Chapter 1: Introduction

The introduction of the book starts with the motivation and the main puzzle that the book aims to address: Why do ethnically and religiously more diverse countries have lower fiscal capacity and taxation levels? In this book, I offer an answer to this question that bridges the literatures on state building and identity politics, emphasizing a political economic approach to the population's illegibility. Hence, I complement the existing literature's emphasis on taxpayer-preference-based explanations with an emphasis on a tax-administration-based explanation, where the difficulties in obtaining information from diverse populations make it more difficult for the state to build an efficient tax administration . The chapter discusses the existing literature and identifies the gaps in the literature on state capacity and taxation in diverse contexts with weak states, before it explains how this book aims to fill these gaps in the literature by offering a theory and testing it with empirical evidence.

Chapter 2: A Theory of Fiscal Capacity Building Under Indirect Rule and Diversity

Chapter 2 presents the theoretical framework. I first define the concepts that I use. I offer a clear definition of what I mean by fiscal capacity and introduce the two configurations of diversity I use throughout the book: dissimilarity, which means that the population has a different ethnic or religious identity from the core/dominant group, and heterogeneity, which means that there are higher numbers of uniquely dissimilar groups. Next, I identify two necessary factors for strong fiscal capacity to exist, high legibility of the population (or high informational capacity of the state) and high bureaucratic capacity. After I theorize how diversity contributes to illegibility and undermines bureaucratic capacity, I discuss why diversity undermines the bargaining power of the central state against local intermediaries in places with indirect rule. The next subsection discusses how the costs of investing in fiscal capacity should be higher under a more diverse population due to lower legibility, weaker bureaucratic capacity and lower bargaining power that diversity causes. Next, I discuss why interstate wars are critical events for evaluating these arguments. Having laid out the theoretical framework, I describe the main case of the book, the Ottoman Empire and formulate two sets of hypotheses I test in the book. The first of these expects diversity to undermine fiscal capacity building and the second expects diversity to increase costs of investment in fiscal capacity building. I conclude the chapter by delineating the scope conditions of the theory.

Chapter 3: Research Design and Empirical Strategy

Chapter 3 serves multiple purposes. First, it engages in a discussion of how best to answer the research questions in the context that I study. I first revisit the hypotheses and other expectations of my theoretical framework. Next, I justify the case selection (the Ottoman Empire) given the research questions and the scope conditions of the theory. Then, I introduce the main quantitative dataset of the book that I will use in multiple chapters, and introduce the main dependent and independent variables that measure diversity (Ethnolinguistic Similarity and Ethnolinguistic Fractionalization) that I use throughout the book. After introducing my qualitative data, which is mostly built on my original archival work in the Ottoman archives in Istanbul, I discuss how to best address the research question with these data and how to integrate these historical quantitative and qualitative data. I conclude the chapter with a detailed discussion of potential biases in archival research in historical work and how combining large-N quantitative research with in-depth qualitative analysis of archival documents can mitigate any potential biases.

Chapter 4: Background and the Problem in the Comparative Context

In this chapter, I describe the background of the problem, and situate the case of the Ottoman Empire in the historical as well as the comparative context. After I describe the problem of acquiring information in diverse societies, I go on to describe how indirect rule is often a consequence of such information problems, but also how indirect rule further undermines the state's capabilities. I also describe similar informational problems in the cases of other diverse countries, the Habsburg Empire, the Romanov Empire, and the Ancien Regime France in order to discuss how these cases were similar or different than the case of the Ottoman Empire and what insights they can offer for the book's theory.

Chapter 5: Diversity Undermining Legibility: Local Knowledge, Informational Capacity, and Administering Censuses

Chapter 5 focuses on how the population's diversity undermines its legibility to the state. It tests the first step of the argument proposed in the paper. First, I discuss the two measures of legibility (or informational capacity) I use in this chapter: census completion and age heaping. In the first empirical section, I demonstrate that in the Ottoman Empire, more diverse areas had lower rates of census completion, indicating that they were less legible. Next, I complement these large-N analyses with an in-depth qualitative analyses of Ottoman archival documents that demonstrate how the state agents had trouble collecting information about more diverse populations. Finally, using several cross-country and subnational datasets, I show that higher levels of diversity are associated with lower levels of legibility measures. These measures of legibility, which I obtain from secondary sources, include age heaping, aggregate informational capacity, statistical capacity, and the extent and quality of cadastral surveys.

Chapter 6: Diversity Undermining Bureaucratic Capacity and Bureaucratic - Administrative Reform

Having tested the first step of the argument, that diversity undermines the population's legibility in the previous chapter, in Chapter 6 I test the second step of the main argument of the book, that diversity impedes bureaucratic capacity. I use several large-N datasets to demonstrate that the population's diversity undermines bureaucratic capacity building. The first of these datasets is a dataset of governors from the late Ottoman Empire. Using this dataset, I first show how the ethnic identities (and thus the languages they speak) of the governors restrict their appointment patterns based on the diversity of the local population. Next, I show how the reassignment of governors is restricted in more diverse places. These findings indicate that a diverse population prevents the optimal assignment of bureaucrats and undermines bureaucratic capacity. Using another dataset on bureaucratic-administrative reform in the Ottoman Empire, I demonstrate that reforms were less likely to be successful where the populations were more diverse. I complement these large-N analyses with an extensive qualitative analysis of Ottoman archival documents. Finally, I use a cross-national dataset on bureaucratic capacity and demonstrate that more diverse countries have weaker bureaucratic capacity.

Chapter 7: Taxing Diverse Populations: Evidence from Tax Revenue and Fiscal Capacity Data

Chapter 7 tests the first set of main hypotheses of the book regarding how diversity undermes fiscal capacity building. I first introduce my local-level fiscal revenue dataset. Using this dataset in the Ottoman Empire, I demonstrate that fiscal capacity building efforts were less successful in areas where the populations were more diverse. Another implication of these results is that the core/dominant group, which was in a position of power and made policy, had to tax its own members because it was too costly to tax other groups due to issues of illegibility. I conduct several robustness checks for this analysis and rule out several alternative explanations that can explain the patterns I identify. Then, I utilize individual-level property surveys from the Ottoman Empire to confirm that members of the core/dominant group had higher tax burdens compared to minorities. In the next section, I engage in a detailed analysis of Ottoman archival documents to provide further evidence that the Ottoman state had difficulty taxing diverse populations. Finally, I use various cross-national and subnational datasets from other contexts, and demonstrate that diversity is associated with lower fiscal capacity and lower tax revenues. In these analyses, I use different measures of fiscal capacity. In this section, I also demonstrate that more diverse states are more likely to rely on indirect taxes and resource revenues, while less diverse states are more likely to rely on direct taxes, which require stronger informational and bureaucratic capacity.

Chapter 8: Diversity Increases Costs of Investment in Fiscal Capacity and State Capacity

In this chapter, I evaluate the hypotheses that expect diversity to increase the costs of investment in fiscal capacity building. In order to systematically assess these hypotheses, I use two quantitative datasets and then complement this with a qualitative analysis of archival data.. First, with a cross-sectional dataset on local-level expenditures of the central state, I demonstrate that the expense-to-revenue ratio of the central state's budget is higher in more diverse Ottoman provinces. I interpret this as evidence that the state had to invest more in a more diverse administrative unit to be able to extract the same amount of revenue. Next, using the same dataset, I analyze the types of expenditures of the state. I demonstrate that where the populations are less diverse and it is easier to acquire information, the state is more likely to focus on expenditure items that are more likely to sustain the state's control and increase fiscal capacity in the long term, such as infrastructure and education. By contrast, in more diverse places, the state's expenditures are more heavily focused on security, which can help coercive tax extraction and sustain control in the short term, while not having much benefit in the longer term. Then, I utilize an original dataset on bureaucrat salaries. I show that the state had to pay higher wages to the bureaucrats in more diverse provinces, indicating that the costs of building bureaucratic capacity are higher under diversity. Finally, I conduct an in-depth qualitative analysis of documents from Ottoman archives to complement the large-N analyses.

Chapter 9: Indirect Rule: Diversity Changes Bargaining Power in Favor of Local Intermediaries Against the State

An important aspect of the argument in this book is that diversity reduces the central state's bargaining power against local intermediaries. Primarily relying on in-depth qualitative analysis of archival documents, I provide evidence in favor of this argument. I demonstrate that where the populations were more diverse, the Ottoman Empire had to rely more heavily on indirect rule through intermediaries. I also demonstrate how this reliance increased these intermediaries' bargaining power against the state and in turn, the intermediaries were able to more successfully resist state building attempts, undermine population counts, and undermine state capacity. Finally, I show that the heavy reliance on these intermediaries for taxation in their areas of influence undermined the state's fiscal capacity in those areas.

Chapter 10: Evidence from Tax Farming [under development]

This chapter is closely connected to Chapter 7 on state's fiscal capacity, and Chapter 9 on indirect rule and bargaining power between the central state and local intermediaries. Tax farmers were often local intermediaries who had an advantage compared to the state's agents in collecting taxes from their areas of influence. Possessing better information and networks in their locality, they could collect taxes more cheaply compared to any outsider agent that the state could assign from the center. In this chapter, I utilize tax farming data from different localities in different time periods in the Ottoman Empire to demonstrate that the state was more likely to delegate tax collection to such local intermediaries in more diverse places. Finally, I demonstrate how the onset of war changes tax farming contracts as it often shifts the bargaining power between the state and the local intermediaries.

Chapter 11: Comparative Evidence [under development]

This chapter complements Chapter 4, where I situated the problem in the historical and comparative contexts. I consider examples from other countries to discuss how diversity was a critical factor in shaping fiscal capacity building and tax policy. I will pay particular attention to the late Habsburg and Romanov Empires, France, the United Kingdom, and Spain between the 17th and early-20th centuries. Countries with different levels of diversity and fiscal capacity, examples from these contexts provide further support for tracing the link between diversity and taxation. Finally, I conduct a case study of 20th-century Tanzania, which early on adopted the Swahili language as a lingua franca. I utilize process tracing to show how the adoption of Swahili enabled better acquisition of information for the state's agents and how it shaped the Tanzanian state's development. While some of the sources I use in this chapter are primary sources, I mostly rely on secondary sources.

Chapter 12: Homogenization and Nation-Building

This chapter focuses on the aftermath of the empires and examines to what extent ethnic/religious homogenization and nation-building projects are responses to the issues of illegibility and the resulting weak state capacity in diverse contexts. I pay particular attention to the ethnic and religious homogenization attempts during the late Ottoman Empire, which sought to assimilate, or ethnically cleanse the minority populations under its rule. I then turn my attention to the early Turkish Republic, and discuss how the ethnic homogenization attempts during the nation and state building processes of the early republic were related to the inefficient state that the republic had inherited from its predecessor. Among the homogenization projects I will examine are the forced population exchange between Greece and Turkey (1922), the pogroms against Jewish populations of Thrace (1934), the Wealth Tax against non-Muslims (1942), the forcing of non-Turkish minorities to speak Turkish in public starting in the late 1920s, and the pogroms against non-Muslim, especially Greek, minorities (1955). I also pay significant attention to the economic aspects of these homogenization projects and how they aimed to create a ‘national bourgeoisie’ by transferring wealth from the minorities to the Turkish-speaking Muslim populations.

After discussing the case of the late Ottoman Empire and Turkish nation building and homogenization, I situate the Ottoman/Turkish case in a comparative perspective, by examining other nation-building projects in comparison to the Ottoman/Turkish case. I examine other core/dominant group nationalisms in the post-imperial and post-colonial settings in newly independent states. In addition to this, I also discuss how the lack of investment and lack of public good provision in diverse areas may have fueled separatist minority nationalisms.

To link these discussions to the broader picture where polities start as empires and end up as nation states, I rely on Tilly's hypothesized trajectory from empires to national states, and how diversity of the populations has propelled rulers to follow a coercion-intensive path of state development. Tilly explained that empires and city states eventually converged into the mid-sized state form, the national state, as the national state was more efficient in resource extraction and was more successful in fighting wars and surviving in the state system. Empires, by definition, are larger and more diverse compared to national states. Hence, the findings in this book add an extra layer to the discussion of the efficiency of the national state in extracting taxes in Tilly's framework. The national state was more efficient also because it was less diverse. This is likely one reason why the large and diverse Habsburg, Romanov, and Ottoman Empires eventually fragmented into many smaller and more homogeneous national states. It was too costly to hold them together as heterogeneous units with a small core/dominant population.

Chapter 13: Legacies [under development]

In this chapter, I consider the legacies of past diversity and variations in state capacity on contemporary outcomes in the Turkish Republic. For identification purposes I focus on the areas of the Ottoman Empire that are situated within the borders of the modern Turkish Republic. I demonstrate that where the population was ethnically and religiously more diverse during the Ottoman Empire's rule, state capacity and public good provision are weaker today.

This analysis is feasible due to the fact that the areas that were more diverse during the late Ottoman Empire are not necessarily diverse today. Hence, I can argue more convincingly that the patterns I find can be attributed to the legacies of past levels of diversity, rather than today's diversity. In order to better identify the legacies of the past, I consider outcomes from the early Turkish Republic, starting from the 1920s. However, the contemporary statistics are more detailed and more reliable compared to the statistics from early and mid-20th century so most analyses focus on the contemporary period (late 20th and early 21st centuries). Finally, I also demonstrate with contemporary survey data that past levels of diversity in the areas that people live shape their attitudes towards the state, taxation, and public good provision today.

In the final part of this chapter, I discuss how the legacies of the lack of investment in state and fiscal capacity, due to diversity, can also be observed in other post-imperial and post-colonial peripheries.

Chapter 14: Conclusion

In concluding in Chapter 14, I revisit the puzzles that the book aimed to address, and reevaluate the book's theoretical contribution in light of its empirical findings. This is followed by an extensive discussion regarding the generalizability of the theory based on the scope conditions I identified during the development of the theoretical framework. Next, I discuss potential areas for future conceptual and theoretical developments before discussing best practices for data collection in similar historical work in other contexts. Finally, I examine the potential policy implications of the findings in the book, particularly emphasizing the need for an ethnically and religiously more diverse bureaucratic apparatus, more pluralism, and more power-sharing in diverse states.