1 Introduction

In order for a state to survive, it must be able to provide adequate defense from external predation. There is reason to believe that the cost of providing adequate defense is increasing in the total amount of capital (wealth) within the geographic boundaries of the state. If so, capital accumulation provides a private benefit to the owner, but imposes an external cost on society. As with any externality, the optimal policy is to tax capital in order to finance defense spending (Thompson, 1974, 1979; Hendrickson et al., 2018; Hendrickson, 2019). This policy simultaneously limits the amount of capital to an amount that the state can effectively defend while simultaneously raising revenue to pay for defense.

The ability to implement this optimal tax policy requires that the state has the fiscal capacity to do so. In the Middle Ages, rulers faced the same constraint on defending their territory, but the bureaucratic infrastructure of the modern state did not exist. Capital taxation was not feasible. Nonetheless, one would expect that the states that survived the selection process of war and plunder were those that developed policies and institutions that effectively tried to replicate this optimal tax regime.

In this paper, I argue that the enforcement of usury restrictions can be understood as an attempt to limit capital accumulation and generate revenue in much the same way as an optimal tax on capital.Footnote 1 Lending at interest during the Middle Ages and early Renaissance was officially prohibited. Nonetheless, secular rulers often allowed certain groups of people to lend in exchange for something akin to a licensing fee. Other secular rulers subjected those who violated restrictions on lending to “fines” or lump-sum taxes. These fines were effectively no different than a licensing fee since they were a fixed sum and not a function of how much lending took place. At the same time, rulers imposed maximum interest rates on these lenders.

The licensing fees and fines are a barrier to entry that effectively grants a monopoly to the lender. Maximum prices, or in this case interest rates, limit the monopoly profits of the lenders. The combination of long-run entry restrictions and maximum interest rates allowed rulers to limit capital accumulation to a level consistent with their ability to provide defense. The fines represented a source of government revenue that could be used for defense-related resources. As such, this system can be seen as an attempt to replicate the outcome of an optimal tax on capital when fiscal capacity is limited.

I then examine the incentives of rulers to maintain such restrictions, confiscate the wealth of the usurers, or invest in state capacity to replace usury restrictions with formal taxation. Thus, the paper contributes to the literature on the historical origins of state capacity and provides some basis for thinking about endogenous state capacity. The paper does so by drawing on the broad macro-historical accounts of the development of state capacity, but also by focusing on the incentives and decisions of the rulers.Footnote 2 Thus, while much of the literature on usury restrictions attempts to explain why such restrictions existed, the purpose of this paper is to explain the microeconomic underpinnings of the implementation of usury restrictions and how rulers respond to these incentives.

Given the theory, I examine the historical experience of usury restrictions in pre-modern England, Italy, and France. The historical evidence shows that the implementation of usury restrictions in England and Italy are consistent with the idea that such restrictions were an alternative to capital taxation. The subsequent departure from usury restrictions is consistent with my model of ruler behavior. The historical evidence for France demonstrates an inconsistent pattern of applying usury restrictions in a way that replicates the optimal tax. Nonetheless, the fact that the French resorted to expropriation can be explained by my model of ruler behavior. In particular, my model of ruler behavior suggests that rulers will resort to confiscation when feudal revenues are sufficiently low and when the ruler’s discount rate is sufficiently high. There is significant evidence of the difficulty in generating feudal revenue. In addition, the political struggles of the French ruler provide reason to believe that the French kings during this period had high discount rates.

Furthermore, since my theory of usury restrictions as an alternative to capital taxation is used to explain the financing of defense, this implies that the use of expropriation should result in inadequate financing and military failures. Again, this is evident from the French historical experience. The crown often lacked financing and was correspondingly ill-equipped for war during this period. Overall, the French experienced a lack of military success prior to adopting something akin to a tax on capital – and the resumption of its difficulties when the tax was eliminated.

2 Background and related literature

This paper focuses exclusively on usury restrictions in the Christian world. Before discussing the related literature, it might be useful to review the Catholic Church’s position on usury from its founding through the period covered in this paper (the period to 1500). While there is no explicit ban on interest in the New Testament, the Catholic Church started prohibiting interest in the fourth century (Rubin 2017, p. 83). In the late twelfth and early thirteenth century, in response to the pressures of the Commercial Revolution, the Church intensified its prohibition when it “proscribed excommunication for usurers, refused usurers burial in Christian grounds, and interdicted usurers’ offerings” (Rubin 2017, p. 84). At the end of the thirteenth century, the Church continued to crack down on usurers through further canon law punishments. Nonetheless, the Commercial Revolution brought about many attempts to circumvent usury laws. As Rubin (2017) points out, the weakening of the Church as a legitimizing agent gave Western European rulers the ability to modify usury restrictions and enforcement during this period. Even though these attempts did not weaken the official doctrine, they did weaken the restrictions in practice. Ultimately, the endorsement in the sixteenth century of charitable banks created by the Franciscans in Italy known as monte di pietà officially weakened the Church’s stance on usury (Pascali, 2016; Rubin, 2017).

There is a literature on usury laws and the political economy thereof that primarily focuses on the role of the Catholic Church in establishing usury laws and the comparative development of states with different sources of religious legitimization. Among the first group, Ekelund et al. (1989) argue that the Catholic Church adopted usury rules as a form of rent-seeking. However, subsequent authors have argued that this explanation is not consistent with the data (Reed & Bekar, 2003). Others, such (Posner, 1995) and (Rubin, 2009), suggest that usury laws were motivated by the Church’s use of social insurance. The presence of social insurance might encourage the poor to take on excessive risk. By outlawing interest or by imposing a maximum interest rate, this would deter some level of risky lending. Both (Glaeser & Scheinkman, 1998) and (Reed & Bekar, 2003) argue that the church imposed usury restrictions for reasons related to consumption smoothing. Nonetheless, they differ in their reasoning. Glaeser and Scheinkman argue that usury laws can maximize social welfare ex ante. Reed and Bekar offer a public choice argument. They argue that credit markets can potentially substitute for church charity as it relates to consumption smoothing. As a result, the Church had an incentive to limit lending markets. Koyama (2010b) argues that the persistence of usury restrictions is due to the coalition of beneficiaries consisting of secular rulers, the Church, and merchant-bankers. The basic idea is that usury laws not only satisfy the Churches objectives, but also allow rulers to achieve their objective of raising revenue and creates a barrier to entry and therefore monopoly rents for bankers.

Bueno de Mesquite and Bueno de Mesquita (2023) offer a novel political economic explanation for the Church’s use of usury restrictions. In particular, they point to the Investiture Controversy and the resolution thereof in the Concordat of Worms in 1122 as informative for the Church’s enforcement of usury laws. By the eleventh century, there was a struggle for power between religious and secular leaders in Europe. This manifested itself in a dispute about who was allowed to choose the bishops (including the pope). This debate was important because bishops could be used to keep the local community loyal to the church, but could also be a potential source of legitimation for secular rulers. The Concordat at Worms granted the pope the right to nominate the bishops. However, the secular rulers had the right to accept or reject a nomination. If the secular ruler rejected the pope’s nominee, the income of the diocese would accrue to the ruler until a bishop was nominated and accepted. This agreement was binding between the pope and the rulers of the Holy Roman Empire, France, Italy, England, and Burgundy. Bueno de Mesquita and Bueno de Mesquita argue that this agreement had significant effects on the bargaining power of the secular rulers depending on the income of the diocese since that income accrued to the ruler in the event a nominee was rejected. This implies that economic development enhanced the bargaining power of secular rulers. In the aftermath of the Concordat at Worms, the Church banned usury. Bueno de Mesquita and Bueno de Mesquita suggest that this reduced the supply of lending and increased the cost of borrowing, which discouraged economic development and limited the bargaining power of secular rulers.

