The entrepreneurial state
How state competition – through war – can drive institutional progress
David Schönholzer writes about state development for Issue 13. Read it online here.
One of the most fundamental facts of life is that where you are born is a huge determinant of how your life will turn out. Someone’s chance of leading an economically secure life is vastly higher if they happen to be born in Zurich, Seattle, or Kyoto rather than Mumbai, Nairobi, or Caracas.
But why are some places much better at providing an environment for the development and adoption of technology than others? Much of the divergence between rich and poor places can be traced back in time to the emergence of a set of institutions in Europe, which have partially spread around the world. Northwestern Europe was already substantially richer than other world regions well before the Industrial Revolution, but there is an ongoing debate in economic history on the reasons for Europe’s divergence from the rest of the world.
While geography and culture have likely played some role, as well as expropriation through colonialism, the nature and the development of the state is the most important factor behind economic fortunes. ‘Little else is requisite to carry a state to the highest degree of opulence from the lowest barbarism, but peace, easy taxes, and a tolerable administration of justice: all the rest being brought about by the natural course of things’, wrote Adam Smith back in 1755.
He was right. Secure property rights and a limited state monopoly on violence have withstood the test of time (although after the experience of the twentieth century we may add a welfare state to this list). This stability fostered the investment and invention that drove living standards higher.
In striking contrast to Smith’s peace as an essential ingredient for states to foster prosperity, however, one popular explanation comes from Charles Tilly, who stressed the centrality of war as a driver of state formation: ‘War made the state, and the state made war’. After all, since 1400, more than 1,000 wars have taken place in Europe, with millions of casualties and thousands of leveled towns.
Why would war, an inherently destructive activity, foster development? One story sees wars as the bitter ‘Gifts of Mars’ that helped to propel Europe to early riches by decimating its population and thereby increasing each surviving farmer’s land available for cultivation. Another claims the constant threat of war drove farming households to seek refuge in walled cities, leading to urban growth, and in turn to development.
Even put together, these explanations cannot be complete, since they imply more war is always good. The benefits of war came, of course, at an enormous cost: as well as the deaths of countless soldiers and peasants, European wars also cut short the lives of future inventors, engineers, and statesmen who might otherwise have driven technological progress more quickly. (To give one example, the astronomer William Gascoigne, inventor of the telescopic sight, was killed at the age of 32 at Marston Moor during the English Civil War.) And they ruined prosperous cities that would have served as nexuses of trade and progress. Any explanation for the emergence of wealth that relies on conflict needs to generate a sufficiently big benefit to counterbalance this.
War acting as a selection mechanism for more effective states might provide this counterbalance. While wars had costs, more effective states tended to win out from them, leading to a virtuous cycle of more effective institutions.
The death and birth of states through European history
New research conducted by Eric Weese and myself explains how competition, including through war, led to creative destruction among European states. We use the newly available research edition of the Centennia Historical Atlas, which collects thousands of historical maps in the largest and most accurate database on historical state boundaries in the world. Each year between the years 1000 and 2002 has ten maps – a total of more than 10,000 maps – showing the exact borders of every state in Europe, Western Eurasia, the Levant, and North Africa. When territory changes hands from one state to another in between maps, it may be the result of a war, inheritance, or treaty.
These maps have been painstakingly assembled and digitized over the course of decades by Frank Reed at Clockwork Mapping, opening up an entirely new avenue for research in economic and political history. We vetted the data with the help of two economic historians, whose accounts of individual cities matched the data closely. While some states had less control over their territory than others and some borders were more porous than others, it gives us the best view of ‘boots on the ground’ – de facto control by states – of every acre of land in Europe since the Middle Ages.
We combine these maps with data on the populations of 2,206 European cities between 800 and 1850. A successful state grows its cities’ populations by attracting immigrants from other states; by preventing deaths from war, disease, and poverty; by seeing more births; and through internal immigration.
