Rulers, Religion, and Riches: Why the West Got Rich and the Middle East Did Not (Cambridge Studies in Economics, Choice, and Society)
Page 19
Other heresies abounded in the centuries prior to the Reformation. In fifteenth-century England, the Lollard movement spread the ideas of John Wyclif (d. 1384). Like Gerson, Wyclif wrote during the Great Schism between Avignon and Rome. Wyclif was a supporter of the rights of lay rulers over the papacy – he claimed that lay lords had the right to take the property of undeserving clergy – giving him significant influence with poorer parish priests. Wyclif went as far as to attack the doctrine of transubstantiation, a key point of doctrinal attack for the Protestants. The Church brutally suppressed the Lollard movement that Wyclif inspired in the half-century after his death. Their predecessors, the Waldensians, met a similar fate in the twelfth and thirteenth centuries. The Waldensians, who took a vow of poverty, rejected open displays of Church wealth and the worldly lives of churchmen. They gained some measure of influence in France, Spain, and Italy, but the Church and its allies suppressed them wherever they arose. In 1192, they were ordered in France to be put in chains; in 1194, they were banished from Aragon and the populace was forbidden to furnish them with shelter or food; and a decree of death by burning was enacted against them at the Council of Gerona in 1197.13
Perhaps the most serious challenge to the Church came from the Prague preacher Jan Hus (c. 1372–1415), who led the anti-Church movement that bore his name in the early fifteenth century. Hus challenged the rights of sinful churchmen to keep their positions and wealth, denounced the morality of what he perceived to be a corrupt clergy and pope, and translated Wyclif’s heretical writings into Czech. For this the Church excommunicated Hus in 1410, although he continued to speak out against church abuses such as the offering of indulgences to anyone who supplied funds for a crusade against the king of Naples. These positions ultimately caused the Church to burn him at the stake in 1415.14 The Hussite movement that followed over the next century established rival churches throughout Bohemia based on the denial of the Roman hierarchy. The Roman Church was able to limit the spread of the Hussite movement – going as far as to send a crusade to Bohemia – and Hussite influence never extended beyond Bohemia. The Hussites simply never gained much traction outside of Bohemia because the Church was quickly able to levy punishment against anyone that showed Hussite sympathies.
It is striking that the Church rather easily suppressed all of the attempts at reform prior to the invention and diffusion of the printing press. A. G. Dickens (1968, p. 51) explicitly makes this contrast between these movements and the Reformation: “Unlike the Wycliffite and Waldensian heresies, Lutheranism was from the first the child of the printed book.” But disentangling the role the press played in the spread of the Reformation from other causes is no small task. For example, is it possible to separate the role of the press from, say, the increased selling of indulgences? In order to make a causal claim connecting the spread of printing to the Reformation, it is necessary to dig deeper. I did just this in a 2014 article in which I collected and analyzed city-level data on printing presses, Reformation status, and economic characteristics. The following section provides a brief overview of that analysis.
Testing the Effect of the Printing Press on the Reformation
The central focus of the analysis conducted in Rubin (2014b) is confined to the Holy Roman Empire, which was the birthplace of both printing and the Reformation.15 It is useful to concentrate on the Holy Roman Empire because this is where the Reformation initially spread. The most important aspect of the Reformation to analyze is its initial spread, because so many other previous attempts at reform were never able to get off of the ground. Once the Reformation spread throughout the Holy Roman Empire, it took on a life of its own elsewhere. Henry VIII was able to bring the Reformation to England to suit his own dynastic and financial purposes (see Chapter 7). Like Henry VIII, the Swedish King Gustav I (r. 1523–1560) confiscated land from the Church during the Swedish Reformation. In France, Calvinist churches rapidly spread in the west and south in the 1550s. The French monarch violently suppressed these Protestants, known as Huguenots, until they agreed on a series of peace edicts in the 1570s–1590s. Similar movements occurred in the Low Countries, where William of Orange co-opted the new religion in part as propaganda favoring the Dutch Revolt from Spanish rule (see Chapter 7). In each of these cases the Reformation spread for reasons that had little to do with the reasons for its emergence. But this is precisely the point: without the initial spread of the Reformation, Henry VIII, Gustav I, William of Orange, and their Protestant counterparts would never have had such an opportunity.
For the sake of statistical analysis, studying the Holy Roman Empire is also convenient because there was substantial variation in religious choice within the empire. Without any variation in religion, it is difficult to discern what the driving factors in adoption were; how do we know what convinced a city to adopt the Reformation if all (or none) of them did so? The variation in the Holy Roman Empire thus allows one to test the importance of various factors that could have caused a city to adopt the Reformation. Table 6.1 provides initial evidence; it lists cities in the Holy Roman Empire with populations of at least 20,000, along with their religious affiliation in 1600 and whether the city had a printing press by 1500. It is immediately noticeable from this table that a majority of the larger cities in the Holy Roman Empire had printing presses. This is not surprising. Printing spread outward from Mainz soon after its invention in 1450, and printers generally moved to large population centers where demand for printed works was greatest. This is the primary reason why A. G. Dickens’s (1974) oft-cited claim that the Reformation was an “urban phenomenon” might be a spurious connection. If the printing press were indeed a significant causal factor in the adoption of the Reformation, then cities that were likely to adopt the Reformation were also likely to be large, since large cities were more likely to have a press.
Table 6.1 Cities in the Holy Roman Empire (Population ≥ 20,000)
Cities (with population ≥ 20,000) with Printing Presses by 1500 Cities (with population ≥ 20,000) without Printing Presses by 1500
City Population (in 1500) P/C (by 1600) City Population (in 1500) P/C (by 1600)
Prague 70,000 C Tournai 35,000 C
Ghent 55,000 C Lille 26,000 C
Cologne 45,000 C Mechelen 25,000 C
Nuremberg 38,000 P
Bruges 35,000 C
Brussels 33,000 C
Augsburg 30,000 P
Antwerp 30,000 C
Breslau 25,000 P
Lübeck 25,000 P
Regensburg 22,000 P
Strasbourg 20,000 P
Vienna 20,000 C
Note: Population data from Bairoch et al. (1988); P/C indicates whether a city was Protestant or Catholic by 1600.
More than population must be considered to understand the connection between the printing press and the Reformation. For example, a quick glance at Figure 6.5 indicates that proximity to Wittenberg played a role in a city’s likelihood of adopting the Reformation.16 It is also possible that cities that housed universities were more likely to reject the Reformation – many universities were church strongholds – but also adopt the press early due to their scholastic nature. Indeed, Hyojoung Kim and Steven Pfaff (2012) argue that university students were important links connecting pro- and anti-Reformation ideas from the universities to the broader spread of the Reformation. They find that cities housing numerous students from the universities at Wittenberg and Basel (Zwingli’s intellectual home) were more likely to adopt the Reformation, while cities housing students from Cologne and Louvain (Catholic strongholds) were less likely to adopt the Reformation.17 Table 6.2 also suggests this possibility; it lists the eleven cities in the Holy Roman Empire that housed a university by the time that Gutenberg invented the movable-type press in 1450. All six of the university towns that also housed a bishop or archbishop remained Catholic. Meanwhile, four of the five university towns that were not bishoprics or archbishoprics adopted Protestantism. This suggests a dual effect of universities on adoption of the Refor
mation: those in Catholic strongholds may have been more able to fend off the Reformation, while those not in Catholic strongholds were more likely to view the Reformation positively. Of course, this is hardly concrete evidence for such an effect – it is evidence from only eleven cities – but it does indicate the need to take a number of factors into account before making a causal claim connecting the spread of printing to the Reformation.
Figure 6.5 Printing and Protestantism in Western and Central Europe
Sources: Reprinted from Jared Rubin, “Printing and Protestants: An Empirical Test of the Role of Printing in the Reformation,” The Review of Economics and Statistics, 96:2 (May, 2014), pp. 270–86. © 2014 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.
Table 6.2 Cities in the Holy Roman Empire with Universities by 1450
City Population (in 1500) Bishopric P/C (by 1600)
Prague 70,000 Y C
Cologne 45,000 Y C
Vienna 20,000 Y C
Erfurt 19,000 N P
Leuven 17,000 N C
Leipzig 10,000 N P
Rostock 10,000 N P
Heidelberg 8,000 N P
Trier 8,000 Y C
Würzburg 7,000 Y C
Dôle 5,000 Y C
Note: Population data from Bairoch et al. (1988); P/C indicates whether a city was Protestant or Catholic by 1600.
Fortunately, social scientists have a method called multiple regression analysis, which allows one to address these problems. In a nutshell, multiple regression analysis provides a best-fit prediction for how one variable affects another while holding all other variables constant. In other words, the results could tell us: Given that a city has population X, has a university, is home to a bishop, and many other things,18 what is the average probability that this town adopted the Reformation if it had a press? What is the average probability that it adopted the Reformation if it did not have a press?
These questions are answerable even if non-observable features might affect a city’s likelihood of adopting both the printing press and the Reformation. For instance, literacy rates are unknown for the period prior to the spread of printing, especially in small towns. But pre-press literacy might have affected the adoption of the press and the adoption of the Reformation. More literate towns were more likely to adopt the press, and it is quite possible they also had a greater desire for the Reformation, due possibly to greater engagement in humanist philosophy or greater awareness of church corruption. Alternatively, more literate towns might have been Catholic strongholds due to many churchmen being literate, meaning that such towns were less likely to adopt the Reformation. Either way, it is possible to disentangle the connection between the spread of printing and the Reformation without any concrete evidence on pre-press literacy through an econometric technique known as two-stage least squares regression analysis (2SLS). This technique permits the separation of the two effects by first estimating the features that affected whether a printing press was present in a town, then using this information to estimate the effect of the printing press on the likelihood of a town adopting the Reformation.19
The 2SLS regression analyses employed to test the connection between the press and the Reformation provides very strong results. They indicate that the mere presence of a printing press prior to 1500 increased the probability that a city would become Protestant in 1530 by 52.1 percentage points and Protestant in 1600 by 29.0 percentage points, all else being equal. These results far surpass the “95 percent confidence threshold” normally considered necessary for a result to be statistically significant, indicating that there is a strong relationship between the spread of the printing press and the spread of the Reformation. Table 6.3 lists some of the other regression coefficients that are statistically significant with at least 95 percent confidence (it lists only statistically significant coefficients). These coefficients can be interpreted as follows: if a city had a characteristic in the leftmost column (i.e., it had a printing press, was a free imperial city, had a university, or was a bishopric), its probability of adopting the printing press or the Reformation was affected, all else equal, by the corresponding number. Likewise, one can multiply a city’s log distance to Mainz or Wittenberg or its population by the number in the corresponding column to see how these variables affected the likelihood of press or Reformation adoption. As expected, being further away from Wittenberg made adoption of the Reformation less likely, university towns were much more likely to adopt the printing press than non-university towns (by 36.5 percentage points, although they were not clearly more or less likely to adopt the Reformation), and bishoprics were also more likely to adopt the press (by 21.4 percentage points) than non-bishoprics while they were less likely to adopt the Reformation (by 11.9 percentage points).
Table 6.3 2SLS Regression Coefficients – Effect of City Characteristics on the Reformation
Did a city adopt the printing press by 1500? Did a city adopt the Reformation by 1600?
Printing Press by 1500 – 29.0%
Log Distance to Mainz –19.2% –
Log Distance to Wittenberg 8.8% –34.3%
Log Population in 1500 12.7% –4.3%
Free Imperial City – 30.8%
University Town 36.5% –
Bishopric 21.4% –11.9%
While no counterfactual history exists to indicate whether an event like the Reformation would have occurred without the press, these results suggest that the printing press was necessary for the Reformation to occur when and where it did. Consider again the fate of previous attempts at reforming the Church. The Church rather easily and brutally suppressed the Hussite movement, Lollards, Waldensians, and others. The presence of these movements indicates the possibility that the seeds of discontent existed for centuries. The primary difference between Luther’s movement and his predecessors is that Luther had the press.
The deleterious effect of the Reformation on the legitimizing power of the Church raises the question: Who replaced the Church in propagating political rule in Protestant states? Rulers must have looked elsewhere for propagation. Who did they turn to, and what effect did this have on laws and policies?
Propagation by the Economic Elite: A Protestant Phenomenon?
The decline of religious legitimacy following the Reformation paved the way for different propagating agents to increase their say in governance. The agents in the best position to replace the Church were the economic elite that comprised parliaments – merchants, urban commercial interests, and the landed elite. Protestant rulers ended up turning to these elites more frequently following the Reformation than their Catholic counterparts.
The economic elite provided one of the most important – and most expensive – alternatives to religious propagation. Elite support was especially important during times of war, when kings needed both money and loyalty. Indeed, a growing literature suggests that beginning in the seventeenth century, the need for defense against increasingly large and well-organized states created a common interest between rulers and the economic elite to provide greater provision of public goods, especially defense, which in turn required greater access to revenue.20
While protection against foreign invasion may have provided some of the initial impetus for the growth of large-scale fiscal institutions, the economic elite also wished to receive other things which improved their economic standing. In some cases, this meant investment in naval protection, which increased trade and prevented attacks from outside forces, or investment in poor relief, which reduced vagrancy and other social ills. If the elite were powerful enough or if the ruler were weak enough, they could ask for the biggest prize of all: secure property rights and freedom from arbitrary encroachment on those rights. When a weak ruler refused these rights, the elite were sometimes powerful enough to revolt. This occurred numerous times in medieval and early modern Europe – the English Baron’s Revolt of the early thirteenth century (which culminated in the Magna Carta), the Spanish comuneros revolt (1520�
��1521), the Dutch Revolt of the 1570s, and the English Civil War of the 1640s are cases in point.
Jan Luiten van Zanden, Eltjo Buringh, and Maarten Bosker (2012) argue that the precarious position of medieval European rulers is precisely why parliaments formed when they did. In the twelfth-fourteenth centuries, as the power and wealth of the economic elite in the cities grew to rival that of the nobility and the Church, the city leaders, nobility, and clergy formed parliaments in order to collectively bargain with the king. In return for tax revenues, kings agreed to constrain themselves – that is, they agreed to not arbitrarily encroach on property rights and would only ask for tax revenue during times of war or fiscal crisis.21 The first parliaments arose in Spain in the late twelfth century after the Reconquista of parts of Spain from the Muslim Umayyads. King Alfonso IX (r. 1188–1230) called the first parliament (Cortes) in León. He called for a meeting of the important citizens – nobility, bishops, and elected citizens – in order to stabilize his regime, which had low legitimacy due to the fact that the citizens were newly conquered and owed no allegiance to him. Parliaments consisting of Church leaders, nobles, and city leaders quickly spread to England, France, and Portugal in the thirteenth century and the rest of Europe soon after, helping rulers gain a consistent stream of tax revenue in return for limited rights, including veto power over new taxes. This “king and council” template was the dominant form of governance in almost all parts of Western Europe until 1800.22 Roger Congleton (2011, p. 192) notes that “the paucity of governmental alternatives analyzed by enlightenment scholars shows how narrow the range of governance was in Europe in the late-medieval and early-modern periods. Neither Hobbes, Locke, Montesquieu, Rousseau, Kant, nor von Humboldt took the time to analyze representative or parliament-dominated systems fully, in large part because they had never seen one operate.”