Rubin (2010; 2011; 2017) examines the implications for comparative development from differences in the evolution of usury restrictions under Christianity and Islam. Specifically, he argues that Western countries did not have to rely on religious authorities for legitimating their rule. As a result, Western countries were able to relax interest restrictions over time. This created a mechanism by which finance and commerce were able to develop in the Western world relative to the Middle East.

This paper extends the literature on usury restrictions to examine why rulers chose to impose usury restrictions in the particular way that they did. In other words, whereas a lot of the literature on usury attempts to explain why such restrictions were put into place, I attempt to provide some microeconomic foundations to understanding why usury restrictions were enforced in a particular way across both time and space. My argument emphasizes the role of national defense. In particular, I consider how usury restrictions compare to optimal tax policy when the government faces an adequate defense constraint. The paper is therefore related to literature on the role of war and the development of policies and institutions to deal with defense-related concerns.

The general argument that war and national defense were a main source of the growth and development of the modern state is one that stretches across disciplines (Lane, 1958; Peacock & Wiseman, 1961; Ames & Rapp, 1977; Batchelder & Freudenberger, 1983; Brewer, 1989; Rasler & Thompson, 1989; Downing, 1992; Ertman, 1997; Hoffman & Rosenthal, 1997; Tilly, 1975, 1985, 1992; Ferguson, 2001; Kiser & Linton, 2001; Thompson & Hickson, 2001; Karaman & Pamut, 2013; Morris, 2014; Gennaioli & Voth, 2015). If war and national defense are, at least in part, responsible for the development of the modern state, then a natural evolutionary hypothesis suggests itself. Those states that have thus far survived wars and other conflicts are those who have developed institutions, policies, and militaries capable of survival. As such, institutions and policies that remain in these states might find their origin and existence in the desire of the states’ historical rulers to provide adequate defense from external predation. Furthermore, one might find other historical policies and institutions that are the antecedents to these modern alternatives. As with any evolutionary argument, this does not necessarily imply that such policies and institutions were designed by benevolent rulers capable of ingenious solutions. On the contrary, these policies and institutions were likely the result of trial-and-error and experimentation. What makes defense-based institutions particularly interesting to examine is that there is an obvious selection mechanism. The inability to provide adequate defense results in observable military failure. States that choose policies and institutions that were not up to the task of contributing to an adequate defense ultimately lose military battles. Such states either disappear, are absorbed by another state, or are forced to adapt.

Along these lines, Ames and Rapp (1977) argue that wars typically require emergency financing and therefore need to rely on borrowing. Nonetheless, they point out that new tax policies and regimes sometimes emerge during wartime and these policies and regimes tend to be the most sticky. Thompson (1974) examines a model in which some amount of capital within a state’s geographic boundaries is coveted by potential adversaries. He derives the optimal tax policy in this context and argues that the U.S. tax system is consistent with his optimal tax regime. Thompson (1979) argues that other policies can be explained by this theory as well, such as subsidies for capital that otherwise would have been undervalued due to things like price controls during wartime. A common critique of this argument is that while this explanation fits the theory, this sort of argument never seems to be put forth by policymakers and advocates of observed tax policies. Thompson and Hickson (2001) argue that this is because tax policy and other institutions emerge out of an evolutionary process for the state. In fact, they argue that vital policies and institutions are not always understood in their time and are undermined by ideological opposition. However, this provides testable hypotheses. If these institutions are critical, then ideological eliminations of these policies and institutions should lead to military defeats and/or an end (at least in the near term) of economic growth and development.

In the intervening time between the publication of Thompson’s original paper on taxation and the fully articulated evolutionary approach to this argument, Thompson and others have explored the role of particular policies and institutions in the context of defense externalities. For example, Hickson and Thompson (1991) argue that guilds in medieval Europe are one such institution. While most argue that guilds were inefficient monopolies, Hickson and Thompson point out that guilds often imposed maximum prices and minimum quality requirements. In the short run, this amounts to something akin to optimal monopoly regulation. However, it does create a long-run entry restriction. Furthermore, these guilds also often served as a local militia when needed. Apprentices were dimissed for reasons related to fitness, but never for lack of ability to perform the duties of their job. In addition, the length of apprenticeships seem to coincide with the typical ages observed under a military draft. The long-run entry restriction is a quantity-based restriction on capital accumulation as well as a form of compensation for guild members for serving as a local line of defense from external predation.

As Ames and Rapp (1977) note, since wars lead to very large, but temporary increases in expenditures, borrowing is a frequent source of emergency finance. Thompson (1995) argues that suspensions of gold convertibility during war and the subsequent resumption of convertibility at the previous par value were critical to emergency finance under the gold standard. The commitment to resumption at the previous par value anchors the price level. Thus, any expansion of the central bank balance sheet due to government borrowing results in a real expansion of the central bank balance sheet and allows the government to tap an adequate and open-ended source of financing. In contrast, the lack of commitment to the resumption of convertibility implies that an expansion of the central bank’s balance sheet will (eventually) lead to a reduction in money demand. Under this circumstance, the central bank’s expansion of the balance sheet in nominal terms results in a reduction of the size of the balance sheet in real terms since the price level increases by a greater magnitude than the balance sheet. Government borrowing is therefore subject to a borrowing constraint in the form of inflation. Hendrickson (2020) contrasts the experience of Sweden’s use of the early Riksbank as a source of emergency lending with Britain’s use of the Bank of England. In those early years, the Riksbank never adopted a policy of a temporary suspension of convertibility. As a result, the use of the Riksbank to finance wars was subject to various constraints. He finds evidence that military defeats were preceded by the binding of these borrowing constraints.

Other policies and institutions have been found to be consistent with the defense externality argument as well. For example, Batchelder and Sanchez (2013) examine the Spanish encomienda during the period of colonization. As part of this policy, the Spanish crown granted the right to conquistadors to temporarily collect a tribute from the native population. This policy devastated the native population. It is therefore hard to understand why this sort of policy persisted for such a long period of time. The authors argue that capital accumulation in the colonies of the New World created a defense externality for Spain. Capital within Spain was easier and less costly to defend than capital thousands of miles away. The encomienda can therefore be interpreted as a way to convert human capital abroad into durable domestic assets. They conclude that the atrocities committed against the native population in this process result, at least in part, from the Spanish crown’s attempt to deal with the defense externality.

Hendrickson (2019) argues that U.S. maritime policy is consistent with the defense externality argument. His argument is that some forms of capital might be complementary to defense. An example is capital employed in shipping. Since navies typically operate at a reduced capacity during peacetime, a state might benefit from being able to expand its sea-based operations quickly. Existing ships and trained crews engaged in private shipping have the ability to serve as a naval auxiliary during wartime. An optimal tax policy in this environment is for the government to tax shipping at a lower rate or to provide subsidies to these firms during peacetime in exchange for service during times of war. He notes that the long history of U.S. maritime policy is consistent with this objective. He also shows that similar subsidization policies are in place for other industries, such as aviation, which is capable of providing an air force auxiliary.

Finally, as it relates to defense, Rouanet (2023) presents a novel explanation of wartime price controls. In particular, he argues that wartime price controls and forced sales during the revolutionary period in France accomplished two things. First, they helped with the financing of the war effort. Second, the price controls and forced sales benefited urban consumers at the expense of rural producers, which generated important support for the war effort. Rouanet (2023, p. 19) supports this argument by presenting evidence that price controls “may have raised resources equivalent to 10% of GDP” and that the fate of the price controls were tied to the political influence of the residents of Paris.

My argument in this paper is also related to work that argues that under certain conditions the cost of enforcing the property rights of an asset might exceed the benefits (Allen, 2002). In such cases, the value of the asset might be reduced in order to enforce property rights. This argument has subsequently been used to explain human sacrifice (Leeson, 2014) and the practice of burying the dead with their valuable possessions (Harris & Kaiser, 2020).

This paper extends this literature on defense externalities by examining the characteristics of historical usury restrictions. In particular, I focus on the microfoundations of the restrictions and evaluate their ability to finance adequate defense in comparison to optimal capital taxation in the presence of a defense externality.