So how did territorial competition affect states according to this data? Among states with at least one city, the average state survived less than a century. The radical Anabaptist Christian Münster Republic existed for just over a year between 1534 and 1535, before its rebellion was crushed by a confederation of local nobles and the city’s prince-bishop. Revolutionaries overthrew the Prince-Bishopric of Liège in 1789, at the same time as the French Revolution, but the Republic of Liège that they founded was destroyed by a joint Prussian-Austrian invasion two years later. Still, it outlasted the neighboring United Belgian States, which lasted for just one year, between January and December 1790, before its former Austrian rulers crushed the new state and regained control over the territory.
These states did not last in the ruthless market for governance, but their ‘deaths’ were matched by an inflow of ‘newborn’ and expanded states that arose out of the remnants of their forebears – the return of the Prince-Bishopric of Münster and Liège, or the restoration of Austrian imperial rule.
This steady churn was accentuated by periods of mass extinction of states, such as when Napoleon swept the German lands after the French Revolution with his iron broom (including both the Prince-Bishopric of Liège and Austrian rule in Belgium). But even putting these periods aside, the European state system was constantly in flux throughout the entire period between 1000 and 1850.
Creative destruction in the European state system
The process of displacement and reinvention of states throughout European history resembles the process of creative destruction popularized by economist Joseph Schumpeter: ‘the process of industrial mutation that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one’. In models of creative destruction, companies develop new technology over time to gain an edge over their competitors. This leads to three effects: firstly, technology improves as businesses with better technology enter the market; secondly, those with worse technology leave, freeing up the resources they had been using for other companies to use instead; and thirdly, firms with superior technology gain market share at the expense of their less effective rivals.
All three of these forces are at work in competition between states as well. New states come into existence based on rules that allow for a more favorable economic environment, for example when the mercantile Dutch Republic replaced the absolutist Habsburg Monarchy in the Low Countries. Existing states may be more likely to disappear (through war or revolution) if they are unable to raise taxes to pay soldiers to defend themselves, or otherwise cannot command the loyalty of their population and control their territory. Finally, better states may foster more economic growth, funding military and diplomatic efforts to gain territory at the expense of rivals.
A tale of two cities: the costs and benefits of switching states
The enormous detail available in the Centennia Historical Atlas allows us to set up a research design that isolates the costs and benefits of territorial competition. This research design is exemplified by the fates of the cities of Prague and Nancy during the early seventeenth century.
In 1618, four emissaries of the Catholic emperor of the Holy Roman Empire arrived in Prague to restrict the rights of local Protestants, only to be promptly defenestrated – thrown out of a window – precipitating the beginning of the Thirty Years’ War. After an initially successful suppression of the resulting Bohemian Revolt by the emperor and his allies, the Swedish king Gustavus Adolphus sought to support the Protestant cause and liberated Prague once again in 1631. But less than a year later, Prague ended up where it started: under the yoke of the Holy Roman Emperor.
So while Prague has never left the banks of the Vltava River, it de facto switched from rule by the Holy Roman Empire to the Bohemian estates, which invited Frederick V of the Palatinate of the Rhine to become their king, then back to the Holy Roman Empire, then briefly to Swedish control (a largely Saxon army under Gustavus Adolphus’s overall command occupied and sacked the left bank of the city, but never managed to control the other side of the river), and finally back to the Austrian Empire.1
All of this happened in such a short time span – just 14 years – that the costs of war, and the wider costs of switching from one ruler to the next, dominated any potential benefits from better governance by some of the intermittent rulers. Consequently, by comparing the size of Prague over the course of the seventeenth century to cities that did not switch between states over this period, we can isolate the costs of territorial competition: the extent to which Prague’s population drops over this period relative to other cities gives us a measure of the costs of conflict. Prague lost more than half of its 100,000 inhabitants over the course of the seventeenth century, whereas cities not exposed to conquest, revolution, or any other switch during this period grew on average.
The city of Nancy went through far fewer and less bloody switches between states during this period. In 1634, taking advantage of the general disorder of the Thirty Years’ War, the French king annexed the Duchy of Lorraine and its capital of Nancy, ruling it for decades. Although Lorraine only became a permanent part of France in 1766, its experience in the seventeenth century serves as a natural experiment for the effect of French rule as opposed to that of the Duchy of Lorraine.
Comparing Nancy to non-switching cities can then serve as a means for testing the potential benefits of territorial competition: as Nancy grew faster on average under French rule than under the Duchy of Lorraine, this suggests that France was relatively better for long-run city growth than Lorraine. Since France annexed Lorraine, permanently ending the existence of Lorraine as a state – pushing it out of the ‘market for governance’ – the fate of Nancy represents a data point in support of the theory of creative destruction in state competition. Nancy may have suffered for a while after annexation, but over the long run, the improvement in governance may have been worth it. By contrast, Prague during the 1600s is evidence of too much switching.
A continent of switching cities
Of course, Prague and Nancy were not the only European cities that changed states between 1000 and 1850. In fact, out of more than 2,000 cities whose populations we can track in this period, 80 percent changed hands at least once. Many cities saw several government changes, and some of them went through dozens.
The two European regions with the highest number of switches are Northern Italy and the Low Countries, which includes cities like Ghent. For example, Genoa was an oligarchic trading republic from the eleventh to the eighteenth centuries, was invaded by Revolutionary France in 1797, and was constituted into the Ligurian Republic until 1805, when it was annexed into Imperial France. It liberated itself through rebellion in 1814, but was integrated into the Kingdom of Sardinia in 1815 by the Congress of Vienna. And so on.
Having grown rich from maritime trade during its Golden Age from the twelfth to fifteenth centuries, Ghent was not able to recover to its peak population size until the nineteenth century. During the intervening years, it had the bad luck of being conquered many times, and governed by a near constant parade of unstable, low-quality rulers. This is where Smith’s edict of peace as an essential feature of good governance shines through most clearly: perhaps Ghent would be a major European city today, in terms of population size, had it enjoyed more stable rule.
It may be easy to find individual city histories in support of one or another theory of long-run growth. But does the history of cities switching ruler or state across the European continent in this almost 1,000-year period provide support for creative destruction?
Our data suggests it does. On average, cities that switch – so often through revolutions or conquests – see population drop by almost 30 percent relative to cities that did not switch at that time. Remarkably, in the time before the switch, these cities grew at the same speed as other cities. This suggests that the comparison is reasonable.2 We can also see that the costs of switching are not permanent: average city population tends to recover in the century after switching. It seems the experience of Prague was not unique – territorial competition generally came with an initial cost.
But switching had benefits too, even when it came through violence. Before, we estimated the initial impact of switching between states – averaged across revolutions, diplomatic annexations, state reconstitutions, conquests, and so on. If we account for this, we can then isolate the potential benefits of being under the new, incoming government.
A very different picture emerges: after accounting for the transitory cost of switches, switching to another state is associated with a permanent 20 percent increase in population size.
We find three sources of benefits due to territorial competition: first, some of the benefits are due to better, or more effective, states taking over more cities over time. The French occupation of the Rhine during the Napoleonic Wars makes for a striking example of these benefits. In the 1690s, the region was burdened by toll booths that required payment of different state customs duties every six miles. The French did away with this remnant of the Holy Roman Empire, which allowed for easier trade and subsequent city growth in the region.
Second, cities end up being matched with states that seem to be a better fit for them, perhaps because they care more about the city for religious, nationalistic, or other ideological reasons. For example, the creation of the Dutch Republic put the cities of the Low Countries under the stewardship of a state that cared deeply about their welfare – unlike the Spanish Habsburgs, who had ruled the region before the Dutch Revolt and who had little regard for anything other than the tribute they sought to extract from these cities. (The Habsburgs continued to rule much of Belgium, which did not experience the same economic boom after the Dutch Revolt that the Dutch Netherlands did.)
And third, some benefits to switching seem to be due to a pure change-in-governance effect: possibly, the breakup of patronage networks in the city tied to the former state in power helps to set free previously suppressed productive forces, no matter the quality of the state or of the match. For instance, after the demise of the Old Swiss Confederacy in 1798, the Helvetic Republic briefly took over governance in Switzerland. Although many Swiss resented it for its centralism and forced a partial reconstitution of the Swiss Confederacy, many oppressive political arrangements of the Old Confederacy that stymied city growth had been permanently removed.
