Paintings, Persecutions, and Political Development

by Noel Johnson and Mark Koyama

A Mysterious Masterpiece

Among the paintings in the Royal Collection of the British Monarchy is a winter scene by Pieter Bruegel the Elder titled “The Massacre of the Innocents”. As often is the case with Bruegel, it is a masterful recasting of a traditional theme into a contemporary setting. In this case, taking the story in St. Matthew of Herod’s attempt to kill the infant Jesus by ordering the slaughter of all children in Bethlehem below the age of two and locating it in a Flemish village in the mid-sixteenth century. This was not an arbitrary choice. The Reformation, which we traditionally date to 1517 and the actions of Martin Luther in Wittenberg Germany, was in full swing in the Spanish Netherlands. The official monopoly held by the religion of the ruling Spanish Habsburgs, Catholicism, was under threat. Protestantism was coalescing with a nascent Dutch national identity and the Spanish Crown was having none of it. In 1565 Philip II reintroduced the death penalty for heresy in the Netherlands. In 1567 he sent the Duke of Alba, along with 10,000 troops, to the Low Countries via the Spanish Road to suppress Dutch Revolt and restore Catholicism as the sole legitimate religion.

The painting, which was completed at the end of 1567, presumably shows the ransacking of a Protestant village by the Duke and his soldiers. There are some idiosyncrasies, however. First, the soldiers appear more concerned with killing livestock than with murdering innocent children. One soldier carries away a dead goose (see the top figure below), another is seen skewering a boar (the third figure below). All while the despondent villagers look on. Some other perplexing scenes show women crying over packages or desperately trying to prevent soldiers from taking pots and other common household items. What could be going on here?

Detail from altered painting
Detail from unaltered painting


Detail from altered painting
Detail from unaltered painting

In 1998 this mystery was solved when, during routine conservation work, it was discovered that many of the original scenes in the version of the painting in the Royal Collection had been painted over. Or, more accurately, the painting had been sanitized so that the most gruesome and inhumane acts of the soldiers were censored out. This was verified using infrared imaging of the original painting, but we can also see the changes by comparing the version in the Royal Collection with later copies contained in other museums. The first two figures above show an example of the changes. The soldier in the altered painting in the left panel appears to be stealing the dead goose while the actions of the villager—apparently dragging his child toward the soldier, makes little sense. In the unaltered version shown in the right panel, the goose is revealed to be a child, most likely a son who the villager is now, tragically, attempting to convince the soldier to spare in exchange for the life of his daughter.

In the details shown in the bottom two figures we see that the mysterious packages in the altered version of the painting are, in fact dead babies. And the “pot” being clutched by the villager is, in fact, her baby who is being ripped away by a soldier. In the background of the detail the left panel shows soldiers spearing farm animals to death when, in fact, the original scene reveals that they are, in fact, massacring babies.

Why was Bruegel’s original painting censored? To understand this we need to understand more about the historical context. We know that by 1577 the painting was in the the collection of Rudolf II, the Habsburg Holy Roman Emperor, King of Hungary and Croatia, King of Bohemia, and Archduke of Austria. We also know that in that year he ordered the painting to be altered. We don’t know exactly why Rudolf did this. Given that he was a famous patron of the arts and a frequent dabbler in the occult sciences, however, it seems unlikely that it was mere squeamishness.

A more likely explanation is that Rudolf was trying to protect the legitimacy of his rule. The seat of Rudolf’s power was Bohemia, and in sixteenth century Bohemia, like in most of Europe, rulers derived their legitimacy from religion. Up until 1517, this meant the Catholic Church. In exchange for the Church giving legitimacy to a ruler via the pulpit and, more often than not, law, rulers promised to enforce the monopoly of the Catholic Church over religion. This is the main reason why Philip II used so much force in the Netherlands against Protestants. By choosing the Protestant religion, his subjects were also rejecting the source of his legitimacy to rule in the Low Countries. And, indeed, after eighty years of war, the Spanish Habsburgs were finally defeated and forced to recognize the Dutch Republic as an independent state.

Rudolf II faced a similar situation in Bohemia that his maternal uncle, Phillip, did in the Spanish Netherlands. Bohemia was a region where dissent from the Catholic faith had long been a fact of political life. In 1575 Rudolf’s father, Maximillian II, had been presented with a decree called the “Bohemian Confession” by the national assembly. According to this document the main religious sects (Catholic, Lutheran, Ultraquist, and Unitas Fratrum) in Bohemia were all granted the right to worship as they pleased. Maximillian died the next year, but his son and successor, Rudolf, was under tremendous pressure from the very beginning of his reign to confirm the acceptance of the principles of the Confession. Like Philip in the Low Countries, Rudolf was faced with a choice—fight to eliminate religious dissent so as to maintain the existing equilibrium based on religious legitimation by the Catholic Church. Or, widen the bounds of what was considered legitimate belief so as to avoid conflict. In the end, Rudolf chose to avoid conflict by issuing the “Letter of Majesty” in 1609 which formally endorsed the principles of the Confession and granted a virtually unprecedented amount of religious freedom in Bohemia.

