Abstract

International trade can have profound effects on domestic institutions. We examine this proposition in the context of medieval Venice circa 800–1600. Early on, the growth of long-distance trade enriched a broad group of merchants who used their newfound economic muscle to push for constraints on the executive, that is, for the end of a de facto hereditary Doge in 1032 and the establishment of a parliament in 1172. The merchants also pushed for remarkably modern innovations in contracting institutions that facilitated long-distance trade, for example, the colleganza. However, starting in 1297, a small group of particularly wealthy merchants blocked political and economic competition: they made parliamentary participation hereditary and erected barriers to participation in the most lucrative aspects of long-distance trade. Over the next two centuries this led to a fundamental societal shift away from political openness, economic competition, and social mobility and toward political closure, extreme inequality, and social stratification. We document this oligarchization using a unique database on the names of 8,178 parliamentarians and their families’ use of the colleganza in the periods immediately before and after 1297. We then link these families to 6,959 marriages during 1400–1599 to document the use of marriage alliances to monopolize the galley trade. Monopolization led to the rise of extreme inequality, with those who were powerful before 1297 emerging as the undisputed winners.

I. Introduction

Venice has always presented two faces. As a great medieval trading center, its wealth was used to build not only beautiful architecture but also remarkably modern institutions. This is nowhere more obvious than in the Doge’s palace, whose grand Sala Maggiore housed a parliament (established in 1172) composed of the rich merchants that monitored and constrained most of the Doge’s activities. But after climbing up to the top floor of the palace, one enters the clandestine rooms of the secret service. With each passing decade after its establishment in 1310, this secret service was used to buttress the powers of a smaller and smaller number of families whose spectacular wealth was fed by international trade. This article tracks the evolution of Venice’s pre-1300 growth-enhancing institutional innovations and then the city’s dramatic post-1300 shift to political closure, social stratification, and extreme inequality at the top end. Our main thesis is that these developments were the outcome of the rise of international trade. International trade led to an increased demand for growth-enhancing inclusive institutions but also led to a shift in the distribution of income that eventually allowed a group of increasingly rich and powerful merchants to capture a large fraction of the rents from international trade.

Two strands of the literature are particularly relevant to this thesis, one dealing with medieval European trade (Greif 2006b) and the other with the Atlantic trade (Acemoglu, Johnson, and Robinson 2005). Medieval Europe experienced a massive expansion of long-distance trade during the Commercial Revolution of 950–1350 (see de Roover 1965; Lopez 1971; North and Thomas 1973). At the same time, medieval Europe embarked on a set of major institutional reforms that laid the groundwork for the rise of Western Europe. Greif (1992, 1994, 1995, 2005, 2006a,b,) establishes a causal connection between institutions and long-distance trade: Europe’s initial institutions facilitated the expansion of long-distance trade and, more important for our thesis, the resulting expansion of trade created a demand for novel trade- and growth-enhancing institutions. These included property rights protections that committed rulers not to prey on merchants (Greif, Milgrom, and Weingast 1994), a nascent Western legal system that included a corpus of commercial law known as the Law Merchant (Milgrom, North, and Weingast 1990), publicly provided monitoring and enforcement of commercial contracts (González de Lara 2008, 2011), and self-governing bodies such as business corporations. All of these are hallmark institutions of modern economic development.

Turning to early modern Europe, Acemoglu, Johnson, and Robinson (2005) show that where pre-1500 political institutions placed significant checks on the monarchy, the growth of the Atlantic trade strengthened merchant groups to the point where they were strong enough to further constrain the power of the monarchy. The English Civil War and the Glorious Revolution are the most famous examples (Jha 2010; Acemoglu and Robinson 2012, chapter 7). After 1800, this improvement in property rights raised urbanization rates and income per capita.1

The foregoing research is pervaded by two themes that will be important for our analysis. First, institutional change occurs not because it is efficient, but because it is advanced by powerful special interests.2 Second, as trade grows it affects the domestic distribution of income and hence the relative power of competing special interests. This change in relative power drives institutional change.3

To deepen our understanding of the effects of long-distance trade, via income distribution, on long-run institutional dynamics, we turn to a detailed historical and statistical examination of Venice. The broad outlines of Venetian history that we use to support our thesis are as follows. Through “a series of fortuitous events” in the ninth century, Venice became politically independent (Cessi 1966, p. 261). Together with Venice’s unique geography, this positioned it to benefit from rising trade between Western Europe and the Levant. These two factors combined to enrich Venetian merchants, who used their newfound economic muscle to push for institutional change.

The two key dates for improvements in institutions that constrained the power of the executive are 1032, which marks the end of a de facto hereditary dogeship, and 1172, which marks the establishment of a Venetian parliament that became the ultimate source of political legitimacy. Contracting institutions also displayed extraordinary dynamism during the Commercial Revolution, in part to deal with the commitment and enforcement problems that come with doing business abroad (Milgrom, North, and Weingast 1990; Greif, Milgrom, and Weingast 1994; Greif 2006b) but also to deal with the unique demands placed on capital markets by long-distance seaborne trade. This risky trade required large capital outlays, and this in turn led to the development of new business forms and legal innovations that supported the mobilization and allocation of capital. One particularly famous innovation was the limited liability contract known as the colleganza in Venice and the commenda elsewhere in Europe. It was the direct precursor of the great joint stock companies of a later period. Importantly for our thesis, it allowed even relatively poor merchants—who had neither capital nor collateral—to engage in long-distance trade and profit from it.

These institutional improvements made Venice wealthier overall, and also led to other substantial changes in the Venetian distribution of income. For one thing, the riskiness of trade together with the widespread involvement of Venetians in this trade created a great deal of income churning—mostly rags to riches but also some riches to rags. For another thing, a small group of merchant families grew spectacularly wealthy.

This brings us to the great puzzle of Venetian history. During the period 1297–1323, a defining epoch in Venetian history known as the Serrata or “closure,” Venetian politics came under the control of a tightly knit cabal of the richest families. It was, in Norwich’s (1977, p. 181) words, the triumph of the oligarchs. Furthermore, by the early 1330s this political closure had spilled over into an economic closure that excluded poorer families from participation in the most lucrative aspects of international trade. Finally, by 1400 the political and economic closure had created a society characterized by a new emphasis on rank and hierarchy. In short, after 1323 there was a fundamental societal shift away from political openness, economic competition, and social mobility and toward political closure, extreme inequality, and social stratification.

To understand this puzzle, we construct a model that highlights the key role played by the evolution of income distribution. To this end, we introduce political and coercive institutions into a version of the Banerjee and Newman (1993) framework in which wealth dynamics are driven by occupational choice under wealth constraints (see also Galor and Zeira 1993).4 In our model, as was the case in medieval Venice, political power is tied to mercantile wealth. Along the model’s equilibrium path there is economic and political mobility until the wealthiest merchants are powerful enough as a group to restrict entry into political markets. However, long-distance trade continues to generate wealth for up-and-coming merchants, which poses a political and economic threat to the wealthiest merchants. To prevent this without triggering a revolt, the wealthiest merchants co-opt the nouveau riche by allowing them into the Great Council. This larger coalition then restricts participation in long-distance trade to Great Council members. Barriers to entry into both political and economic markets are erected. The resulting evolution of the distribution of income (and hence of coercive power) permanently supports this outcome.

We show empirically how this replicates the sequence of historical events associated with the Serrata of 1297–1323. The key outcome of the Serrata was the creation of a hereditary nobility that had the exclusive right to sit in the Great Council and used this right to restrict participation in long-distance trade. To deepen our understanding of the Serrata, we develop a database of the 8,178 elected members of the Great Council in the period immediately preceding the Serrata (1261–1296). We use this to show that mobility was indeed eroding the political position of the wealthiest families. In particular, they were losing seats to up-and-coming families who had not previously been involved in politics. Building on Kedar (1976) and González de Lara (2008), we code up hundreds of colleganza contracts for long-distance trade that have survived from the period 1073–1342. We use these to show that economic restrictions enacted during the Serrata were effective not only in restricting the use of the colleganza to the newly created nobility but in restricting it to the most powerful of these nobles.5 We then turn to the galley trade, the most lucrative aspect of long-distance trade. After the Serrata, control over state-sponsored galley convoys was restricted to nobles. To finance them, nobles abandoned the colleganza in favor of within-family financing and marriage alliances with other wealthy noble families. We track 6,959 noble marriages recorded during 1400–1599 using techniques from social network theory (Jackson 2008). We show that families who dominated the post-Serrata galley trade were the most important families in the marriage network (as measured by eigenvector centrality). We also show that these same families dominated the Great Council during 1261–1296. Thus, those who were powerful before the Serrata emerged from it as the undisputed economic, political, and social victors.

This article has three points of contact with the vast literature on Venetian history. First, the article deals with the so-called myth of Venice. In its strongest form, the myth states that the civic-minded Venetian patriciate acted selflessly in the interest of all Venetians and that the Serrata was not a major point of discontinuity. See Grubb (1986) and Martin and Romano (2000) for reviews of the literature. The myth has faced a number of criticisms, of which this article is one. Our post-Serrata analysis borrows threads from influential studies by Queller (1986) and Ruggiero (1980) discussed below, while our interpretation of the 1297–1323 Serrata is closest to Cracco (1967). However, we agree with Lane (1968) that Cracco’s emphasis on class struggle is misplaced and that more evidence is required to support his thesis. Rather than appeal to class struggle, we focus on special interest politics and institutions as in Acemoglu and Robinson (2006) and Greif (2008). Second, the article deals with Venetian social history. We integrate the late fourteenth century social transformation emphasized by Romano (1987) and Chojnacki (1997) into our broader thesis. Third, this article has implications for the literature on Venice’s long-term stability, particularly Venice’s ability to put a lid on interclan rivalries (e.g., Lane 1971, pp. 259–260; Greif 2006b, section 6.4.2; González de Lara, Greif, and Jha 2008). Our analysis of the distribution of economic rents complements Greif’s as well as González de Lara’s (2008, 2010, 2011) analyses of the self-reinforcing nature of constraints on the Doge and the role of policies to sustain rents from international trade. Note that because the very important issue of interclan rivalry is dealt with by these authors, we have little to say about it here. Finally, what sets us apart from the existing literature is our central thesis, namely, that international trade had profound long-term impacts, via wealth distribution, on domestic institutions. We support this thesis with systematic evidence covering eight centuries and tracking Venetian families’ political representation, involvement in international trade, and intermarriage.

Sections II–III review constraints on the executive and the rise of contracting institutions during Venice’s early history. Section IV presents the model. Section V reviews the Serrata and presents our empirical results about political mobility and the use of the colleganza. Section VI reviews the post-Serrata galley trade and our empirical results about marriage alliances and inequality. Section VII concludes.

II. The Rise of Constraints on the Executive

Throughout the ninth and first half of the tenth centuries, Venice experienced a slow revival of long-distance trade (McCormick 2001, pp. 630–638; Findlay and O’Rourke 2008, p. 84). This trade required Venetian merchants to cooperate in mobilizing resources, and in this period we already see numerous examples of Venetian convoys traveling throughout the Mediterranean (McCormick 2001, pp. 523–529). Furthermore, Venetian naval strength was growing. Venetian navies fought the Arabs in southern Italy in 827, 828, 840, and 842, though often unsuccessfully. However, by the 860s, Venetian naval power had become an effective deterrent to Arab naval actions (Nicol 1988, pp. 26–33). Explaining the origins of this success in collectively mobilizing resources is beyond the scope of our article because it would require both cross-cultural and cross-regional comparisons. We therefore restrict ourselves to two conjectures. First, (Greif 2006b, pp. 25–26) argues that when comparing Western Europe to the Islamic world, Western society made more allowance for individualistic as opposed to kin-based organizations, legitimized these organizations without appeal to religious authority, and thrived under the radar of relatively weak as opposed to strong states. All of these factors are pertinent to Venetian success.6 Second, unlike many Western European cities, the geography of the Venetian lagoon created an environment that discouraged agriculture and encouraged seaborne trade. Success in the latter required Venetians to cooperate among themselves.7

II.A. Constitutional Change I: The End of Hereditary Doges (810–1032)

Long-distance trade picked up substantially in the second half of the tenth century as a result of events in Western Europe and the eastern Mediterranean. To the west, rising incomes led to a resurgence of trade, especially along the Rhine and Danube Rivers north of Venice and the Po Valley emptying into the Venetian lagoon.8 To the east, between 961 and 969, a resurgent Byzantium regained control of the eastern Mediterranean sea lanes, notably conquering Crete and Cyprus. As Pryor (1988, p. 111) writes of these conquests: “Christian reconquest of the Muslim possessions along the trunk [main shipping] routes in the tenth and eleventh centuries laid the foundations for the later Western domination of those routes, with all that implied. The reconquest thus appears as one of the most fundamentally important historical processes in Mediterranean history.” Larger scale trade between Venice, Constantinople, and the Levant quickly reemerged. Figure I shows the main eastern Mediterranean trade routes.

Figure I

Medieval Shipping Lanes of the Eastern Mediterranean

The rise of long-distance trade had an important implication for Venice: it allowed a relatively large number of merchants to become rich and demand civic recognition. Evidence of this can be gleaned from the lists of endorsers of dogal documents. Endorsing a dogal document was a sign of having arrived in society. In the second half of the tenth century the number of endorsers per document increased considerably. Castagnetti (1992a,b) has carefully tracked the names appearing in three extant Venetian dogal documents from 960, 971, and 982. These documents were endorsed by 65 people in 960, 80 people in 971, and 128 people in 982. More interestingly, the percentage of endorsers belonging to families whose names had never before appeared in any Venetian document is high, averaging 59%.9

While these newly rich merchant families were not individually powerful, within 60 years of reopening the Mediterranean sea lanes to Christian shipping, they were collectively powerful enough to significantly constrain the power of the Doge. To analyze this process, one must bear in mind that dogal institutions in this period present two faces. On one hand, Doges were weak in that they were elected and often murdered or forced into retirement by their opponents. They were not autocrats. See, for example, Greif (1995, p. 738).10 On the other hand, Doges had wide-ranging powers that no other Venetian commanded. Cessi (1966, p. 270) describes the dogal system of the time as “quasi-tyrannical,” and Lane (1973, p. 90) writes that “the Doge was a monarch of unlimited power.”

These two contradictory faces of dogal power pose a problem for us. To analyze constraints on the executive we must first establish that the executive was in fact at least somewhat powerful, even if not as powerful as claimed by Cessi and Lane. To do so we focus on one of the more important powers that a monarch can have, namely, the ability to appoint a family member as successor. Specifically, we consider succession from 810 (when the first Doge recognized by Byzantium was elected) until 1328 (when the last Doge of the Serrata period died).

Figure II illustrates the dynastic connections among Venetian Doges from 810 to 1328.11 Time is measured horizontally, and the length of each box corresponds to the length of the term in office of one Doge. For each Doge we go back in time to his most recent predecessor with whom he had a family connection. Curves above the box mark connections between father and either son or brother. Curves below the box mark connections involving a son-in-law or nephew. We break the Figure II bars at the two key constitutional crises of 1032 and 1172. We define a dynasty as a set of Doges who pass on the Dogeship within the family at least twice. In the figure, we mark each dynasty with a distinct color. There are three dynasties between 810 and the introduction of the first constraints on the Doge in 1032. The striking fact is that during this period, every Doge had a direct family relationship to another Doge and most Doges belonged to one of three dynasties.

Figure II

Dogal Dynasties

The first dynasty, the Participazio, consists of Agnello Participazio, his sons Giustiniano and Giovanni Participazio, as well as Pietro Tradonico, who had married into the Participazio family, and Pietro’s nephew Domenico Tribuno. The four boxes that are not colored in the figure in this early period are also Participazio, but it is not clear from contemporary sources whether they were related to the earlier Participazio.12 The second dynasty, the Candianos, held the Dogeship for four successive generations (Pietro Candiano i, his son Pietro Candiano ii, his grandson Pietro Candiano iii, and his great-grandson Pietro Candiano iv). This was followed by Pietro Candiano iv’s brother (Vitale Candiano) and son-in-law (Tribuno Menio). The Orseolo were the third and final dynasty of the period. Doge Pietro Orseolo i was succeeded by his son Pietro Orseolo ii in 991, who in turn was succeeded by his son Otto in 1009. As was common for Doges, Otto used his position to appoint brothers to the most important church positions. One brother was appointed head of the Venetian church (patriarch of Grado), and another was appointed to a rank just below this (bishop of Torcello). In 1026, an already unpopular Otto blocked the appointment of a Flabanico family member to an important church position (bishop of Olivolo), which sparked a successful revolt led by Domenico Flabanico and resulted in Otto’s exile. Otto almost regained power during 1031–1032, but Domenico Flabanico prevailed and became Doge in 1032.

The election of Flabanico as Doge was a transformative moment in Venetian history. He was a wealthy silk merchant, and most subsequent Doges over the next many centuries were also merchants involved in long-distance trade. Flabanico’s election thus represents the triumph of the merchants. Further, his reign ushered in two de facto constitutional innovations that significantly constrained the powers of Doges. First, the election of the Doge was to be respected in full: a Doge would no longer be allowed to appoint his successor. Second, Doges were henceforth required to consult with a two-member dogal court of judges and abide by the court’s decisions (see Lane 1973, p. 90; Cessi 1966, p. 263 and 270). The constitutional principles embodied in these changes were not new—in principle, a Doge’s successor was elected rather than appointed and was accountable to judges. What was new was the willingness of subsequent merchant Doges to respect constitutional principles. This willingness is apparent in Figure II. Comparing the period 810–1032 with 1032–1328, there is a dramatic fall in the number of dogal successions, that is, in the number of lines connecting boxes. Furthermore, there is only a single dynasty after 1032.

The driving force behind the 1032 constraints on the executive was long-distance trade and the broad-based economic and political power it brought to a growing group of merchants. It is no coincidence that the reforms came relatively quickly after the opening up of eastern Mediterranean sea lanes to Christian shipping.13

II.B. Constitutional Change II: The Establishment of a Parliament (1032–1172)

From 969 on, Venetian long-distance trade expanded steadily; however, the growth of trade accelerated after 1082. In that year, a remarkable confluence of events on distant shores propelled Venice into a period of unprecedented prosperity and globalization.

In 1071–1081, Constantinople was again in decline, weakened by Seljuk Turks in Anatolia and facing invasion by the Norman kingdom in southern Italy. A desperate Byzantium enlisted Venetian naval aid to stop the Normans’ Adriatic crossing. Byzantium was in dire straights, and Venice extracted a heavy price for its naval involvement. The Golden Bull of 1082 granted Venice duty-free access to 23 of the most important Byzantine ports and granted Venetian merchants property rights protections from the caprices of corrupt Byzantine administrators. Most important, the Venetians were given buildings and wharfs within Constantinople. Venetians thus became the first foreign traders in Constantinople to have their own quarter. See Brown (1920) for an English translation of the Golden Bull and for details of the Venetian quarter.

The 1082 Golden Bull was crucial to Venice’s maritime expansion. As a result, the number of newly rich merchants who were eager for civic participation increased. In the 1090s we again see a jump in the number of new names endorsing dogal documents (see Castagnetti 1992b and especially Castagnetti 1992a, pp. 625–626 and 637–638).14 The time was ripe for merchants to once again flex their political muscle.

After the reign of four unrelated and long-lived Doges (see the middle bar in Figure II), the Michiel family held the Dogeship for 53 of the 75 years leading up to 1171.15 Toward the end of this period, Venetian–Byzantine relations had become increasingly acrimonious, and tensions came to a head on the night of March 12, 1171, when the Byzantine emperor rounded up 10,000 Venetians residing in the empire and announced that they were being held for ransom.16 In September 1171, Doge Vitale Michiel ii launched a large armada that was to blockade and harass Constantinople until the hostages were released. The plan failed miserably, and in May 1172 the fleet returned in utter disarray. Venetian frustration was palpable, and much of it was directed against the Doge. At a gathering on May 27, he was mobbed and assassinated. It had been almost two centuries since a Doge had been murdered, and the unexpected assassination left a power vacuum which the dogal court and leading merchant families immediately filled. As in Jones and Olken (2009), the assassination of a powerful ruler produced a transition to a less autocratic regime.

The first major change was the introduction of a limited-franchise elected parliament known as the Great Council.17 With this constitutional change in place, the new legislative body used its power to increasingly constrain the power of the Doge over the next few decades. Many of these constraints were formalized in the oath of office that the Doge now publicly swore to uphold. The oath explicitly listed what the Doge could not do, for example, expropriate state property or preside over cases against himself. The Great Council added to this list with the election of each new Doge (Hazlitt 1966, p. 437; Madden 2003, pp. 95–101). Furthermore, in all important decisions the Doge was required to consult with a strengthened six-member dogal council that was elected by and accountable to the Great Council. As Madden (2003, p. 98) notes: “In short, by 1192 the doge could do almost nothing without approval of the council.”18

The establishment of the Great Council and the constraints imposed by the dogal oath of office were major institutional innovations. For Norwich (1977, p. 90) these were “arguably the most important reforms in Venetian history.” They dramatically limited the power of the Doge and arrogated his powers to a large group of families who owed their wealth and power to long-distance trade.

III. Institutional Change: The Rise of Contracting Institutions and Inclusive Growth

The two centuries following 1082 were ones of extraordinary dynamism for contracting institutions. By the early fourteenth century, financial innovations included: the appearance of limited-liability business forms; thick markets for debt (especially bills of exchange); secondary markets for a wide variety of debt, equity, and mortgage instruments; bankruptcy laws that distinguished illiquidity from insolvency; double-entry accounting methods; business education (including the use of algebra for currency conversions); deposit banking; and a reliable medium of exchange (the Venetian ducat). All these innovations can be related directly back to the demands of long-distance trade.19

Equally important is the development of a supporting legal and enforcement framework. The most discussed of these is the Law Merchant, which is universally accepted as the foundation of modern commercial law (Berman 1983). Its very scope—the use of a court of peers to adjudicate commercial disputes between merchants traveling in distant lands—means that the Law Merchant was a direct and immediate response to the needs of long-distance trade (Kadens 2004).20 The Commercial Revolution is also viewed as a key driver of the development of the modern Western legal tradition. This tradition has its origins in a legal revolution that occurred in the period 1075–1122 (Berman 1983; Landau 2004). While a general discussion of the origins of this legal tradition is outside the scope of this article, a comment on timing is appropriate. Civil law was not in use anywhere in Europe in 1000 (Radding and Ciaralli 2006), but reemerged in Europe in the second half of the twelfth century when communes began writing statutes governing their constitutions and commerce (Landau 2004). Second, the first half of the twelfth century witnessed an explosion of secular legal documents. Such documents were rare in 1050 but common by 1150.21 Venetian law developed rapidly thereafter: its codification was begun under Doge Enrico Dandolo (1192–1205) and completed in 1242 by Doge Jacopo Tiepolo (1229–1249). See Besta and Predelli (1901).

Thus, civil law and commercial documents both appear just after long-distance trade began its explosive growth. There were other developments in Venetian contracting institutions in this period. See, for example, González de Lara (2008, 2011). Here we simply conclude that the expansion of trade after 1082 was accompanied, especially toward the end of the twelfth century, by a remarkable set of innovations in contracting institutions.

III.A. The Colleganza as an Institutional Response to the Demands of Long-Distance Seaborne Trade

We now take an in-depth look at one particular contracting innovation—the colleganza. This was a predecessor of the joint-stock company and is viewed by economic historians as one of the key commercial innovations of medieval times, if not the key innovation (e.g. Lopez and Raymond 1967, p. 174). Our main aim is to draw out the implications of long-distance trade for the evolution of income distribution and set the stage for the empirical work to come. For further details on the colleganza and comparisons with other contemporary commercial contracts, see Lopez and Raymond (1967), Pryor (1987), and González de Lara (2010).

To understand why the colleganza was such an innovation, one must first understand the mechanics of long-distance trade. Ships typically left Venice at the end of March when the winter storm season was finished and the prevailing winds had turned favorable. If all went well, ships arrived in Constantinople by the end of April, spent three weeks collecting merchandise for the voyage home, and arrived back in Venice by July. The goods brought back were then sold to merchants traveling to the late summer fairs in Central and Western Europe. See Lane (1966, 1973, pp. 69–70). Such a trip, if on schedule, would have earned enormous profits—over 100% and sometimes much more. Although big returns could be had, there were also big risks. Death abroad from illness, shipwrecks, and piracy were common. There was also substantial business risk associated with the thinness of markets. Ships often traveled from port to port for months while merchants searched the hinterland for merchandise. A merchant who arrived in Acre a month late might find that the market was over for the year and be forced to dump his goods at fire sale prices. Thus, luck and also the business skills and effort of a traveling merchant could make the difference between huge profits and huge losses.

The colleganza was a solution to three key problems of long-distance trade. First, this trade required large amounts of capital relative to most other contemporary private commercial activities, such as agriculture or manufacturing. Second, collateral was problematic because, unlike in agriculture or manufacturing, the capital literally sailed out of sight. Third, the complex unforeseeable circumstances and large risks involved required balancing high-powered incentives for traveling merchants with risk sharing between them and the investors.

Although there were many variants of the colleganza, we describe only the simplest and most common of these. There are two parties, the traveling merchant and the investor (or sedentary merchant). In Venice, the sedentary merchant gives cash or wares to the traveling merchant, who then boards a ship with other merchants for an overseas destination, say, Constantinople. In Constantinople, the traveling merchant sells the wares and uses the proceeds to buy other wares for resale in Venice. A colleganza specifies the names of the two parties, itemizes the capital contributed by the sedentary merchant and/or gives it a value (this is the “joint stock”), and states how profits will be split. The contract sometimes provides specific instructions, for instance, an itinerary of ports to be visited, but very often leaves the traveling merchant a very high degree of freedom. Once the traveling merchant brings or sends the wares back to Venice, the accounts of the voyage are settled and the relationship is dissolved. In the archetypical colleganza, the sedentary merchant provides all the capital and receives 75% of the profits. The traveling merchant contributes no capital and receives 25% of the profits. If there are losses, these come out of the sedentary merchant’s capital. However, the sedentary merchant’s obligations are limited by his initial investment. Restated, the colleganza provides limited liability and, specifically, the liability is limited to the joint stock specified in the contract. This was a major innovation over Roman law and is widely recognized as the origins of the great joint stock companies of a later period.

Figure III provides a typical example. The sedentary merchant, Giovanni Agadi, puts up the joint stock of 300 pounds of Venetian pennies, an unimaginable sum for an ordinary Venetian. The traveling merchant, Zaccaria Stagnario, is to board a privately owned ship that will travel in convoy to Constantinople. No other commercial instructions are given: Stagnario is in charge of all other decisions (including continuing his voyage to “any other place that seems good to me”) and this is why high-powered, profit-sharing incentives are needed. The profit split is expressed in fractions: three fourths for the sedentary merchant and one fourth for the traveling merchant. If instead of profits there are losses, this downside risk is entirely borne by the sedentary merchant: “at your own risk from sea and people.” The traveling merchant faces stiff penalties for failure to pay back the sedentary merchant.22

Figure III

Colleganza between Zaccaria Stagnario and Giovanni Agadi, August 1199

Original parchment document (left), reproduced with permission from Archivio di Stato di Venezia (document identifier: San Zaccaria, Pergamene, b. 35 pergg., n. 160). Authors' translation from the Latin (right), based on the transcription of the parchment in (Morozzo della Rocca and Lombardo, 1940, document 444).

There is much that is not specified in this contract, so much so that the contract is hard to understand except in the context of supporting institutions that developed to support merchants traveling in colleganza. This point comes out in Pryor (1987, chapters 3 and 4), who reviews the resolution of hundreds of colleganza disputes to flesh out the full set of “rules of the game” surrounding the colleganza. In addition, González de Lara (2008, 2010, 2011) reviews the private and public institutions that supported the colleganza in Venice in the thirteenth century. Thus, the colleganza is not just a contract, it is an innovation that created a demand for other supporting institutions.

III.B. Economic and Political Mobility: The Role of the Colleganza

The discussion of this section has emphasized that long-distance trade was exceptionally complex and risky and could make or break a merchant. It has also emphasized that the institutional response—the colleganza—allowed poor merchants to enter the game. Indeed, most historians have commented on this feature of the colleganza, for example, de Roover (1965, p. 51), who writes: “In a great many cases, the tractores [traveling merchants] were ambitious young men who were willing to take heavy risks in order to accumulate sufficient capital to join eventually the ranks of the stantes [sedentary merchants].”

As a result of the widespread engagement of the population in long-distance trade and the economic mobility it entailed, newly rich merchants flowed into political power throughout the twelfth and thirteenth centuries. This is a famous feature of Venetian society at this time. See Hazlitt (1966, p. 216), Lane (1973, p. 20, 89–90), Norwich (1977, pp. 182–183), Ruggiero (1980, p. 4), and Lopez (1971, pp. 67–68, 70). In Madden’s (2003, p. 3) words, “the membrane of Venetian nobility was permeable. Indeed, nobility in the sense of a group of families with a hereditary claim to political authority did not exist at all. In Venice, wealth, not land, defined nobility; commercial skill, not military prowess.”

The life of Zaccaria Stagnario provides an example of the economic and political mobility that was possible at this time. His grandfather Dobramiro was a Croatian slave who was freed when his Venetian owner died. His father, Pancrazio, was a helmsman. In 1199 we find Zaccaria traveling in colleganza to Constantinople (this is the document we reproduced and translated in Figure III), and this experience paid off handsomely when he moved there after the 1204 conquest. By 1207 he held office as councilor to the first Venetian podestà in Constantinople and was rich enough to be a sedentary merchant in two colleganzas for the large sum of 200 Byzantine hyperpeppers, an amount equal to seven months’ salary of the Duke of Crete. Ironically for the grandson of a slave, these colleganza were for travel to the Black Sea fortress of Sudak, a slave-trading center. Upon his return to Venice, Stagnario integrated himself into the highest social and political circles. In the words of Robbert (1999, p. 35): “Zaccaria, the grandson of a slave, represented the new man in Venice who climbed to the top because of his business skills.”23

III.C. Colonial Empire and Nobility Rents, 1082–1297

On April 12, 1204, the blind Doge Enrico Dandolo ordered his galley beached under the imposing walls of Constantinople. He urged his men up and over, where they entered the history books as the first foreigners ever to enter Constantinople by force. Constantinople fell, and in the upheaval that followed, Venice grabbed a vast swath of colonies spread throughout the Aegean, eastern Mediterranean, and Black Seas. Over the next half century, upward of 70,000 Venetians migrated to these colonies, creating a vast commercial network within a colonial empire.

To run its colonial empire, Venice quickly established a colonial bureaucracy. At its apex stood a relatively small number of chief colonial administrators.24 They occupied extraordinarily lucrative offices: large salaries were paid by the Venetian state (e.g., Robbert 1994, tables 4–8) and officeholders “usually mixed business with politics” (Lane 1973, p. 141). Rösch (1989, pp. 160–161) documents that within just three years of the 1204 conquest, nine Venetians had already earned vast profits. By the time of the Serrata in 1297–1323, chief colonial administrators were often drawn from the richest families of the newly formed nobility (O’Connell 2009, chapter 2). We therefore refer to the benefits of officeholding as nobility rents. There were many other forms of nobility rents, and we focus on this one for simplicity.25

IV. The Puzzle and a Model

We have described a virtuous circle: long-distance trade created a constituency that supported improvements in public and private institutions, and these improved institutions supported the further growth of trade. However, the fourteenth century witnessed a marked decline in economic, political, and social mobility. In the years 1297–1323, the Serrata created a closed hereditary nobility, and in the decade after 1323 this nobility put a stranglehold on the most lucrative lines of long-distance trade.

To understand the events of 1297–1323 and their long-term consequences for Venice’s institutions, we develop a model that highlights how wealth dynamics interact with politics to drive institutional change. In this section, we present the key ingredients and implications of our model. A complete formal presentation can be found in the Appendix.

We build on Banerjee and Newman (1993), in which individuals are motivated by their own material well-being and the bequest they leave for their children. Initial wealth limits the occupational opportunities available to credit-constrained individuals, and this in turn affects wealth dynamics. We tailor the occupation and investment opportunities to our Venetian setting and also add in political economy considerations. In the model, all Venetians can initially participate in international trade. Those without much capital of their own can either remain in Venice as craftsmen or become traveling merchants by signing a colleganza contract with a sedentary merchant, who will put up the required capital. Those with intermediate wealth can finance their own voyage. Finally, the very wealthy can be sedentary merchants in multiple colleganza. International trade is risky, and the success or failure of commercial voyages drives economic and social mobility.

As usual, this type of model is tractable if one focuses on parameter configurations such that children’s occupational choices depend on their parents’ wealth bracket (low or L, medium or M, and high or H) and the success or failure of their parents’ projects, but not on the parents’ specific wealth levels within each bracket. Rather than analyzing every possibility, we focus on a case that captures key elements of the evolution of international trade and political institutions in Venice. Figure IV represents this case on a simplex where we can follow the evolution of Venice’s wealth distribution. The share of low-wealth individuals, denoted PL, is measured along the horizontal axis, the share of high-wealth individuals, denoted PH, is measured along the vertical axis, and the share of middle-wealth individuals is given implicitly by PM = 1 − PLPH. Assume an initial wealth distribution with a mixture of low- and middle-wealth individuals, but very few high-wealth individuals, which corresponds to a point like A on the simplex and characterizes Venice in its early days. In addition, parameters for the returns from international trade and the probability of success and failure are chosen so that, consistent with our earlier discussion, international trade creates substantial mobility, some of it downward but mostly upward. The corresponding differential equations, and the patterns of intergenerational mobility that underlie them, can be found in the Appendix.

Figure IV

Evolution of the Wealth Distribution

Starting from point A, over time commercial success allows some middle-wealth individuals and their children to join the high-wealth group, while failure makes others join the low-wealth group. This makes the wealth distribution move upward and rightward on the simplex toward point B.26 When the size of the group of high-wealth individuals who operate as sedentary merchants increases sufficiently, this reduces their profits. In the model, this is captured in a very simple way. Each sedentary merchant signs colleganza contracts with μ traveling merchants, drawn from the group of low-wealth individuals. When the wealth distribution crosses the PL = μPH line at point B, all low-wealth individuals are now working as traveling merchants. Any further increase in high-wealth individuals (PH) creates competition among them for traveling merchants, which bids up costs and reduces expected profits for sedentary merchants. This increases wealth churning: the poor are now more upwardly mobile and the rich are now more downwardly mobile. As a result, the economy moves toward the steady state at C.

The high degree of intergenerational mobility and churning that characterizes this steady state was preempted, as we shall see, by political developments. To capture these developments, we add in coercive political economy considerations. In line with the historical evidence that membership in the Great Council was initially tied to commercial wealth, in our model the council is made up of individuals who are born to wealthy families as well as those with a more modest background who become wealthy over the course of their lifetime through commercial success. However, Great Council members can vote to prevent further entry into the council by making membership hereditary. This allows existing Great Council members to keep all of the rents associated with political power for themselves, but may trigger a violent revolt by those who are excluded. We model the revolt technology in a simple way so that its outcome depends on the relative size of the groups supporting the revolt and opposing it. A vote to close the Great Council benefits its PH members and harms the PM individuals with middle wealth who could potentially gain entry for themselves and their children through commercial success. Thus, a revolt against such political closure succeeds whenever βPH < PM, where β > 1 captures the fact that Great Council members are more powerful because they control the state’s coercive capacity. Since PM = 1 − PLPH, the revolt condition becomes forumla. If this condition is not met, then a revolt is defeated and its participants are hanged in St. