic, political, and cultural systems. The term ethnocentrism was used to define the exaggerated tendency to think the characteristics of one's own group to be superior to those of other groups. Later research dissected ethnocentrism and found that it stems from the Western scholar's imposition of her values on research (Hofstede, 2001). A related term, orientalism, was introduced to describe an outcome where one group "others" another group based on the differences in identity (Said, 1978). The constructivist theory provides a remedy to ethnocentrism in cross-cultural studies by examining the "other" through the other's own lens and not through that of the researcher, which led to the emergence of historical studies that views the subject from the subject's point of view.
Our aim is not to provide a superficial comparison in institutional quality between the Ottoman Empire and European kingdoms. This paper intends to shed light on the decision-making mechanism of the Ottomans and explain the conditions that led them to particular decisions with respect to the selected institutions. It is a fundamentally economic idea to investigate the circumstances in which decisions are made, because the circumstances affect the cost-benefit calculations associated with every decision. Therefore, we would like to understand the circumstances in which the Ottomans made the institutionally relevant decisions. In this context, Tamura (2002) can be used to explain our approach to the study of Ottoman institutions. He develops an endogenous growth model and shows that the late adoption of agricultural technology in China compared to Europe is a function of relatively large land availability in China and not differences in human capital. In fact, in Tamura's (2002) model, human capital is endogenous to the relative availability of arable land in China and Europe. Therefore, the "why-the-West" question can be answered by utter desperation of the West in terms of available resources. We believe that discussions such as whether a realm was desperate enough to make a particular decision or introduce something new can be integrated into the Northstyle institutional approach.
The novelty of this paper lies in the synthesis of the selected institutional characteristics and the known outcomes. While these institutional characteristics were not perceived as weaknesses during the expansion, the seeds of the downturn may have been sown during the expansion. The question is as to which institutional characteristics of the Ottoman Empire should be chosen for the study. While we do not claim that the chosen institutional characteristics are the most important, they do fulfill the following criteria. First, there have been nominal accounts of them in the literature. Second, these institutional characteristics were introduced or strengthened during the rise. Third and most important, the institutional structures examined in this paper enable us to discuss whether they increased accountability in the decision-making process or led to the emergence of a credible opposition to the sultan, such as council, nobles, or clergy. While we are not using the term accountability as we understand it in today's post-modern societies, the existence of any group whose stand must be taken into account by the sultan implies the existence of some degree of executive constraint. However, the third criterion requires further refinement. Any opposition that takes away the sultan's power in an area and replaces that with its own power does not promote accountability; this is merely swapping absolute power between two alternative executives. Therefore, when the decisionmaking mechanism changes and brings forth a credible opposition to the sultan, it is important to examine whether the opposition represents the views of a wider segment of the society or only those related to its own interests.
Our approach to the decline of the Ottoman Empire based on the mentioned institutions is decidedly different from the existing studies. Despite the danger of oversimplification, one can observe that the existing literature on the fall of the Ottoman Empire belongs to one of the two groups. One group emphasizes the role of the rising capitalism in the West First, where the Empire falls victim to the then emerging ways of the capitalist system (Kasaba, 1988;Wallerstein, Decdeli, & Kasaba, 1987). While even this group recognizes the weaknesses of the Ottoman institutions, the West and its capitalist ideals are identified as the ultimate culprit. The other group, represented by Pamuk (2004), praises the Ottomans for their pragmatism and flexibility where pragmatism defended the traditional order and flexibility enabled its longevity. This view suggests that despite the overall institutional weakness selective institutional reforms were effective in enabling the Empire to survive into the nineteenth century.
In this paper, we focus on three Ottoman institutions. First, the succession process is discussed, because the process through which the leader emerges is important to assure that the Empire will have an able leader. Second, we examine the power structure in the Ottoman government, in other words, the Ottoman elite. We identify who the elite were and how they ascended to their status. Third, the landownership-military-public finance triangle is discussed. This is an area of institutions that was significantly affected by external circumstances as well. In all three institutional realms, we focus mostly on the period of rise, in other words, the fourteenth through the sixteenth centuries.

Succession process and the imperial household
The process of succession through which the next leader emerges is an important institution for internal stability and external image of an empire. Independent of its implicit or explicit nature, even when there were rules for succession, they were not written in stone in any empire. Neither the absence of succession rules nor changes in them necessarily weakens an empire. In this section, we examine the environment in which the succession rules emerged in the Empire.
Especially during the critical era of the rise, the succession process in the Ottoman Empire was affected by the choice between Islamic and Turkic traditions of family life. This view puts the timehonored emphasis on the harem life (the sphere of Sultan's wives) in perspective. Often, powerful imperial wives are credited for the downfall of the Empire because of intense competition and intrigue among them; however, it was the institutional setup of the royal household that gave wives the incentive to take advantage of their environment, sometimes against the best interest of the Ottoman Empire. There was, however, once a Turkish identity beneath that of the Ottoman. In order to illustrate the significant changes in customs from the Turkic to the Muslim tradition, this section starts with background information regarding Turkic customs and continues with the Ottoman imperial household structure and succession.
Turks were nomadic people who lived in loosely connected tribes in Central Asia. Starting in the fourth century, as drought devastated their homeland, the Turkic tribes migrated in all directions. Those that moved westward were influenced by and influenced other cultures, predominantly the Persian and Arab powers of the day. Turks who had a shamanistic religion in their homeland began converting to Islam in the late seventh and early eighth century and transformed themselves in another 400 years as a legitimate Islamic power. 1 While there are no definitive explanations for the mass conversion of Turks to Islam in a relatively short period of time, alternative ideas have been suggested. For example, surrounded by Muslim powers plus the Byzantium, the Turks did not have much choice in this matter if they wanted to survive. Another possible explanation suggests that, fearing Turkish military valor, Arab powers provided economic and strategic privileges to Turks in exchange for converting to Islam (Quataert, 2000). A related reason is that the warrior ideology in Islam, jihad, must have appealed to the Turks, who had a warrior tradition of their own in their native Central Asian steppes. Additionally, the Ottoman ambition to become leaders of the Muslim world through caliphate may have helped to increase the Ottoman commitment to Islam (Inalcık, 1993).
As Arab power was winding down in the Middle East toward the end of the tenth century, a confederation of Turkic tribes was emerging under the leadership of the Seljuk Empire. The Turkish victory in the battle of Mankizert (Malazgirt) in 1071 opened Anatolia (Asia Minor) to Turks, at which time the Muslim identity of Turks had been established for at least two centuries. When the Moghuls defeated the Seljuks in 1243, beylicks, or principalities, appeared in Anatolia under different Turkish beys (leaders) (Quataert, 2000;Shaw, 1976;Vucinich, 1965).
Among the many Turkic principalities in Asia Minor were the Ottomans, whose origin goes back to Osman Gazi, son of Ertugrul Gazi who died in 1288. 2 As the Ottomans moved northwest toward the Byzantine in Asia Minor, Orhan Gazi (reign: 1346-1360), Osman Gazi's son, captured the strategic city of Bursa in 1326, which became the Ottoman capital. While the rise of the Ottomans as the leader of other Turkic beylicks cannot be explained by their numbers or military strength, their ability and willingness to make strategic decisions brought the Ottomans to the top. They dared to cross the Dardanelles and settled in Europe (Thrace), which was a bold move for the mid-fourteenth century. Orhan Gazi made other pragmatic and strategic decisions that would help strengthen his position against both the other Turkish beylicks and the Byzantium. Moving to Thrace and seizing Gallipoli won respect for the Ottomans among the other Turkish tribes. Orhan Gazi made use of his proximity to the Byzantine by helping Kantakouzenus with his enemies and marrying his daughter, which started a tradition for the Ottoman leader to take Christian-born wives. Additionally, the first Ottoman harem was introduced during the reign of Orhan Gazi (Feroz, 2003;Pierce, 1993;Quataert, 2000). Mardin (1969) suggests that even the early Ottomans strived to distinguish themselves from the other Turkic beylicks by substituting tribal norms that were based on kinship with those of established Islamic states.
By the fifteenth century, the Ottomans abandoned the bulk of their Turkic traditions and assumed a distinctive Muslim identity (Vucinich, 1965). There are three important differences in the dynastic family structure between the Turkic and the Ottoman tradition. First, polygamy did not belong to the Central Asian Turkish culture, although it was known even in the pre-Islamic Mesopotamian antiquity (Karduman, 2014). Starting in the seventh century, the diffusion of Islam in the Middle East and North Africa allowed polygamy (up to four wives) and changed the Ottoman dynastic family structure as well. Second, unlike the Turkic beys during the pre-Ottoman era, the Ottomans avoided marrying Turkic women. Women from the Balkans or Russia were preferred in the harem, a practice inconsistent with the Turkic tradition. The lineage of the ruler's mother, important in both ancient Turkic and early Ottoman tradition, became irrelevant by the fifteenth century as the Ottomans started a new tradition of marrying exclusively non-Turkic women (Inalcık, 1993). Third, in the pre-Islamic Turkic culture, women, especially the spouse of the leader, enjoyed more freedom as equals to their spouses as compared to the Western and Islamic societies of the time. The Ottoman culture, on the other hand, isolated women in the imperial harem (Yilmaz, 2004). 3 There were, however, exceptionally powerful women.
At a pinnacle in Ottoman history, about a century after Orhan Gazi, the Ottoman harem became a strong social institution during Mehmet II's (the Conqueror) fifteenth century reign. The Seraglio was a ruthless and isolated Machiavellian environment full of rivalry, intrigue, and influence, creating opportunities for murder, collusion, and corruption. Top women in the harem developed relationships with powerful government officials such as grand viziers to accomplish their ultimate goal of pushing their own son to the sultan position. The most notoriously powerful women emerged in the post-Mehmed II era as products of the established harem. Probably the most infamous of all wives was Hürrem Sultan or Roxelana (1502-1558), a Russian, for whom Sultan Süleyman (the Magnificent) changed the harem protocol. After she became Haseke Sultan (a title reserved for a woman who gave birth to the sultan's child), Sultan Süleyman married her, instead of sending her and the child to a province, according to custom. Hürrem Sultan bore most of Sultan Süleyman's children and used her influence on the sultan to have grand viziers and sons (shehzades) from other women killed. Another powerful wife in the harem, Kösem Sultan (Anastasia, 1589-1650), enjoyed a 30-year rule as de facto regent Valide Sultan (a title reserved for the mother of the reigning sultan), because her son Murad IV was in his early teens when he became Sultan. 4 She made alliances with janissaries as well as grand viziers and was eventually murdered by a harem rival. Safiya Sultan (1550-1605) arranged for her husband Sultan Murad III to be preoccupied with the harem's distractions and ruled under his name for 10 years. She allegedly had 19 competing shehzades killed to prevent any challenge to her son's ascension to the throne (Pierce, 1993).
Polygamy practiced by the Ottoman sultans made the introduction of implicit or explicit succession laws impossible. Succession of the next Ottoman sultan was a result of complicated internal intrigue and collusion between the women of the harem, higher-ranking government officials and the military. In the 1450s, during the time of Sultan Mehmed II, fratricide or the murder of the new sultan's brothers and sons became common and justified by the highest religious official, Şeyh-ül Islam, to strengthen the central authority of the sultan and to prevent civil or military uprising. 5 Sultan Ahmed I (r. 1603-1617) ended fratricide and started the period when brothers of the reigning sultan were kept in kafes, a sort of housearrest at the palace (Inalcık, 1993). Not only the brothers and children of the sultan, but the sultan himself became increasingly Seraglio-bound, especially following Sultan Süleyman, with a possible exception of Murad IV (1623-1640). Other sultans mainly relied on the talents and wisdom of grand viziers who entered the state service as devshirme (İnalcık, 1997;Quataert, 2000). Clearly, neither Turkic nor the Ottoman traditions implied a clear path of succession. While the designation of the son (through primogeniture or seniority) was an ancient Turkic-Moghul tradition, it was not always honored. Especially after the ruler's death, it was common for brothers (or uncles) in the ruling family to refute heir apparent (Inalcık, 1993). However, de facto succession outcomes were different among the Moghuls, the Seljuks, and the Ottomans. The Turkic-Moghul tradition implied that the leader shared power and territory with sons as well as with other male members of the ruling family. In this tradition, leadership required achievement and virtue as well as support of tribal chiefs through relations and negotiations. This gave the male members of the family the opportunity to display their leadership skills and win the respect of the elders as well as the general population. Therefore, the Turkic succession setup is nearly the opposite of the Ottoman tradition in which the harem setup produced a large number of potential heirs. While sometimes Ottoman shehzades were allowed to hold a provincial office, they were not allowed to have their own households. Essentially, their households were on loan from Istanbul. Mehmed III (r. 1596-1603) was the last prince to hold provincial office (Kunt, 2007). All shehzades except for the one who was to ascend to the throne were either kept in complete isolation from the rest of the world or killed. Neither fratricide nor forced isolation had a place in the Turkic, Moghul, or European tradition, unless the immediate family member committed a grave transgression. In contrast, isolation of heirs and fratricide were practiced in Byzantium as well as other Muslim empires (Inalcık, 1993).
The above discussion shows that the Turks went through a fundamental cultural transformation by converting to Islam, which introduced polygamy and therefore the harem into the imperial household. The power struggle among the women in the harem over who would be Valide Sultan is sometimes interpreted from the equality point of view. As opposed to the general belief that the Ottoman imperial women were segregated and powerless, this view states that the hierarchy and power struggle in the harem was similar to that among men (Pierce, 1993). We argue that these characteristics of the Ottoman imperial women do not imply superiority over their Turkic tribal counterparts, where the leader's wife was openly involved in social life and her male children (and other elite male members of the ruling tribe) had opportunities to show their leadership skills necessary to become the next leader without resorting to intrigue and power struggles.

Power structure: Identity of the Ottoman elites
In this section, we discuss the characteristics of the Ottoman power structure and question whether this structure was conducive for the eventual development of checks and balances in the Ottoman government. The use of the rather recent political term checks and balances may be questionable with respect to the fifteenth or sixteenth century political structure, as it would be difficult to find a kingdom or empire where the head of the state did not have absolute power. Therefore, we emphasize the term "conducive" in the above sentence and introduce two basic criteria. The first criterion is whether some groups or segments of the Ottoman society successfully represented a counterpoint vis-à-vis the Sultan and affected at least some of the executive decisions. The second criterion implies that the rise of another segment of the society against the sultan's absolute power should not only represent this particular segment's interests, but also those shared by the society at large.
In order to understand the power structure in the pre-modern Ottoman Empire, one needs to first investigate the nature of the ruling class around the sultan in an elementary fashion. Grand vizier, akin to Lord Chancellor in England, was the second most powerful person in the Empire. The Ottoman divan was also comparable to England's Council of Ministers inhabited by clergy and noblemen. Divan consisted of grand vizier, ministers who were military judges of Rumeli and Anatolia (kadiasker), the judge of Istanbul, the minister of finance, and the chief of the janissaries. Later the office of Şeyh-ül Islam (the supreme religious authority), Reis-ül-Küttap (the minister in charge of foreign relations), and Kapudan Pasha (admiral of the fleet) were added to divan. Comparable to other kingdoms or empires in the West, the executive and legislative power resided with the Sultan despite divan (İnalcık, 1997;Quataert, 2000).
While a superficial comparison of the Ottoman and a European empire's power structure shows similarities, there were important differences. When we ask the question as to who the members of divan were, the answer was unequivocally a Muslim who knows the Ottoman way through language and customs (Shaw, 1976). However, whether they were born as Ottomans was not important; in fact, being born as a Turkish-Muslim subject may have been a serious impediment for a person's career aspirations. With the exception of ulema (clergy), almost all members of the ruling class during the pre-modern Ottoman Empire were converts.
The devshirme system provided extreme social mobility to non-Muslim (overwhelmingly Christian) young males who were handpicked, provided a solid education, and assimilated into the Ottoman culture. Some of them joined the administration, even rising to the position of grand vizier and others joined the military (Quataert, 2000). Most devshirme came from the Balkans, specifically Bulgaria, Albania, and Bosnia. As early as the thirteenth century, during the time of gazis (warriors for faith), devshirme were serving alongside Ottomans and Turkic tribes in the Ottoman army. Some argue that Orhan Gazi (r. 1326-62) failed to build a standing army consisting of Turkoman tribes because they neither had the discipline of fighting in the infantry nor the willingness to be loyal to anyone else but their own bey (leader) (Feroz, 2003). 6 The practice of devshirme increased after Orhan Gazi until under Mehmed II (the Conqueror, r. 1451-81) and became the main source of infantry (Shaw, 1976). 7 There was a deliberate lack of power provided to subjects who were born as Turkish-Muslim in the Ottoman Empire. Muslim as well as non-Muslim (Christians, Jews, etc.) subjects of the sultan made up reaya, which was the tax-paying class. According to Mardin (1969), Turk was a derogatory term which implied the lower tribal culture. While there is no definitive explanation for the Ottoman aversion to the promotion of Turkish elites, there are some suggestions that may shed light on this issue. One of these suggestions is the competition among the Turkic tribes in Asia Minor (Anatolia) around the tenth century. At that time, the Ottomans were neither the strongest nor more populous Turkic tribe in Asia Minor, compared to the Karaman or Germiyan. However, the Ottomans were known as being pragmatic. Introducing a new strategy in confiscated territories, once they gained control over an area, they made it more stable and reduced taxes to win the support of their subjects. Even though the early Ottoman rulers were calling themselves gazis, they had no problem with hiring Byzantine generals and soldiers. While the Ottomans were fighting other Turkish tribes over supremacy in Asia Minor, they were also trying to expand toward the direction of the Byzantium and the Balkans. The ambitious Ottomans were not much liked by the Turkic groups in Asia Minor, which led to the continuation of tribal problems well into the fifteenth century. The Turkic groups went as far as appealing to the Moghul leader Tamerlane (Timur) who defeated Sultan Bayezid in 1402 near Ankara. Therefore, the longlasting distrust among the Turkic tribes in Asia Minor may explain the Ottoman aversion toward the Turkic identity (İnalcık, 1997;Quataert, 2000). Under these circumstances, enabling other Turkic tribes to have a place in the Ottoman administration would have led to a partly non-Ottoman Turkish aristocracy. Especially in a vast empire with active frontiers, devshirme with no baggage or community was preferred to Turkish aristocracy (Hathaway, 2005;Kafadar, 1995).
Denying Muslim Turks an elite status went as far as disallowing their marriage to the sultan's daughters and instead promoting such marriages to the initially devshirme upper class. Avoiding marriages within the ruling class to undermine the position of the Turkish elite started with Bayezid II, continued through Selim I and Mehmed II and finally strengthened under Süleyman the Magnificent when his grand vizier married an Ottoman princess. Needless to say, with such a family connection, the influence of viziers became substantial (Pierce, 1993). As discussed in the previous section, avoiding Turkish-Muslim marriages in the imperial household was also practiced with respect to the harem. While it is understandable that most slave girls in the imperial harem were of non-Turkish, non-Muslim origin, most Valide Sultans were also of non-Turkish, non-Muslim origin (İyigün, 2013;Toledano, 2014). The ethnic background of Valide sultans during the period  shows that only about 13 percent of the known Valide sultans were of Turkic origin. Additionally, Valide sultans of Turkic origin mostly concentrated during the pre-mid-fifteenth century period. Shaw (1976) suggests that the ethnic identity of the harem influenced the geography of the Ottoman expansion. As the Balkan ethnicity grew in the harem, Ottoman raids turned to North Africa and the Middle East. In fact, İyigün (2008, 2013) provides empirical verification for Shaw (1976) suggestion.
Our discussion shows that the members of the military and bureaucracy, which were the most important components of the state structure, were preferred to be devshirme, those who were taken from their families at a young age to be raised by the Ottoman State. An important exception to the prominence of devshirme was ulema that had to be Muslim-born. Ottoman sultans needed clergy to rule on laws as being consistent with sharia (Islamic law). Therefore, Sultans such as Bayezid I and Mehmed II had their brothers executed based on the fetva (legal opinion) provided by ulema, who declared fratricide as admissible for the sake of the Ottoman state's stability. As the sultan had to be accountable to sharia, ulema became a political force in the Empire and were increasingly mixed into seraglio politics (Feroz, 2003). Though kanun (sultan's law) and örf (customs) were also part of the judicial system, all facets of life such as inheritance, commerce, and finance followed the Islamic law (Kuran, 2011). Additionally, kanun became increasingly subject to fetva by Şeyh-ül Islam especially after sixteenth century, as ulema were trying to bring the sultan's authority under legal supervision. Ulema represented the Sunni brand of conservatism and became an impediment to government reforms starting in the seventeenth and throughout the nineteenth century (Tezcan, 2010).
The 290-year delay in the use of printing press in the Ottoman Empire can be used as an example to illustrate ulema's role in obstructing technological innovations. The first basmahane (printing house) appeared in Istanbul in the fifteenth century based on the know-how of Jews expelled from Spain. Until Ibrahim Muteferrika in 1727, the printing houses were largely managed by Greek, Armenian, and Jewish subjects of the Empire and books were printed in the languages of these subjects. Ibrahim Muteferrika, a Hungarian convert, introduced the first Muslim-managed printing press based on his successful argument with Şeyh-ül Islam of the time that important Islamic written work was about to get lost due to the shortage of calligraphers. However, because of the protest of the calligraphers, ulema eventually allowed only non-religious books to be printed (Göçek, 1987). To give an idea in terms of the use of the printing press in non-religious works, only 142 books were printed between 1727 and 1838 (Hanioğlu, 2010;İnalcık, 1997;Watson, 1968).
Even though the following example comes from a period beyond this paper's focus, which is ulema's reaction following the reading of the Edict of Gülhane (Tanzimat Fermanı) in 1839, it illustrates ulema's grip on the Empire's policies. Tanzimat aimed to support, among other things, the integration of non-Muslims into the Ottoman society by enhancing their civil liberties and granting them equality before the law throughout the Empire. It was customary to invoke God's blessings following the reading of a Hatt-i Hümayun (a note by the sultan); however, the ulema-selected preacher invoked the blessing and protection of Muslims without any mention of other religious groups or the proposed reforms. The financial independence of the ulema ended with waqfs (pious charities), after which even Şeyh-ül Islam became a civil servant (Tezcan, 2010).
In addition to ulema, janissaries were the other powerful group in the Empire whose revolt led to the dethronement of sultans and occasionally to regicide. Between 1589 and 1603, the janissaries rebelled six times, firmly establishing their power. Often janissaries demanded the head of a court favorite to end their rebellion. Following his defeat in Poland for which he blamed janissaries, Osman II (Young Osman) was planning to replace the institution of janissary with an ethnic Turkish standing army, aided by mercenary units. In 1622, the janissaries colluded with ulema and other higher ranking bureaucrats to kill the teenage Osman II to avoid losing money and status. Though this was the first regicide in the Ottoman history, between 1603 and 1703, six out of nine sultans' reign ended with dethronements. 8 The janissary revolts took place in a time when the Ottomans were fighting on two fronts, against the Safavids in the east and Habsburgs in the west (Tezcan, 2010). Needless to say, the Ottomans were not successful in either front. Therefore, ulema and janissaries were the only segments of the Ottoman society able to take a credible counterposition to the sultan. Interestingly, both groups were very near to the sultan, yet at times blatantly rebellious. In the next section, we will discuss the economic reasons that explain the rebellious attitude of janissaries. The ultimate question is whether the political power enjoyed by ulema and janissaries as well as the strong collusion between them can be interpreted as a sign of a society that was moving toward secular modernity. Kafadar (1995Kafadar ( , 2007 as well as Tezcan (2010) believe so and argue that the ulema-janissary cooperation served as an unwritten constitution to limit the power of the sultan. We, on the other hand, argue that having alternative powers confronting the sultan would have led to modernity only if these powers represented a broader segment of the society instead of being overwhelmingly focused on their own benefit. Therefore, powerful ulema and janissaries who were hanging over the sultan's head as the sword of Damocles do not pass the test of being conducive to opening the imperial decision-making process to wider segments of the society.
Compared to the European powers, one group that does not exist in the Ottoman society in its full capacity is noblemen. Kılıҁbay (1985) asserts that instead of asking the question as to why the Ottomans could not transform themselves into capitalism, we should ask as to why the Ottomans could not even get feudalism right. A major impediment to the emergence of noblemen or as later called in England landed gentry was Islam's mandate regarding inheritance rules. As opposed to the European inheritance laws that implied the rights of primogeniture, Islam requires the division of the estate among the by-Sharia recognized family members. The use (or misuse) of waqfs to prevent an increasing division of land implied that some landowners choose waqfs with small expenses, rolled the rest of the income into the waqf and provided the income generally to the eldest son. However, this practice was not enough to sustain landed gentry. Most important, the type of land ownership changed from the early modern to the expansion period. While mülk (private property) having no tax or military obligations was more common in the earlier Ottoman statehood, later miri (state land) dominated the land ownership structure (Kılıҁbay, 1985). Therefore, landowners were never elevated to the elite status along with janissaries and ulema.

Structure of landownership, Military, and public finance
In empires starting out in the middle ages, military structure was closely related to the patterns of land use. Granting the use of land revenues for services rendered was common in pre-modern states, as all empires needed the revenues from land to pay salaries of administrators and soldiers. Therefore, changes in the land use or tax collection patterns were potentially capable of wreaking havoc on public finances. When especially the Western empires approached the pre-modern period, the Renaissance and beyond, their military structure and patterns of land use drastically changed. The Ottomans' proximity to Europe made the rapid changes in the former's external environment even more challenging to deal with. Our primary interest in this section is consistent with that of the previous sections in that we aim to observe the Empire's responses to the internal and external challenges as well as to understand whether these responses were conducive to a more open and transparent Ottoman society with strong public finances.
Agricultural production was the main source of the Ottoman revenue, which was then allocated among the administrators (beys, viziers, etc.) and soldiers as salaries. Arable Ottoman land was state-owned and divided into tımars (fiefs). Because occupants had tenancy right only, fiefs were not hereditary and could be confiscated upon the tımar holder's death (İnalcık, 1997). Initially, the Ottoman tax system seemed to be working rather well. For example, at least compared to other European powers, the tax burden in the Ottoman territories was lighter than in neighboring European kingdoms and this fact was widely known at least in Central and Eastern Europe. Additionally, up to the seventeenth century, Ottoman tax officials conducted surveys on tımars with respect to population and land characteristics as well as productivity to assess the Empire's tax base. These surveys were then used in military and administrative appropriation decisions (Quataert, 2000).
The recipients of tımar revenues had to provide military service. In fact, the per-person tımar revenue was the amount of money necessary to maintain a cavalry man and his horse for a year. In the Ottoman Empire, such cavalrymen were called sipahi and were the most important military unit in the Ottoman army starting in the days of gazis. Sipahis fought during the war season which was spring and summer and managed their tımar holdings in peace times. In the allocation of tımar holdings, higher ranking commanders and government officials were provided with more valuable revenue units. Peasants who worked in the field were basically serfs.
The changing needs of the Ottoman army around the mid-sixteenth century led to a long and painful overhaul of the Ottoman agricultural system. Sipahis and janissaries comprised the most important components of the land forces in the Ottoman army. While sipahis were maintained by timar holdings, janissaries received their salaries from the sultan. Changes in the military technology such as increased use of firearms and the necessity of maintaining a standing army played a role for the declining relevance of sipahis. More importantly, by the mid-sixteenth century the Ottomans were focused on territorial expansion as well as maintaining the confiscated territories in a vast empire that reached from North Africa and the Middle East to the Balkans and central Europe. Therefore, the Ottoman army was in need of fire-armed infantry.
It was clear that introducing and maintaining an infantry-focused army needed cash infusions. Tax farming or the çiftlik system (iltizam) was introduced to provide the needed revenues for maintaining a strong infantry. In this system, the government held auctions at specific times and places for the right to collect the taxes of a district with arable land. Officials would determine the value of the land and the highest bidder would pay the state in cash. The successful bidder had the authority to collect revenues from the land as return on his investment. One characteristic of tax farming was that the bidders themselves often had financiers, which later complicated the system. In the late sixteenth century, the tımar system was largely abandoned in favor of tax farming and different types of leasing agreements (Quataert, 2000). 9 The çiftlik system introduced a number of intermediaries into agriculture, which led to increased complexity regarding claims and obligations. The new system was clearly detrimental for peasants. Under the çiftlik system, they faced the subcontractors of tax farmers whose main interest was to maximize tax returns, pay off debt and renew contracts. Peasants could not acquire any protection against them and had to plant according to the wishes of the çiftlik owners. Because peasants were still attending the previously tımar-land, it may seem that their life had not changed from the sipahiera. However, the average peasant household's economic and social status significantly declined because of their increased tax obligations under the çiftlik system in which peasants were paying taxes to at least three groups of people. First, the regular tithe and the animal tax had to be paid partly to the tax farmer and partly to the government. Second, local administrators and provincial governors imposed new taxes called tekalif-ı sakka to augment their incomes. Third, ayans (local notables) kept armed irregulars and made illegal demands for protection. Therefore, the countryside of the Empire developed a unique power structure consisting of ayans, armed irregulars, the members of the old prebendal class, and even kadıs (Islamic judges). Because kadıs often colluded with ayans, it was very difficult to bring forth çiftlik-related cases to the local judicial system (Jennings, 1979).
The central government was relying increasingly on its government officials to impose the new rules in agriculture; however, these officials either did not know much about the local conditions or became local powers themselves. The enlarging power vacuum at the local level was filled with provincial administrators and other local figures that were expanding their power and wealth. In later centuries, ayans were appointed as tax collectors (muhassıls) or as administrators of bureaucrats' estates (mütesellim). At the same time, some ayans and bureaucrats were becoming tax farmers. These local powers had the knowledge of their region and used it to take advantage of decentralization in agriculture (Quataert, 2000).
It seems that the çiftlik system changed the production patterns as well. In this system, the tax farmer had de facto ownership of land, which motivated çiftlik owners to focus on their profits. The market forces dictated that there had to be a change from grain to cotton and maize wherever possible. In later periods, such as the end of the eighteenth century, almost all export crops came from çiftliks. However, there is not much evidence for large-scale farming or increasing revenues for the İstanbul government. Most çiftlik owners were indebted to a number of private financiers who guaranteed tax farming contracts, provided owners supervised the operations and organized the buyers of cash crops. 10 Therefore, çiftlik production was positioned to reflect the interests of financiers and not those of çiftlik owners or the government.
As such, tax farming did not have to be ill-fated in the Ottoman Empire (Balla & Johnson, 2009). In a comparison of tax farming in France and the Ottoman Empire, they conclude that the Ottomans did not have the necessary judicial and financial structure to make tax farming work for them. French tax farmers organized themselves into a coalition of tax agents during the seventeenth century. Because the legal transfer of ownership was made relatively easy, they were able to build limited partnerships and find capital at a lower cost. Clearly, the Ottoman Empire was vast which would have made the transaction cost of a coalition prohibitively high. However, a system of multiple tax farmer coalitions would have emerged, if there were supportive legal and financial structures.
When it was understood that the çiftlik system was not going to generate the intended revenues, large-scale discriminatory taxation started. Coşgel's (2006) study of taxation in some of the Middle Eastern Ottoman territories during the sixteenth century suggests that the Ottomans often resorted to discriminatory taxation that gave the producers the incentive to switch to products without discriminatory taxes. The resulting inefficiencies in production contributed to the worsening of the already inequitable income distribution in these territories (Coşgel, 2006). The inflationary period 1556-1625 led to worsening public finances and more corruption. Along with tax collections, exports earnings especially from grains declined. While initially it was the contraband trade that reduced export earnings, later high-ranking Ottoman officials participated in illegal trade as well. Miri (state land) was being transformed into mülk (private property). Even the last tımar holders could sell a part of the tithe they were collecting in kind and benefit from favorable market conditions after they fulfilled their government obligations. Following large deficits in the late sixteenth century, avariz (emergency levies) and cizye (non-Muslim poll tax) increased. The avariz became permanent and was increased by more than 12 times between 1582 and 1681, and cizye 6 times between 1574 and 1691. Abuses in tax collection led to widespread dissatisfaction with government.
Additionally, large deficits and higher inflation rates led to various uprisings. Some ayans in the Balkans rebelled against Istanbul by withholding taxes. In the 1590s the Celali (bandits) revolts started in Anatolia, which continued until 1650 and undermined state authority. Sipahis who lost their tımars joined Celali and terrorized the country side. Macedonia and Bulgaria experienced their own brand of revolts perpetrated by Christian bands (Goffman, 2002;İnalcık, 1997). The Istanbul government was apparently at a loss especially when it came to dealing with janissary revolts. In 1589, the treasury debased the silver currency and janissaries revolted because of their declining standard of living and lower wages (İnalcık, 1997;Kasaba, 1988). To remedy their financial hardship, they enlisted their family members into janissary for more salary and became members in guilds to learn a profession. The intermittent wars against Austria from the seventeenth century onward necessitated the use of janissaries and mercenaries who caused trouble between wars. These wars with the Habsburgs not only ruined the Ottoman finances, they also made it clear to the Empire's rivals that the Ottomans were far from the days when they swept through territories (İnalcık, 1997).
Even Muslim reaya was concerned about the weak public finances and how far the government could go to increase revenues. As the Ottoman finances worsened, the fear of sultan's confiscation of private property led to the degeneration of waqf. Because of the religious nature of waqfs, there was assurance against their confiscation or arbitrary taxation by the government. Therefore, waqfs became a tax evasion tool for the Muslim upper class (Kuran, 2012;Mardin, 1969;Quataert, 2000).
Therefore, if we consider that the overhaul of the agricultural system aimed to modernize the Ottoman military structure, not only it failed to accomplish this goal but also wreaked havoc with the social and economic life of the country side and cities. Busy with worrying about government revenues and dealing with rebellions, the Ottoman administration was only minimally engaged in innovating military technology, an unusual strategy for a territorial expansion-minded empire. When facing the Austrian and Russian threat in 1729, the Ottomans turned to military innovation out of desperation. There are records of European experts who were invited to modernize the Ottoman army. However, the extent of the administration's commitment is questionable. Originally a French officer, a convert known as Ahmed Bey came to Istanbul, entered in Ottoman army service in 1731 and established a school of military engineering in 1734. He became a pasha (general) and humbaraci (bombardier). However, a couple of decades later when Baron de Tott came to Istanbul in 1768 to provide consultations for the Ottoman army, there was no evidence of Ahmed Bey's work (İnalcık, 1997). Selim III (r. 1789-1807) continued with modernization efforts and with the aid of French experts he introduced a new army (Nizam-ı Cedid). Resenting the growing French influence, the janissary-ulema coalition was joined by the ayans in 1805 and Selim III was deposed in 1807.
The discussion in this section raises a couple of important points. First, even though we observe the emergence of new power structures against the authority of the Istanbul government, the then newly empowered çiftlik owners, ayan, kadıs, and financial intermediaries did not lead to a more transparent society with improved equity. Second, while the Ottomans changed the agricultural and therefore the revenue system almost at the same time as in Europe, they did not have some of the European economic and judicial structures in place. Third, among other things, weak public finances made it difficult to keep up with the latest military technology.

Summary of discussions and new points to consider
When we look back at the selected institutional characteristics of the Ottoman Empire that are examined in the previous sections, we observe as the first common theme the aversion to Turkish elite. Clearly, the attempt to keep Turkish-Muslims from rising in the imperial household did not completely eliminate the threat against the sultan's sovereignty. The ulema-janissary collusion was perfectly capable of entering the scene as the only credible domestic power against the sultan.
In terms of the succession process, the Islamic influence made the harem structure possible where women of Turkic origin were avoided and the selection process of the next heir to the Empire was increasingly based on an intricate power struggle between the women of the harem and the high-level bureaucrats. It is highly questionable that this particular succession structure was capable of choosing the best heir for the Ottoman throne. Regarding the power structure, except for the ulema, the Ottoman sultan was surrounded by Ottoman subjects with devshirme background. The ulema were the only elite group with a Muslim background. When we connect the Ottoman power structure with landownership, we observe that there was no private property and no rural noblemen. In Europe, the presence of the landed gentry initiated power sharing between kings and noblemen. While private landownership coupled with mostly primo-geniture inheritance laws in the West enabled the country estates to remain intact, the absence of both in the Ottoman realm made the central government responsible for managing arable land through a highly centralized structure. The inability of the Istanbul government to centrally manage a vast amount of arable land led to the decentralizing attempt of tax farming, which in turn significantly decreased government revenues. We conclude that the Ottoman decisions with respect to succession, the structure of elites, and landownership-related issues were not conducive to building a more transparent society where the voices of larger segments of the society could have been heard.
In the case of the Ottoman institutions examined in this paper, the sum is more than its parts. In other words, significant negative externalities emerged from these institutions. Even though Ottoman sultans tried to avoid the emergence of the Turkish-Muslim elites, they still had to deal with two other significant powers that often colluded against sultans. The ulema-janissary coalition created a persistent headache for sultans as the former kept the Istanbul government busy with internal affairs while the external world had become increasingly unknown to the Ottomans. The Ottomans were not part of the age of discovery, mercantilism, capitalism, or industrialization. Taking the age of discovery as an example, the European expeditions to new continents relied on new navigation and shipbuilding technologies. In other words, there were significant positive externalities associated with expeditions from which the Ottomans could not profit. As a rare strategic move toward Europe, the Ottomans supported the Calvinist movement, probably in the hope that the Reformation would weaken Europe. Another example is the sixteenth century cooperation between the Ottomans and the French, which brought together Süleyman the Magnifient and Francoise I of France against their mutual enemy Charles V, King of Spain and the Holy Roman Emperor (Isom-Verhaaren, 2013). Still, considering the Empire's fixed target on Europe and particularly Vienna, the Ottomans were surprisingly incapable of interpreting the political, social, and technological changes in Europe. 11 While the Ottomans were trying to fight on three fronts (the ulema-janissary coalition, the Safavids in the east and the Russians in the north), there was a new world economic system being built of which the Ottomans were not a stakeholder but in which they were eventually integrated by the European powers as a marginal participant. While often the blame is put on the capitalist system for the marginalization of the Ottoman Empire in the world economic system, the Empire did not encourage commerce among its Turkish-Muslim subjects in that commerce was looked down as degrading in the Ottoman society (Göçek, 1996). Therefore, this paper's examination of the selected Ottoman institutions highlights the areas that should receive attention in future research. These areas include the avoidance of Turkish elites, the society's scorn for commerce and the Ottoman worldview during the fifteenth and sixteenth centuries.
There may be lessons that can be drawn from the examination of the selected Ottoman institutions for modern day governance. 12 First, a general lesson is that institutions are only as good as the incentive structure that they provide. Replacing the previous institution with a new one could lead to inefficient outcomes, if the new institution provides perverse incentives just like the previous one. Therefore, rather than either preserving the tradition or initiating a change for the sake of change, policy-makers should think about the incentives that are provided by the current and the proposed institution. Second, a lesson specific for the Ottoman Empire is that during the conception of the later reforms (seventeenth through nineteenth century) the motivation was to recapture the glory of the previous centuries. It seems that the glorification of the rise of the Ottoman era is somewhat present in today's Turkey (Evrensel, 2013). Such tendencies cause concern, because naïve admiration overlooks the fact that a rise energized by sheer force could not last in the face of opponents that transformed themselves in the realms of military, trade, and finance. Attempting to go back in time instead of trying to understand the forces of current times may not be a successful strategy.