The Duke of Alba
Unaltered painting
Altered painting

It should perhaps come as no surprise, then, that sometime during the first year of his reign, with his subjects asking themselves if their new Habsburg ruler will choose to undermine his legitimacy by abandoning Catholic hegemony, or, smash dissent like his uncle in the Low Countries, Rudolf chose to “censor” Bruegel’s explicit portrayal of the consequences of the latter. This interpretation is bolstered by the fact that the original painting had explicit references to Habsburg culpability in the massacre and that Rudolf had these erased. For example, the top figure above shows a portrait of Philip’s commander in the Netherlands, the Duke of Alba (also known as the Butcher of Flanders). Art historians have suggested that the commander in the lower left of the original painting, with the characteristic droopy eyes and beard of the Duke, was the man himself (see the middle figure above). In the altered version of the painting the beard changes color, is diminished, and the eyes become less dark and hollow (the bottom figure above). The top figure below shows a detail of one of the standards carried by the soldiers in the original painting. It has five gold crosses, similar to the heraldic arms of Jerusalem which the Habsburg’s claimed they had a hereditary right to rule. In the altered painting (the bottom figure below), the cross has been transformed into a lion (in the same scene the mounted soldier in black could also represent Alba, the beard being removed in the altered painting).

Original standard with cross
Altered standard with lion

Religious Legitimation, State Capacity, and Toleration

The history of Bruegel’s painting, and Rudolf II’s censorship of it, encapsulates a turning point in European history. Across the continent rulers were being forced to decide whether to respond to the challenge of the Reformation by doubling-down on enforcing religious conformity or backing away from religious war by relaxing the rules surrounding what constituted legitimate religious belief.

The nub of the issue was that states that relaxed the bounds of religious tolerance relatively quickly found themselves taken down a path in which they no longer had the legitimacy (or in many cases the resources) of the Catholic Church to rule. At the worst, this could result in a ruler losing power. This is what happened to Rudolf. Despite issuing the Letter of Majesty, he was still forced to use his military to suppress Protestants demanding even more religious freedom. Ultimately, he was forced to cede power to his brother, Matthias II, who, despite attempting to maintain the delicate balance between religious groups, also failed to prevent conflict. Indeed, the civil war that began when several Bohemian Catholics were thrown out of the window of the Bohemian Chancery by Protestants who feared that the principles outlined in Rudolf’s Letter of Majesty would not be upheld was to become one of the most violent conflicts in European history—the Thirty Years’ War.

At best, undermining traditional reliance on religion for legitimacy and embracing increased religious tolerance forced states to invest in state capacity and adopt more general rules for governing their subjects. This happened because religious authorities traditionally played a vital role in legitimizing (and sometimes managing) the most important legal and fiscal institutions in pre-Reformation Europe. Examples include the role the Church played in providing welfare and social insurance through its alms and charitable institutions, the ways in which guilds combined membership with religious services, and the exclusion of Jews from many areas of economic and social life. Indeed, the willingness of a subject to pay almost any tax or obey any ordinance from a ruler depended, to some degree, on the church’s ability to legitimize it.

How rulers responded to the challenge of the Reformation had major implications for the course of institutional development. Initially both Catholic and Protestant rulers continued to try to use religion as the main basis of their political legitimacy. This failed, however, in those parts of Europe whether the Reformation had produced a large amount of religious heterogeneity. Their attempts to use religion to buttress the state ran into serious headwinds—examples include Charles I’s failed attempt to impose a common prayer book on England and Scotland. Over time, therefore, one alternative strategy became evident to rulers and political elites and this was to substitute reliance on religion for other sources of political legitimacy. This, however required investment in state capacity by the secular government.

A byproduct of this was that, once religious authority was weakened, then greater religious freedom became feasible. The state was freed from the need to enforce religious conformity. And due to the insatiable desire of rulers for tax revenues – usually spent on costly and fruitless wars – they had an incentive to treat individuals more as interchangeable subjects than as Catholics, Protestants, or Jews (though these identities certainly persisted, they no longer held the central role in governance the once did). To borrow a term from James Scott, this was all part of making subjects more “legible” so as to allow the state to tax and govern without the helping hand of the Church.

These trends are visible in the data. The figure below shows the amount of taxes collected per capita for England, France, Spain, Prussia, and Austria. Per capita taxation proxies fiscal capacity, one of the main components of state capacity. It is an imperfect measure, but relatively easy to collect and compare across locations. Two things jump out. First, state capacity was very low in all European states before 1600. Second, after 1600 fiscal capacity increased faster in countries that abandoned religion as a source of political legitimacy.

Fiscal capacity in five European states

This is not simply a question of the Reformation. As we discuss in our recent book, Persecution and Toleration: The Long Road to Religious Freedom, what mattered was how states in different parts of Europe responded to the Reformation and how the Reformation interacted with the intense military-political competition that intensified across Europe after 1500. The religious tensions and conflicts created by the Reformation created major challenges for all European polities. Some states like Spain succeeded in repressing religious diversity within their borders and remained reliant on religion for political legitimacy. The most successful states, however, in the early modern period were England, the Dutch Republic, and to some degree, France which succeeded in moving away from religion as a source of legitimacy and were also most successful in building state capacity.