3 Usury restrictions and capital taxation

3.1 A model of optimal taxation and a comparison to usury restrictions

I assume that the purpose of the state is to provide defense. The model is concerned with the state’s ability to provide adequate defense. Let G denote a level of spending on the part of the state such that no conflict occurs in equilibrium.Footnote 3 A state that spends this amount is considered to be providing adequate defense. In addition, assume that the level of spending required to provide adequate defense is a function of the wealth owned within the borders of the state. This could be because adversaries covet this wealth or because defense will necessarily involve destruction of wealth. If the model is confined to typical neoclassical assumptions, wealth can be defined as equivalent to the capital stock. Given this description of adequate defense, one can write the cost of adequate defense as \(G = g(k)\), where k is the capital stock, \(g' > 0\), \(g'' \ge 0\), and \(g(0) = 0\).

In keeping with standard neoclassical assumptions, suppose that the only inputs to production are labor and capital and, for simplicity, assume that the labor supply is fixed.Footnote 4 It follows that aggregate production, y, can be written as \(y = f(k)\), where \(f', -f'' > 0\), \(f(0) = 0\), \(\lim \limits _{k \rightarrow 0} f'(k) = \infty\), and \(\lim \limits _{k \rightarrow \infty } f'(k) = 0\). The real interest rate is r and the rate of depreciation on capital is \(\delta \in (0,1)\). The government’s sole responsibility is to finance defense. The relative price of capital is constant and normalized to unity. Given these assumptions, we know the following conditions:

  1. 1.

    A benevolent social planner’s choice of capital satisfies:

    $$\begin{aligned} f'(k^*) = r + \delta + g'(k^*) \end{aligned}$$

    where \(k^*\) is the optimal capital stock. The condition states the the marginal product of capital is equal to the marginal private cost plus the marginal external cost.

  2. 2.

    In a laissez-faire competitive equilibrium, the following condition holds:

    $$\begin{aligned} f'(k_c) = r + \delta \end{aligned}$$

    where \(k_c\) is the equilibrium capital stock in a competitive market. The marginal product of capital is equal to the marginal private cost.

  3. 3.

    \(k_c > k^*\)

  4. 4.

    The optimal tax rate on capital, \(\tau\), is given as

    $$\begin{aligned} \tau ^* = g'(k^*) \end{aligned}$$

Conditions 3 and 4 follow immediately from the first two conditions. Condition 3 states that the laissez-faire equilibrium leads to the over-accumulation of capital given the externality that comes in the form of costs associated with enforcing and defending the property rights of those within the state. Condition 4 is the basic result that when there is a social cost above and beyond the private cost of production, the efficient solution is to levy a tax equal to the marginal external cost. Given that the cost of enforcing property rights is increasing in the amount of capital accumulated, it follows that the marginal tax rate on capital should be equal to the marginal cost of defending it. Define \(u:= r + \delta\) as the user cost of capital.

With this as a backdrop, consider a ruler who is incapable of levying a tax on capital, but who faces the same cost structure for providing adequate defense. Laissez-faire results in an over-accumulation of capital and a lack of revenue for the ruler. This leaves the territory prone to external predation and without a way to pay for defense of the territory’s capital stock. One would therefore expect rulers to experiment with rules, institutions, and policies that limit capital accumulation while simultaneously raising revenue for the government. The enforcement of usury restrictions in the Christian world often entailed a ban on lending, lump sum fees for approved lenders, and maximum interest rates for those that did lend.

One way to think about these restrictions is as though the ruler is charging a license fee for operating in the market as a monopoly. Once the monopoly is in place, the ruler can impose a price ceiling. The imposition of a binding price ceiling can conceivably limit the equilibrium quantity of capital in the same way as a tax. The license fee of the monopolist is the revenue generated to finance defense. Thus, one way to interpret these policies is as an attempt to replicate the efficient capital tax under the constraint of limited fiscal capacity. Nonetheless, the question is whether this is possible and under what conditions. To assess this I work backwards beginning with an existing monopolist and examine the ability of price controls to produce the optimal quantity of capital. I then examine the implications for revenue. I begin with the main result.

Proposition 1

Let \(k_m\) denote the equilibrium capital stock under a monopoly. If \(-k_m f''(k_m) > g'(k^*)\), then there is a maximum rental rate, \({\bar{r}}\), such that the user cost is \({\bar{u}} = {\bar{r}} + \delta\) and \(k = k^*\).

Proof

The inverse factor demand curve under laissez-faire is given as:

$$\begin{aligned} u = f'(k) \end{aligned}$$

Total revenue for the capital owner each period is therefore \(uk = kf'(k)\), and marginal revenue is given as

$$\begin{aligned} MR = f'(k) + kf''(k) \end{aligned}$$

The marginal cost of capital is \(r + \delta\). Thus, the monopolist’s choice of capital satisfies

$$\begin{aligned} f'(k_m) + k_m f''(k_m) = r + \delta \end{aligned}$$

where \(k_m\) is the monopolist’s capital stock.

The monopolist can then charge an interest rate, \(r_m\), such that \(u_m = r_m + \delta\). The choice of \(r_m\) and therefore \(u_m\) satisfies:

$$\begin{aligned} u_m = f'(k_m) \end{aligned}$$

Recall that the optimal choice of capital satisfies \(f'(k^*) = r + \delta + g'(k^*) \equiv u^*\). Since we know that a user cost, \(u^*\), is associated the optimal allocation, it follows that one could set a price ceiling for r such that \({\bar{u}} = u^*\). For this user cost to prevail, the price ceiling must be binding, or \({\bar{u}} < u_m\). This will be the case if \({\bar{u}} < f'(k_m)\). This will hold if

$$\begin{aligned} f'(k^*) < f'(k_m) \end{aligned}$$

From the equilibrium conditions, this inequality can be re-written as

$$\begin{aligned} r + \delta + g'(k^*) < r + \delta - k_mf''(k_m) \end{aligned}$$

Or,

$$\begin{aligned} g'(k^*) < -k_m f''(k_m) \end{aligned}$$

\(\square\)

What this proof demonstrates is that a maximum real interest rate can result in an optimal allocation of capital if the slope of the marginal revenue curve is sufficiently steep at the point in which marginal revenue equals marginal cost relative to the slope of the marginal defense cost curve at the optimal allocation of capital. In other words, this condition holds if the capital stock is sufficiently small (and the real interest rate is sufficiently high) under a monopoly. It therefore follows that one would be more likely to observe usury restrictions as a source of financing in less developed countries, which seems particularly relevant to the historical experience with such restrictions.Footnote 5

This point is illustrated graphically in Fig. 1. As shown in the figure, the equilibrium capital stock under laissez-faire is given as \(k_c\). The equilibrium capital stock with a monopoly is \(k_m\). Finally, the optimal capital stock, \(k^*\), is determined by the intersection of the factor demand curve and the marginal social cost curve, MSC. As shown in the figure, the interest rate is sufficiently high under the monopolist that a binding price ceiling can produce the optimal allocation of capital.

Fig. 1
figure 1

A Price Ceiling Suppose that \(f(k) = k^\alpha\), where \(\alpha \in (0,1)\). The inverse factor demand curve is \(u = \alpha k^{\alpha - 1}\) and labeled D. The marginal revenue curve is \(MR = \alpha k^{\alpha - 1} + \alpha (\alpha - 1) k^{\alpha - 1} = \alpha ^2 k^{\alpha -1}\). This is just a constant fraction, \(\alpha\), of demand. This curve is labeled MR. The marginal private cost (MPC) is \(r + \delta\). The marginal social cost (MSC) is the marginal private cost plus the marginal external cost \(r + \delta + g'(k)\). For simplicity, suppose that \(g'(k) = ak\), where a is some positive constant

Full size image

Assuming that the price control can implement the optimal allocation of capital, the ruler still needs to raise revenue in order to pay for the costs of defending this capital. One way to do this is to charge a fixed license fee, F, to the monopolist. This leads to the following result.

