Most state employees, like police officers, are rewarded with salaries, offered a career ladder of graded appointments with progressively higher remuneration, and follow standardized guidelines while being monitored by other employees above them in the public sector hierarchy. But there are exceptions to this. Sometimes states do not directly hire and monitor employees and instead operate indirectly via private contractors. In the United States, for instance, businesses contract with the state for the management and operation of some prisons.
Private contractors are only in charge of a tiny fraction of government duties today, yet the privatization of state duties was extremely common just a few centuries ago. In 1500, all the state administrations in Europe were organized this way. Let’s call this type of state organization “patrimonial.” Patrimonial administration was particularly ubiquitous in the tax administration. Monarchs sold tax farms —the prerogative to collect specific taxes— to private companies in return for a down payment in cash. Imagine if Amazon, not the IRS, collected taxes. Any revenue above the down payment and the administrative cost of collecting the tax was the company’s profit from the tax farm. The privatization of tax collection had a major disadvantage: tax farming meant that rulers had to surrender some degree of control over public policy because the assent of tax farmers was necessary to implement fiscal policies.
Around 1650, a radical political transformation began to take place. As war became more expensive, rulers across Europe started to attempt to modernize their administrations to make tax collection more efficient. Rulers sought to replace patrimonial officeholders with professional, centrally appointed, and salaried bureaucrats. In some places, like England, early administrative reforms succeeded. The Excise Office, in charge of collecting indirect taxes, was the most bureaucratic organization in Europe. In other countries, like France, reforms failed, and the state remained patrimonial at least until 1789. Before the Revolution, state offices were almost all for sale, and most state employees were independent agents contracting with the monarch on equal terms. Why did early administrative reforms succeed in England and not in France
Why did some administrative reforms succeed while most failed?
The success of bureaucratic reforms demanded a set of permissive conditions be met (as demonstrated by Ertman’s comparative historical analysis). For state bureaucracies to be viable, rulers needed: (i) the ability to compensate or defeat, during administrative reforms, entrenched officeholders with a vested interest in the survival of patrimonialism, and (ii) a high supply of human capital to recruit into the state administration for a wage.
In a book project, The Protestant Road to State Bureaucracy, I explain the asymmetrical historical development of these administrative conditions. I show that the extent to which these conditions were met by the seventeenth century depended on whether states experienced a Protestant Reformation in the early sixteenth century.
The Reformation was a complex and widely studied process (for example, here and here). For present purposes, let’s consider it a package of reforms that included the confiscation of the Catholic Church’s assets and the establishment of state-religions. Through these policies, the Reformation acted as a catalyst that magnified tiny initial differences across polities and accelerated the bifurcation of Catholic and Protestant polities along two different political development paths. The Protestant path promoted the blossoming of administrative conditions that facilitated bureaucratization more than a century later.
The Protestant Reformation involved massive ecclesiastical confiscations. In England, monasteries owned over a third of all the land, and in the span of five years, from 1536 to 1541, all ecclesiastical land was transferred to the crown during the Reformation. These confiscations affected state development paths through two slow-moving mechanisms.
One mechanism was the erosion of the link between war and the sale of offices. Protestant rulers used the windfall from ecclesiastical confiscations to partly finance the religious wars that soon erupted. Their Catholic adversaries, in contrast, financed those wars with a traditional mix of tactics that included selling offices. Thus, war fueled proprietary officeholding in Catholic realms, but that link was broken in Protestant lands. In the long run, proprietary officeholders used their offices to deepen their privileges, and this led to institutionalized venality in Catholic polities and to larger stocks of proprietary officeholders who resisted bureaucratic reforms.
A complementary mechanism was the reallocation of human capital from religious to secular skills. The declining power of the Catholic Church because of confiscations reduced the number of plum jobs in the clergy. This profoundly affected elites’ educational decisions. Anticipating where the jobs were, Protestant elites had their university-educated children take secular, rather than theological, degrees. In the long run, this shifted the supply curve of “useful” human capital upwards in Protestant places, making it less expensive to recruit talented bureaucrats for a wage.