Features of good governance: constraints and capacities
The experience of Nancy gives an example of how French governance at the time compared to that of the Duchy of Lorraine. Averaging across hundreds of cities that switched states like Nancy, we can estimate the average growth effects of each state that was on the receiving end of at least one city switch.
Despite rich scholarship on states throughout European history, it is difficult to learn about the quality of governance of different European states over time, especially for those states that ceased to exist hundreds of years ago. More importantly, these estimates allow us to investigate which features of states are better for city growth.
One of the correlates with higher estimated state effects on city growth is the extent to which the state had an active parliament. Parliaments offered a forum for rulers to bargain over peace guarantees and property rights in exchange for tax revenues – part of Adam Smith’s recipe for high-functioning states. Cities switching to states such as Britain, and England before it, in which Parliament convened roughly every other year, grew faster than those switching to states like Denmark, in which parliaments were nearly nonexistent during this period. These estimates suggest that going from no parliament to one that meets every year leads cities in the state to grow almost nine percent faster per century.3
Of course, this relationship is not perfect: some states had a positive impact on city growth even without a powerful parliament. If the ruler of a high-quality state was not constrained by parliament, they often had access to other capacities that fostered growth. In particular, higher-quality states had higher fiscal capacity: they were able to maintain more debt at a lower interest rate than other states. This helped to secure internal peace in times of crisis and invest into productive capacities such as infrastructure once the Industrial Revolution took off.
One example of a city that benefited from constrained, high-capacity governance is Leipzig under the rule of Saxony. After attaining some degree of autonomy from the Holy Roman Empire during the fifteenth century, the Saxon state developed under a restrained nobility and court whose rulers supported unusually modern civil law and commercial ordinances, sometimes in opposition to special interests. Leipzig hosted one of the largest trade fairs in Central Europe for almost two centuries, supported by legal institutions that protected merchants from expropriation and fraud and guaranteed access to credit. As a consequence, Leipzig experienced substantial population growth on the basis of flourishing industries, especially trade and manufacturing.
A global view: moderate competition and high-quality challengers
Taking all this evidence together, it seems that competition between European states fueled a process of creative destruction in statecraft. But does this mean that more competition between states is always better? Probably not. Indeed, our evidence suggests that the gains from switching from one state to another decrease the more a city does it. While being invaded once a century may shatter ossified patronage networks keeping the city back, a revolution every decade is more likely to put it into a state of permanent chaos and drive people into exile.
That suggests there is a competition sweet spot.
Another important takeaway is that not only may Europe have had close to the ‘Goldilocks’ amount of competition between states, but it also benefited enormously from the ability of states to copy the successful innovations of others through means other than through annexation, and a relatively unified collective intellectual culture. Waves of institutional innovations brought about by newcomers forced other states to adapt and learn if they wanted to survive, just as with useful and technological innovations. The introduction of parliaments, the rise of fiscal capacity, and later on executive constraints and the rule of law went hand in hand with enhanced military might, leading to a virtuous cycle of better governance coupled with larger shares in the market of governance.
The rise of parliaments makes for a powerful illustration of this process, mirroring the spread of ideas like calculation, and technologies like gunpowder. They first arose in the context of the Reconquista in 1126 as a way for Castilian kings to compete for the favors of merchants and farmers in formerly Muslim cities. From there, parliaments gradually spread to the rest of Western Europe, first to Barcelona and then to France. In the words of one study, ‘within the highly competitive European state system, the institution spread in part because it affected sovereigns’ access to resources, and therefore their ability to fight wars’.
In sum, Europe’s secret to early growth may lie in the delicate balance between the costs of war and the creative forces unleashed in the competition between states. Many cities faced destruction throughout European history. But out of their ruins rose, gradually and by no means irreversibly, the foundation for better governance.
While the de facto ruler of Prague changed frequently in this period, it never formally left the Holy Roman Empire. This reflects the Centennia Atlas's focus on boots-on-the-ground control rather than whether a ruler had a legitimate claim to the territory.
One concern with using non-switching cities as a comparison group is that they might be growing more than switching cities because they don’t have to fear a hostile takeover to the same degree that a border city might. To address this possibility, we control for the distance to the nearest border in our statistical exercises, essentially comparing switching and non-switching cities that are at a similar distance from the nearest border.