One way to test the second prediction—that religious toleration increased in places that abandoned religion to legitimize rule—is to look at the treatment of Jewish communities. The position of Jews in medieval Europe was precarious. They were allowed to live in European cities if granted permission from Christian authorities. Often, they were the only source to borrow money at interest in a town—and Christian authorities often exploited this fact by taxing the Jewish community heavily for the privilege. This in turn could exacerbate existing antisemitism which authorities could further exploit to threaten Jews with expulsion or death if they did not accede to their fiscal demands. Thus, a pernicious self-reinforcing equilibrium evolved in many places where Jews provided fiscal services to the state, which reinforced antisemitism, which, in turn, made it easier for the state to extract revenues from them.

Jewish Persecutions, 1100-1800

A consequence of this equilibrium (and many other factors) was that persecution of Jewish communities was endemic in medieval Europe. The above figure shows towns that either attacked or expelled their Jewish communities between 1100 and 1800. Larger circles correspond to more persecution events and triangles represent towns, mainly in Eastern Europe, that had Jewish communities but rarely persecuted them. If we look at the number of persecutions per city-year before and after 1600, cities in states that moved farther away from religious legitimization of rule were much less likely to persecute. For example, the ratio of persecutions per city year after 1600 to before 1600 in England was zero (or rather post-1656 as that was the date at which Jews were readmitted into the country). In France it was 0.10. In Austria the ratio was 1.11, suggesting that Jews were persecuted about the same amount after than before the Reformation. In Italy the ratio was 0.61. In Spain and Portugal Judaism remained entirely proscribed and many of their descendants who had converted to Christianity were persecuted by the Inquisition.

The different institutional paths taken by states that abandoned religion as the primary organizing principle of governance and those that did not also had consequences for economic development. Again, one of the clearest ways to see this is by focusing on Jewish communities. We have excellent documentation of what cities permitted Jews in their borders and which did not throughout the medieval and early modern periods. We combine those data with information on city populations to ask the question: Did cities that tolerated Jews grow faster than those that did not? There are good reasons to suppose this was the case since, on average, Jews were better educated than non-Jews in Europe. Probably because of this, they were over-represented in skilled trades such as long-distance trade and medicine.

Cities that tolerated Jews grew faster after 1600

The above figure shows a plot of century growth rates for 1,792 cities with and without Jewish communities in Europe between 1400 and 1850. These estimates factor out city-level fixed effects, century dummies, and controls for access to transport routes such as seas, rivers, or roads. They suggests that, before 1600, there was a slight growth advantage to having a Jewish community in a city. This is not statistically significant, but could be consistent with Jews playing a useful economic role while also being constrained by stifling antisemitic regulations as well as periodic bouts of persecution. After 1600, however, there is a clear growth advantage to cities that tolerated Jews. By 1700 cities with a Jewish community grew about 5% faster than those without. By 1800 they grew 10% faster. In this paper we provide evidence that this growth advantage was actually causal. Similarly, work by Eric Hornung finds that locations in Prussia that accepted French Protestants fleeing persecution from Louis XIV benefitted economically.

In addition to illuminating the origins of religious freedom and with it modern liberalism, this research holds several lessons for today. First, relying on religious legitimacy was a cheap way to govern in medieval Europe but it came with serious costs. It entailed a reliance on religious identity as a marker of social status and involved the state in enforcing religious conformity—an equilibrium which facilitated periodic spasms of religious persecution. Today, as many groups increasingly argue that personal identity should be one of the main factors determining legal status we should keep this in mind. Moreover, precisely because this arrangement transferred powers away from secular and towards religious authorities, it discouraged investments in state capacity. Given the emphasis on state building in the development literature today, recognizing that there is a link between how a state legitimizes itself and whether there are incentives to building fiscal and legal infrastructure is potentially useful.

Second, the bargain between religious and political authority in medieval Europe came not only at a human cost, it restricted growth. Economic historians such as Joel Mokyr have emphasized the importance of intellectual freedom and a “culture of growth” as prerequisites for sustained economic growth. We argue that one important precondition for this development was the breakdown of the medieval partnership between the state and religious authorities.



  • Noel Johnson

    Noel Johnson is Associate Professor in the Economics Department at George Mason University. He is also a member of the Center for the Study of Public Choice and a Research Fellow at the Mercatus Center. He earned his PhD at Washington University in St. Louis. His interests lie at the intersection of economic history, development, and the new institutional economics. Dr. Johnson's recent research has explored how states build administrative and fiscal capacity; the relationship between state capacity and growth enhancing economic and social outcomes such as religious tolerance or free trade; and how the disease environment has affected economic and social outcomes.

  • Mark Koyama

    Mark Koyama is an Associate Professor of Economics at George Mason University, Virginia and a Senior Scholar at the Mercatus Center. He was a 2017–2018 National Fellow at Stanford University's Hoover Institution.

Leave a Reply