Corollary 1.1

Let F denote the license fee. The license fee must satisfy

$$\begin{aligned} F \le g'(k^*) k^* \end{aligned}$$

Proof

In order for the fee to be incentive feasible, the fee must not be larger than the monopoly profit. Formally, this implies that

$$\begin{aligned} F \le ({\bar{r}} - r) k^* \end{aligned}$$

where the term on the right-hand side measures the monopoly profit with a maximum interest rate equal to \({\bar{r}}\).

Now, note that if the price ceiling results in the optimal allocation, then it must be true that \({\bar{r}} + \delta = r + \delta + g'(k^*)\). It follows that \({\bar{r}} - r = g'(k^*)\). Thus,

$$\begin{aligned} F \le g'(k^*) k^* \end{aligned}$$

\(\square\)

A fixed licensing fee cannot exceed the economic profits earned by the monopoly faced with the price ceiling. Otherwise, the monopoly would never enter the market. Proposition 1 already established that a price control can result in the optimal allocation of capital, \(k^*\). Tax revenue under the optimal tax policy would be equal to \(g'(k^*) k^*\). Corollary 1.1 demonstrates that the fixed licensing fee can raise, at most, the same amount of revenue that the tax would generate.

There is one remaining issue that needs to be addressed. Whether or not a ruler chooses the optimal interest rate ceiling might depend on other competing incentives. For example, if raising or lowering the ceiling generates additional revenue, the ruler might have an incentive to deviate from optimal policy.

Corollary 1.2

A ruler can increase revenue by raising the price ceiling if the demand curve is sufficiently steep.

Proof

Recall that the licensing fee satisfies,

$$\begin{aligned} F \le ({\bar{r}} - r)k^* \end{aligned}$$

For simplicity, suppose that this is binding.

Let \(F_m\) denote the licensing fee charged to a monopoly in the absence of an interest rate ceiling. It follows from Corollary 1 that

$$\begin{aligned} F_m \le (r_m - r)k_m \end{aligned}$$

Again, supposing that this is binding, the ruler can earn more revenue without a price ceiling if \(F_m > F\). This will be true if

$$\begin{aligned} \frac{r_m - r}{{\bar{r}} - r} > \frac{k^*}{k_m} \end{aligned}$$

Or,

$$\begin{aligned} \frac{r_m - {\bar{r}}}{{\bar{r}} - r} > \frac{k^* - k_m}{k_m} \end{aligned}$$

\(\square\)

Corollary 1.3

A ruler can increase revenue by lowering the price ceiling if the demand curve is sufficiently flat.

The proof for Corollary 1.3 follows directly from the proof to Corollary 1.2.

A ruler who raises the interest rate ceiling might generate more revenue, but the equilibrium capital stock falls below the optimal level. This means that the society is less wealthy than its military ability would allow. If the ruler relies on other output-based means of revenue, then this amounts to a reduction in the ruler’s wealth. On the other hand, a ruler who can lower the price ceiling to generate additional revenue will allow too much capital accumulation. This welcomes outside threats and prevents the ruler from providing adequate defense.

3.2 Discussion

While the model outlines optimal policy in the context of a defense constraint and demonstrates how usury enforcement can partially replicate this outcome, it seems unreasonable to assume benevolent leaders who understood optimal tax policy in the context of this model. Nonetheless, it does seem important to explain how such policies might emerge in the context of the governance of pre-modern states. In this section, I provide a sketch of how these policies might have emerged and the implications of the theory. These implications allow me to consider historical examples in the context of the theory.

The theory of governance in this paper draws on the work of Adam Smith as well as (North & Thomas, 1973; Lane, 1975; Ames & Rapp, 1977; Olson, 1993), and (Morris, 2014), among others. The basic idea is that the government’s primary responsibility is protection from both internal and external predation, or what (Ames & Rapp, 1977) refer to as justice and defense, respectively. Taxes are compensation for this protection. Given the primary role of protection, one would expect that policies and institutions emerge to meet this end.

As Weingast (2018) notes, the idea that military competition influences the development of the state is an idea that dates back to Adam Smith. States that accumulate wealth make themselves a target for external predation. The incentive to accumulate wealth depends on the ability to protect that wealth. States that are unable to provide adequate protection will disappear. As Thompson’s (1974) model makes clear, this requires limiting the allocation of capital to that which the state can defend while simultaneously collecting sufficient tax revenue to finance that defense.

Scholars from Adam Smith to Morris (2014) have attributed the fall of the Roman Empire to its inability to defend itself from roving bandits. In the aftermath of the fall of the Roman Empire, feudal societies emerged. Adam Smith argued that the feudal system emerged and survived because it integrated the political system and the economic system with the military structure (Weingast 2018, p. 6). Or, put differently, the states that survived the selection process of plunder were those that developed institutions that aligned economic and political incentives with military objectives. In the terminology of Olson (1993), the roving bandits were replaced by stationary bandits. These stationary bandits differed from the roving bandits in the sense that roving bandits have an incentive to steal as much as possible. Roving bandits that are sufficiently strong, settle down and establish a monopoly on violence. They become stationary bandits. The stationary bandits can continue to extract resources in the form of taxation. However, in contrast to roving bandits, stationary bandits have an incentive to develop ways to provide both justice and defense for the public since they are the source of the bandit’s revenue.

The rise of cities undermined the feudal system. As Poggi (1979) argues, the rise of cities also meant the rise of city elites who challenged the existing feudal order. In some pre-modern states this led to the emergence of what Poggi (1979) refers to as the Ständestaat, or governance by the estates, as city elites gained some semblance of power alongside the existing feudal authorities. In some places, cities were granted their own right to governance in exchange for tax revenue paid to the king. In other cases, fully autonomous city-states emerged. Where cities were given some degree of autonomy, they also had to construct policies and institutions to provide adequate defense for the cities.

The examples used in Sect. 5 represent various stages of pre-modern state development. In the case of the English example, England had a feudal structure during the period that I examine. In Italy, largely autonomous cities are the subject of analysis. Finally, France had a feudal system of governance, but this system was not as centralized as was the case in England.

While much of this description focuses on the big picture aspects of the role of defense in the development of the state, there is also an extensive literature at the micro level of the particular policies, institutions, and tools developed by states to deal with war and defense.Footnote 6 The general idea that rulers had to provide adequate defense does not answer the question as to how rulers went about doing so. It is likely that rulers experimented with various policies and institutions and through a process of trial-and-error and the evolutionary selection mechanism of military defeat that certain policies and institutions survived. Nonetheless, as circumstances like advancements in military technology changed, one would expect that policies, institutions, and governance structures changed along with it.

My defense-based interpretation of usury restrictions is consistent with the following observations. War and defense-related public expenditures made up the majority of public spending. It is therefore plausible that defense considerations played a role in the design of policy beyond the simple desire for revenue and that the evolutionary process of military conflict selected for the sorts of policies that helped facilitate defense. Furthermore, while the costs of defense are ultimately paid with taxation, emergencies require the ability to access large sums of money in a short period of time. Tax smoothing arguments imply that rulers would prefer to borrow to finance these lumpy expenditures (Barro, 1979). This is particularly true during the Middle Ages when a ruler’s hold on power was sometimes tenuous and access to sources of credit was often limited. Granting monopolies to lenders therefore provided rulers with a source of funding in an emergency.

The use of usury restrictions seems to satisfy three basic criteria related to defense in that they (1) limit capital accumulation, (2) raise revenue, and (3) allow for emergency financing through loans. However, this would be true regardless of the intent of the policy. The importance of war and defense and the implementation of these policies might be a coincidence.

One way to test the theory would be to examine whether the evolutionary process of military conflict tended to select for these types of policies. However, clear examples of such selection are not always available. Instead, I will test my theory indirectly.

A ruler implementing usury restrictions as described above face two problems. First, at least in theory, the ruler can generate as much revenue from the usury restrictions as the optimal taxation. Of course, this result depends on share of the surplus the ruler can extract, given the well-known problems of negotiation between bilateral monopolies, and how much revenue can be generated by a tax given the level of state capacity. A ruler who wants or needs to generate more revenue might therefore invest in state capacity in order to collect the tax.Footnote 7 Second, the agreement between rulers and usurers is subject to a last period problem. In other words, in the final period that a usurer has a monopoly, the ruler has an incentive to confiscate the wealth of the monopoly usurer. Overall, what this suggests is that rulers are likely to end this system of usury restrictions in one of two different ways. The first is through an investment in state capacity. The second is confiscation.

In the next section, I develop a model of ruler behavior. The model determines the conditions under which a ruler currently implementing usury restrictions would either invest in state capacity or expropriate the wealth of the usurers. Subsequently, in Sect. 5, I provide historical examples of usury restrictions that are consistent with my tax model. I then examine the decline of usury restrictions in the context of my model of ruler behavior.

4 A model of ruler behavior

In this section, I will assume that a ruler has implemented usury restrictions consistent with the discussion in the previous section. I will then outline the conditions under which a ruler would choose to expropriate wealth and the conditions under which the ruler will choose to invest in a sufficient level of state capacity to implement a more direct tax on capital.

4.1 The ruler’s decision

Suppose that a ruler is implementing usury restrictions as outlined above. The ruler obtains revenue from the usury restrictions of F. The ruler also receives feudal revenue, P. The ruler offers protection to the usurers, which costs \(c_u\). Let \(V_u\) denote the value to the ruler of the state of the world in which the ruler implements usury restrictions. It follows that

$$\begin{aligned} V_u = \gamma \tau ^* k^* + P - c_u + \beta V_u \end{aligned}$$

where \(\beta \in (0,1)\) is the ruler’s discount factor, \(\tau ^* = g'(k^*)\) is the optimal tax rate, and \(\gamma \in (0,1]\) is the fraction of the optimal tax that can be collected through implementation of usury restrictions. Solving for \(V_u\) yields,

$$\begin{aligned} V_u = \frac{\gamma \tau ^* k^* + P - c_u}{1 - \beta } \end{aligned}$$

The ruler faces a decision about how to proceed with policy. The ruler could choose to continue with usury restrictions, invest in state capacity, or expropriate the wealth of the lenders. It is therefore important to consider these alternatives. The cost of investing in state capacity is \(\Omega \in {\mathbb {R}}^+\). In return, the ruler obtains tax revenue. The degree of state capacity is measured by \(\theta \in (0,1]\). In particular, let \(\theta \tau k^*\) denote the revenue generated from an optimal tax policy on capital. Here, \(\theta\) measures the fraction of tax revenue that actually makes its way to the ruler. Higher values of \(\theta\) are therefore associated with higher levels of state capacity. In implementing this tax, the ruler must make concession to the general public in order to collect the tax. The cost of those concessions is given as \(c_{\tau }\). Let \(V_{\tau }\) denote the value to the ruler of a state of the world in which the ruler implements a tax on capital. It follows that

$$\begin{aligned} V_{\tau } = \frac{\theta \tau ^* k^* + P - c_{\tau }}{1 - \beta } - \Omega \end{aligned}$$

Finally, the ruler can always choose to expropriate the wealth of the lenders and then re-implement the usury restrictions. In doing so, the ruler obtains all of the lender’s wealth. However, the lenders will only stick around and continue to lend under the re-instituted lending restrictions with probability \(\rho \in (0,1)\) and the ruler will collect the continuation value, \(V_u\), of the usury restrictions henceforth. It follows that

$$\begin{aligned} V_e = \chi k^* + P + \rho \beta V_u \end{aligned}$$

where \(\chi \in (0,1)\) is the fraction of the capital stock owned by the usurers.

I assume that each strategy is at least plausible in the sense that \(V_u \ge 0\), \(V_{\tau } \ge 0\), and \(V_e \ge 0\).

4.2 The ruler’s behavior

Proposition 2

The ruler will expropriate the wealth of lenders if

$$\begin{aligned} & k^*> \frac{(1 - \rho )\beta P - (1 - \rho \beta ) c_u}{\chi - \beta - (1 - \rho \beta ) \gamma \tau ^* } \\ & k^* > \frac{(1 - \rho ) \beta P + \rho \beta c_u - c_{\tau } - (1 - \beta )\Omega }{\chi - \beta - (\theta - \rho \beta \gamma )\tau ^*} \end{aligned}$$

A ruler will choose to expropriate the wealth of the lenders if this is the best policy option (i.e., \(V_e > Vu\) and \(V_e > V_{\tau }\)). Using the expressions for \(V_e\), \(V_u\), and \(V_{\tau }\), some algebraic manipulation yields the condition shown in Proposition 2. From this proposition, I can derive some important implications for ruler behavior.

There are two pretty straightforward results from Proposition 2. The first is that as feudal revenues decline, it makes it more likely that the condition will be satisfied. The reason is that when the flow of feudal revenues is low, the ruler might be tempted to expropriate all the wealth of the lenders in order to raise revenue. Weak rulers with insufficient feudal revenue and an inability to raise revenue from other sources might therefore resort to expropriating from the lenders. The second result is that a greater fraction of wealth owned by the usurers, the more likely it is that the ruler will expropriate wealth.

Expropriation is not the only policy option for the ruler. An alternative would be for the ruler to choose to invest in state capacity and switch to a capital tax. The following result outlines the conditions in which this type of investment will occur.

Proposition 3

The ruler will invest in the state capacity necessary to implement a tax on capital if:

$$\begin{aligned} \frac{(1 - \rho ) \beta P + \rho \beta c_u - c_{\tau } - (1 - \beta )\Omega }{\chi - \beta - (\theta - \rho \beta \gamma )\tau ^*}> k^* > \frac{c_{\tau } - c_u + (1 - \beta ) \Omega }{(\theta - \gamma ) \tau ^*} \end{aligned}$$

The condition in Proposition 3 follows from the fact that a ruler will prefer to invest in state capacity if \(V_\tau > V_u\) and \(V_\tau > V_e\). Investment in state capacity will bring tax revenue equal to \(\theta \tau ^* k^*\). The ruler will be more likely to invest in state capacity when the return on investment is high. We can think about this in terms of \(\theta\) and \(\Omega\). As \(\theta\) increases toward unity, there is less waste in the collection of taxes. As a result, the ruler has greater incentive to invest in state capacity. A lower cost of investing in state capacity, \(\Omega\), will also make it more likely that the ruler invests in state capacity. Similarly, more patient rulers will be more willing to bear the cost of investing in state capacity now in exchange for the flow of benefits that such an investment provides in the future. Finally, the ruler is more likely to implement this policy if the flow cost of concessions to the ruler’s subjects is lower.

From Propositions 2 and 3, there is another important implication that warrants discussion. Recall that a lower value of \(\gamma\) implies a smaller flow of revenue from usurers. From Propositions 2 and 3, a lower \(\gamma\) has an ambiguous effect on the policy switch since a decline in \(\gamma\) makes one inequality more likely to hold and one inequality less likely to hold in each proposition. Nonetheless, consider a scenario in which \(V_u = V_e = V_\tau\). A decline in \(\gamma\) would cause a direct decline in \(V_u\) and an indirect decline in \(V_e\) through its continuation value. Thus, it follows that when rulers are implementing usury restrictions, but are indifferent between all three policy choices, a lower \(\gamma\) would cause the ruler to invest in fiscal capacity. However, consider a scenario in which the ruler is indifferent between usury restrictions and expropriation, but strictly prefers both of these policies to an investment in state capacity, even after a decline in \(\gamma\). Since \(\frac{\partial V_e}{\partial \gamma } = \rho \beta \frac{\partial V_u}{\partial \gamma }\), it follows that the ruler would switch to expropriation. From Proposition 2, I know that the ruler would prefer expropriation when \(k^*\) is sufficiently large. Thus, for a given \(\chi\), the greater the equilibrium capital stock, the greater the wealth of the usurers and the more likely the ruler is to expropriate when the tax revenue from usurers is low.

Finally, it is possible that the ruler continues to implement usury restrictions.

Proposition 4

The ruler will continue to implement usury restrictions if

$$\begin{aligned} k^* \le \frac{c_{\tau } - c_u + (1-\beta ) \Omega }{\left( \theta - \gamma \right) \tau ^*} \\ k^* \le \frac{\beta (1 - \rho ) P - (1 - \rho \beta ) c_u}{\chi - \beta - (1-\rho \beta ) \gamma \tau ^*} \end{aligned}$$

This result follows from the fact that the ruler will favor the continued use of usury restrictions if \(V_u \ge V_e\) and \(V_u \ge V_\tau\). What this result shows is that if the optimal capital stock in the state is sufficiently low, the ruler will want to continue to use usury restrictions as a way to raise revenue. A more general result is that maintaining usury restrictions requires that the capital stock is sufficiently low. Since capital accumulation is limited by military technology, this implies that one should observe usury restrictions in states that are less developed both economically and in terms of state capacity.

5 Historical evidence

When examining usury restrictions in the pre-modern world, there is a general pattern that emerges. Usury restrictions primarily applied to those of the Catholic faith under canon law. However, rulers often made exceptions by granting various foreigners and non-Christians the ability to lend at interest. When doing so, rulers imposed maximum interest rates on loans. In addition, rulers levied lump-sum fees and/or taxes on the lenders. In some cases, these lump sum payments were assessed on a regular basis. In other cases, they were assessed as taxes when needed or through outright extortion. In many cases, these foreign and non-Christian groups were treated poorly by the population. As part of the system of usury restrictions, leaders often offered protection against such predation for these groups. Nonetheless, in many cases, these groups were often expelled from the country by the leader when their lending was no longer deemed necessary. In this section, I examine three examples. The first two examples consider the institutional structure of Jewish moneylending in England and Italy. The third example considers the experience of Italian (and to a lesser extent Jewish) moneylending in France.

5.1 England

From 1194 to 1275, England had an institution known as the Exchequer of the Jewry. The creation of this institution came in response to an attack on Jewish moneylenders years earlier in which the “apparent aim of the attack was to burn the records of debts owed to Jewish lenders” (Koyama 2010a, p. 382). The institution offered the king’s protection to Jewish lenders and created a record-keeping system of debts owed to Jewish lenders. In exchange, the king levied taxes on the Jewish moneylenders.

It is important to consider how this experience fits with my theory. Lending at interest was legally prohibited by canon law. Jewish moneylenders were not subject to such prohibitions. This meant that Jewish moneylenders had a collective monopoly on lending. While these lenders had a collective monopoly, they were subject to maximum interest rates (Barzel, 1992; Koyama, 2010b). The king also levied lump sum taxes on these lenders.

As I show with my theory, entry restrictions combined with a maximum interest rate can replicate the equilibrium quantity of capital that would emerge from an efficient tax on capital in the face of a defense externality. Lump sum license fees or taxes can then generate as much as revenue as the optimal tax on capital. If the purpose of the observed policy is consistent with my theory, then I would expect that this policy was implemented during a time when the state needed to finance its defense. Furthermore, my model of ruler behavior should help to explain what type of policy immediately followed the Exchequer of the Jewry.

The evidence supports this interpretation. Koyama (2010a) finds that overall government revenues fluctuated over this period primarily due to the cost of war. During the feudal era, rulers often assessed arbitrary taxes called tallages. This was the form of taxation on Jewish moneylenders. Consistent with my theory, these tallages were assessed as a lump sum tax. In total, these taxes did not make up a significant chunk of revenue for the state during normal times. Roth (1987) estimates that these taxes amount to about one-seventh of annual revenue. Stacey (1987) argues that this is somewhat of an understatement. Nonetheless, the data summarized in Koyama (2010a) supports the overall statement that this was a small fraction of overall revenue. There are various spikes in tallage assessments during the 13th century, all of which are in times of war. As Koyama (2010a, p. 385) notes, “the significance of the Exchequer lay less in the size of its contribution ...than in the ability it gave the king to levy huge discretionary taxes during crises.” This important role in financing defense is consistent with my theory.

The Exchequer of the Jewry ended in 1275 and Jews were expelled from England in 1290. My theory of ruler behavior suggests that expropriation will occur when there is a decline in the flow of tax revenue from usurers, if there is sufficient wealth to expropriate. The end of the Exchequer of the Jewry and the expulsion of the Jews fits with this prediction. The primary motivation for the elimination of the Exchequer of the Jewry appears to be a decline in revenue from Jewish lenders. As Koyama (2010a, p. 394) notes, “the tallage ordered by Henry III in 1272 revealed that the Jewish community was unable to provide tax revenues on the scale that it had done previously” and “the disappointing receipts of a subsequent tallage in 1275 made it possible for Edward I to negotiate an end to Jewish lending.” Subsequently, in 1290, Jews were expelled from England and their wealth confiscated. Furthermore, consistent with the evolutionary argument put forth above, it is important to note that the elimination of the Exchequer of the Jewry coincided with the rise of Parliament (Barzel and Kiser (1997), Koyama (2010a)).

5.2 Italy (Tuscany)

During the late thirteenth and early fourteenth century, there was a significant expansion of guilds throughout what is now Italy. Since members of these guilds were required to be Catholic, many Jews had to find a new source of income (Pascali 2016, p. 143). Those Jews who had previously accumulated sufficient wealth and knowledge as merchants, decided to switch to moneylending (Pascali 2016, p. 143). The subsequent analysis applies directly to Tuscany, for which explicit data is available.

Beginning in the late thirteenth century, Jewish lending was explicitly regulated by Italian towns. This lending continued to be significant until the early sixteenth century. The Catholic Church maintained its ban on lending at interest for all Christians during this period of time. Jewish lending was chartered by the local Italian towns. Since Jews were not subject to Catholic doctrine, they were exempt from these rules. As part of the charter, Jewish lenders were required to pay an annual lump sum tax and were subject to a maximum interest rate on their lending (Botticini 2000, p. 167). Evidence suggests that this price ceiling was binding (Botticini 2000, p. 177).

Jewish lending during this period is commonly referred to as pawnbroking since they accepted “movable objects – such as clothes, shoes, jewels, and working tools” as collateral (Botticini 2000, p. 168). However, borrowers were also able to borrow in exchange for a written promise to repay the loan or by designating a co-signer to the loan who would be responsible for paying in the event that the loan was not repaid. Contrary to what one might think given this description, borrowing from these Jewish lenders tended to be used for productive investments (Botticini 2000, p. 176).

Jews were not the only lenders during this period. Christian lenders existed as well. However, Christian lending was often disguised to circumvent usury laws. Furthermore, there were segmented markets for different types of credit (de Roover, 1966). Merchant bankers and money changers coexisted with the Jewish lenders, but were involved in different types of credit. Money changers were often concerned with foreign exchange. Merchant bankers primarily financed international trade and lent directly to governments. Within local communities, small loans were often given between individuals. However, the ability to borrow for private citizens on a magnitude needed for productive investment required using the Jewish moneylenders. Thus, the charters effectively gave Jewish moneylenders a monopoly over this form of lending.

According to the theory presented above, one should expect to find similar restrictions on the behavior of merchant bankers and money changers as on Jewish lenders in order to limit capital accumulation within the state. Koyama (2010b) makes the compelling case that the cost of devising strategies and acquiring the knowledge to circumvent usury laws among Christian lenders created a barrier to entry. In Italy, these entry restrictions were formalized by the creation of guilds for bankers and money changers. The long-run entry restrictions of the guilds served to restrict capital accumulation.

The theory outlined in this paper argues that the particular rationale for policies of a fixed licensing fee combined with a maximum interest rate is an imperfect attempt to replicate an optimal tax on capital in the presence of the defense externality imposed by capital accumulation. Furthermore, in the event of a defense emergency, the spending of the town on defense is likely to outpace its tax revenue. Some of the larger Italian city-states were likely to have access to merchant bankers for financing. However, smaller Italian towns were often obligated join in the defense of the larger city-states, yet would not necessarily have the same access to credit from merchant bankers. As such, Jewish lenders should be expected to play the dual role of financing emergency defense.

Botticini (2000, p. 166) argues that the tax revenue generated from Jewish lenders was used to finance war as well as grain subsidies during difficult times. Recall from Corollary 1.2 that if demand is sufficiently inelastic, rulers can increase revenue by raising the interest rate ceiling. Botticini (2000) examines the relationship between the interest rate ceiling and whether a city collected tax revenue from Jewish lenders. She also examines the relationship between the interest rate ceiling and whether Jewish lenders lent to the city government on favorable terms. In both cases, she finds a positive relationship between the fiscal importance of the Jewish moneylenders and the interest rate ceiling. Furthermore, as it relates to emergency financing, Botticini lists 22 loans obtained by the city of Perugia from Jewish lenders from the late thirteenth century to the early fifteenth century. Of these loans, 11 were related to defense and war.

The use of entry restrictions and what amounts to licensing fees appears to be a way to tax capital to finance defense in pre-modern Tuscany. My model of ruler behavior suggests that a ruler will switch to an explicit tax if revenue begins to decline and/or the cost of collecting a tax on capital declines. This lending began to decrease in importance in the early sixteenth century. The largest form of formal taxation that emerged during this period was in the form of excise taxes. However, with Florence and its surrounding cities engaged in frequent conflicts at the beginning of the fifteenth century, they had to find a new source of revenue. It is during this period that the region created what was its first wealth tax. The emergence of this tax was explicitly the result of the need to finance war (de Roover 1966, p. 23). Beginning in 1427, households were subject to the catasto, a tax levied on the value of property as well as any corresponding accumulated income. The initial tax was assessed as follows. A census was conducted. Each household had to declare their net worth, or the value of their property less any debts they owed. From this net worth, they were allowed to deduct living expense based on the number of people in the household. The first tax assessed was 0.5% of this amount (de Roover 1966, p. 25–6). Net worth was to be reported and taxed every 3 years. However, after the third catasto, the interval between tax events lengthened and was inconsistent. This tax stayed in place until the fall of the Medici in 1495 and replaced by a real estate tax (de Roover 1966, p. 24).

More recent work might also shed light on the switch from usury restrictions to the capital tax. Florence lost approximately 60 percent of its population to the Black Death in the middle of the 14th century (Jedwab et al., 2022). Recurrent outbreaks of the plague continued to reduce the population. A large exogenous change in the population increases the capital-to-labor ratio. Under standard neoclassical assumptions, this would be expected to increase real wages and reduce the real interest rate. Desierto and Koyama (2024) provide clear evidence of an inverse relationship between population and real wages during this period. This implies that the real interest rate correspondingly declined. This is important for two reasons. The first is that the decline in the real interest rate make the sort of usury restrictions described in this paper less effective. The second is that the arguments that pertain to the ruler apply as much to the capital stock as to the capital-to-labor ratio. As the model of ruler behavior demonstrates, if the optimal capital stock is sufficiently large, then the ruler has an incentive to invest in state capacity and implement a tax on capital. From the data presented in Desierto and Koyama (2024), real wages appear to peak in the second quarter of the 15th century. To the extent that this is driven by the decline in the population, a corresponding trough in the real interest rate should occur around the same period. Although it is difficult to compare interest rates across time from the Middle Ages, there is evidence of a decline in interest rates. For example, Homer and Sylla (2005, p. 104–105) write about the 15th century,

Historians offer evidence that the rate of interest was declining in Italy. One states that 5 - 8% “was now regarded as a fair interest rate on commercial loans.”

...Another historian also say that during this century rates on commercial loans in Italy fell: they declined from 12 to 10% and then to 5%. A third historian considers that the normal range of rates for merchants in Italian trading towns was still 7 - 15%. A fourth considers 5% to be indicative of the usual level of prime commercial rates in Venice.

They also note that for Florence, “interest on public debt was reduced by 15% in the thirteenth century to 10% and finally, by 1390, to 5%” (Homer and Sylla 2005, p. 93). It is not clear that this reflected a general decline in interest rates. Nonetheless, it is additional evidence of interest rate declines beginning in the 14th century.

Recall from Proposition 1 that the implementation of usury restrictions as an alternative to capital taxation requires that the capital stock is sufficiently small (or, correspondingly, that the real interest rate under the monopoly is sufficiently high). In the event that usury restrictions no longer replicate the optimal tax on capital, one would expect such restrictions to be abandoned in favor of an alternative. The introduction of the catasto in 1427 is therefore not only consistent with the timing shown in Desierto and Koyama, but also with my model.

5.3 France

The fourteenth-century experience in France is substantially different from the experience in England and Italy. Much of this had to do with the tenuous power of the French king. The king was “merely one of many political entities performing rudimentary state-like functions” (Henneman 1999, p. 101) and “exerted considerable authority in some regions outside this kingdom while having little effective control of certain regions within it” (Henneman 1971, p. 8). At this time, most of the French were of the opinion that the king should live off revenue from his domain, although many acknowledged the king had the power to raise revenue during times of war. Much of the first quarter of the fourteenth century was ruled by Philip IV and three of his sons. The second quarter of the fourteenth century was ruled by Philip IV’s nephew, Philip VI. A list of French kings and their regimes is listed in Table 1. Each regime seemed to oscillate between theories of the king’s fiscal authority (Henneman 1971, p. 27–39). Uncertainty over fiscal authority led to a number of ad hoc taxes and policies, including those aimed at moneylenders.

Table 1 Kings of France
Full size table

Usury was illegal in France during the fourteenth century, although it is not always clear what the term meant. During the reign of King Philip IV, which began in the late thirteenth century, the term referred to lending at interest above and beyond a maximum interest rate. However, at this time, the Catholic Church considered any lending at interest to be usurious and those caught engaging in usury were denied several Church services. These rules of the Church combined with the moral stigma of lending at interest apparently limited most lending to that done by foreigners. During his reign, King Philip IV was indebted to Jewish lenders as well as the Knights Templar. He expelled both groups in 1306. However, Jewish lenders were allowed back into France in 1315 under King Louis X. Italians also played an important role in both international trade and finance in France during the Middle Ages and were often the targets of usury enforcement.

During this period, the French were frequently at war and “the French crown found itself engaged in costly military ventures without a tax system capable of financing protracted wars” (Henneman 1969, p. 16). As a result, Italian (often referred to as “lombards”) and Jewish moneylenders were subjected to various taxes to support such ventures. However, unlike the previous examples, the French kings of this period often subjected these moneylenders to outright confiscation.

Prior to the fourteenth century, Italians (whether merchants or moneylenders) were subjected to relatively low taxes (Henneman, 1969). This changed in the 1300s, both for merchants and moneylenders. I will focus only on moneylenders.

Enforcement of usury restrictions during this period began in 1311, Philip IV re-enacted usury laws and declared legal maximum interest rates and an intent to enforce usury laws. Henneman (1969, p. 27) describes this policy as follows:

Philip had ordered all Italian usurers to leave the kingdom by November, 1311, paying the king what they owed him before they left. Unquestionably the crown wished only to collect a fine, for the Italian money-lenders were too valuable to be expelled completely. In mid-November Philip required Italians who were under arrest and demanded a hearing to appear in the immediate future before a royal court. It was there that they probably arranged to pay whatever finance the government demanded, for by the end of January, 1312, a new usury ordinance included a clause permitting the Italians previously expelled to reside in the kingdom under certain conditions.

While the manner in which this fine was collected was rather unorthodox, it is nonetheless consistent with the theory outlined above. The whole scenario seems an indirect way of extracting a fine for the right to lend from the Italian lenders.

In 1315, following the death of Philip IV, King Louis X raised money by imposing fines on all moneylenders as a measure of usury enforcement (Henneman 1969, p. 28). Subsequently, during the War of Saint-Sardos in 1324, King Charles IV used forced loans to raise money from Italian moneylenders. In 1325, Charles IV raised money by assessing lump sum fines on usurers. Up to this point, the enforcement of usury laws largely occurred during times of “war or political unrest” (Henneman 1969, p. 30). The fines were lump-sum. Both the fines and the defense-based policy are consistent with my theory.

However, beginning in the 1330s, policy towards moneylenders dramatically changed. During 1330, Philip VI declared that it was a crime to lend at a rate higher than 5 percent. The fine was large. Henneman (1969, p. 30) notes that the fine was the “most sizable extortion of the Italian lenders.” What made this remarkable is two-fold. First, the stated maximum of 5 percent made essentially all lending in France usurious. Second, this policy occurred during a time of peace.

Policy took another dramatic shift in 1331 when all debts to Italian lenders were cancelled with the principal owed directly to the king immediately (Henneman 1969, p. 31). This policy is dramatically different than both previous French policy and the theory I presented above. According to my theory, lump sum taxes on the monopoly rent of the usurious lenders is an alternative to optimal capital taxation. Cancelling the interest and seizing the principal is not a tax on monopoly rents, but rather direct confiscation.

In 1333, just three years after Philip VI declared 5 percent the maximum interest rate, he revised the maximum rate upward to nearly 22 percent (Henneman 1969, p. 31–2). This liberalization of policy was only temporary.

At the beginning of what is now known as the Hundred Years’ War in 1337, Philip VI ordered Italian moneylenders imprisoned. The evidence is unclear about what sort of punishment was actually extracted, but Henneman (1969, p. 32) suggests that debt cancellation occurred once again. The first major battle of the Hundred Years’ War was the Battle of Sluys in 1340. That same year, Philip VI confiscated all of the debts of Italian and Jewish moneylenders (Henneman 1969, p. 33).

In 1345, Italian moneylenders were threatened with being arrested again and subjected to a fine. Finally, beginning in 1347 and continuing until 1363, the French crown again confiscated the debts of usurers (Henneman 1969, p. 34).Footnote 8 Henneman (1969, p. 43) concludes that the “prolonged extortion of 1347–63, coinciding with the Black Death and serious military and political disturbances in France, must have destroyed the money-lending business of most Lombards in the kingdom.”

The experience in France differs substantially from the example of England or Italy. In Italy, Jewish lenders explicitly paid a licensing fee for the right to lend. In England, Jewish lenders were subject to varying levels of tallage depending on need to finance war. In both cases, these lenders knew that they were subject to these fees, regardless of the degree of uniformity. In each of those cases, the rulers levied what amounted to lump sum taxes on the monopoly rents of the lenders. This is consistent with the theory that I presented above. Early on, the French policy toward Italian and Jewish lenders appears similar to the English case. However, that ended with the explicit move towards extortion and what amounted to confiscating the principal on loans that began in the 1330s. According to my theory, the shift toward confiscation in 1330 should be associated with an inability of the French to adequately finance defense. This indeed appears to be the case.

The shift to confiscation is indicative of the sort of fiscal problems in France during the thirteenth century and the first half of the fourteenth century, culminating with the capture of King John II. Beginning in the thirteenth century, “The increased level of warfare caused royal expenses to soar” (Henneman 1999, p. 105). However, the French did not have any sort of fiscal infrastructure in place. In the feudal system of government, the king was expected to live off the revenues generated from his royal domain. Henneman (1999, p. 106) describes the general attitude of the French:

To the non-lawyer, the king’s power was based on the feudal-seigneurial tradition: he was the supreme suzerain of the realm. In this role, he could require aid in military emergencies, but the assistance could take the form of personal service. Earlier aide de l’ost or war subsidies had generally taken the form of finances in lieu of personal service, and beginning in 1302 Philip IV adopted this basis for levying taxes.

The term “war subsidies” refers to special taxes that the king could collect during times of war. The legal view of the time was that the king had the authority to collect these war subsidies. However, the French people tended to agree that this was the case only during “outright war” (Henneman 1999, p. 107). The difficulty in collecting these subsidies “meant that France was always unprepared militarily” (Henneman 1999, p. 110).

In 1345, in the midst of the beginning stages of the Hundred Years’ War, the French could not raise enough money. Evidently, this is what caused King Philip VI to fine usurers in 1345 and confiscate their debts in 1347. The inability to collect adequate revenue for the war seems to have been a driving factor. By going after usurers, the king could “collect large sums from the nobility without making an issue of their privileges” (Henneman 1971, p. 222). Overall, there is evidence that “Philip VI was encountering obstacles, not only in raising men and money but in enforcing essential orders” (Henneman 1971, p. 224). The French lost every battle between 1345 and 1347 and each of these obstacles seem to have contributed (Henneman 1971, p. 226–7).

A decade later in 1355, the Estates-General met in Paris at the request of King John II, Philip VI’s son, to raise money for the war. While the Estates came up with a plan to collect tax revenue, they wanted to do so at their own authority. This plan was challenged by local authorities, who were accustomed to collecting taxes on their own. The inability to raise money continued until John II was captured in battle. The effort to ransom the king resulted in more permanent taxes, including the fouage, a household tax levied based on the household’s wealth. The tax “was to pay military salaries ...by putting the most effective fighting men on royal payroll and creating, in effect, Europe’s first standing army” (Henneman 1999, p. 115). Despite the success of the military paid for by this tax, Charles V ended the collection of the tax shortly before his death in 1380. France struggled again to adequately finance defense until the 1430s (Henneman 1999, p. 117).

In short, the usury restrictions in pre-modern France differ significantly from those in England or Italy. The French experience with usury restrictions appears to be connected to the crown’s inability to raise significant revenue. In the absence of sufficient revenue for war, the crown resorted to outright confiscation and ultimately destroyed the moneylending industry at the time. Resorting to confiscation when feudal revenue is low is consistent with my model of ruler behavior. Similarly, the political conflicts between the crown and the nobility during this time might be indicative of a high discount rate on the part of the king, which also makes confiscation more likely. The inadequate financing for the war seems to have contributed to military defeats. In fact, the introduction of something akin to a wealth tax led to adequate financing and subsequent military victories. The subsequent elimination of this tax was followed by difficulty financing the military. The particular way of enforcing usury restrictions in France after 1330 proved to be an ineffective way of financing defense, much like other French taxation during this period.

6 Conclusion

Much of the literature on usury restrictions has focused on the motivations for such restrictions on the part of the Catholic Church or the implications of such restrictions on subsequent financial development. In this paper, I set out to explain why secular rulers enforced usury laws in much the same way across both time and space. In particular, I examine the microeconomic foundations of the pattern of usury restrictions often observed in Western Europe in the pre-modern period. I argue that the creation of entry restrictions and some form of lump-sum fine, license, or tax can be understood as a way to imperfectly replicate the optimal defense-based tax on capital.

It is important to recognize that this explanation does not require that rulers deliberately designed these policies to replicate a tax on capital. A confluence of factors led to the development of the particular type of usury restrictions observed over this period. First and foremost, the lack of fiscal capacity during this period meant that pre-modern states often had to experiment with ways to generate revenue. By granting monopoly lending privileges to Italian or Jewish lenders, this not only opened up a source of credit during the Commercial Revolution, but also provided access to credit for the rulers themselves. The monopoly rents made possible by the entry restrictions created an easy target for taxation. Italian and Jewish lenders tended to be unpopular. Thus, levying taxes and fines allowed rulers to raise revenue when other forms of taxation were either not feasible or subject to substantial resistance. Furthermore, these taxes or fines could be justified as the price of protection to these groups which were often subject to persecution.