Quantitative mechanism-based analysis
What is the evidence for this theory? The literature on state-formation has typically been narrative and qualitative. Departing from this traditional approach, I offer various pieces of quantitative evidence. Although no single exploration provides definitive evidence of causal effects, collectively they document some robust patterns that are consistent with the Protestant Reformation having pushed states into a political development path that facilitated administrative reform in the late-seventeenth and eighteenth centuries. One can think of the empirical approach as “quantitative process tracing.”
Naturally, I cannot do justice in this blog post to the wealth of data and analyses that are part of the book project. I will focus here on some main pieces of evidence. The empirical “plan of attack” begins by establishing a stylized fact. Analyzing the sample of the 13 major territorial states in western Christendom in 1789, only Protestants had begun to build bureaucratic administrations, and all Catholic states were patrimonial (see Figure 1).
I also examine continuous measures of different aspects of bureaucratization, based on expert surveys to political historians, and show that Protestant states in 1789 relied to a greater extent on salaried workers and had more meritocratic recruitment. The association between Protestantism and markers of early bureaucracy persists after controlling for existing explanations for bureaucratization (the intensity of early geopolitical competition, urbanization, and absolutism) and after excluding each individual polity (and each pair of polities) from the sample.
As it is hard to make conclusive inferences based on this small sample, I complement the cross-sectional results with evidence for the operation of two mechanisms linking the Protestant Reformation to the development of administrative conditions that facilitated state reform:
- Mechanism 1: The Reformation made states less likely to sell offices.
- Mechanism 2: The Reformation led to the reallocation of human capital.
In the first step of my mechanism-based analysis, I examine the association between the Reformation and venality. I begin by showing that, in a cross-section of European polities between 1500 and 1789, the sale of offices as a financial instrument was only widespread in Catholic polities (see Figure 2). The association between Catholicism and the sale of offices is statistically significant and persists after controlling for alternative explanations; and after iteratively excluding each polity and pair of polities from the sample.
What explains this pattern? Catholic states sold more offices to finance war, and Protestant states could avoid office-selling because they could confiscate ecclesiastical assets during emergencies. To support this mechanism, I show that in Catholic France proprietary office-holding increased significantly upon the start of new wars in the period 1480-1789. In Protestant England, in contrast, the sale of offices never became widespread. Analyzing annual data on the royal sale of assets from 1485 to 1640, I present two findings. First, the English crown received a higher proportion (and total) revenue from the sale of assets when England was fighting wars. Second, a formal statistical test for an unknown structural break in the time series identifies the year 1537 as the year of the break. This is just a year after the English Reformation confiscated all the Catholic monasteries in the kingdom.
In the second part of the mechanism-based analysis, I document how the Reformation led to the reallocation of human capital from religious to secular skills. Using individual-level data on all the graduates from Oxford from 1500 to 1714 compiled for this paper from a biographical dictionary, I show that after the English Reformation Oxford graduates became more prone to pursuing careers in law (as barristers) relative to careers in the clergy (see Figure 3). This reallocation was more pronounced for students who were originally from counties in which wealthier monasteries were confiscated.
I complement this national-level finding with subnational level evidence on the growth of endowed grammar schools —the British equivalent of the American preparatory school. For this, I create a novel district-level dataset of all endowed grammar schools in England. In a difference-in-differences setting, I show that the confiscation of a monastery at the district-level led to the foundation of more establishments of secondary education.
The most obvious theoretical contribution that this book makes is to the literature on comparative state formation. This book also contributes to a rapidly growing literature, mostly in economics, on the effects of the Protestant Reformation —a literature that has mostly focused on economic effects, neglecting the significant political consequences of Protestantism (but see here).
At the most abstract level, my work illustrates a class of historical mechanisms through which “causes” lead to effects only in the long run. In this type of mechanism, a shock sets in motion slow-moving transformations that incrementally generate conditions that only eventually allow agents to produce change. In the case that I examine, the Reformation broke the link between venality and war and led to the reallocation of human capital from the secular and religious sector. These transformations ultimately made it possible for administrative reforms to succeed. But change only took place centuries later when war —a productive condition— gave rulers incentives to reform their state administrations.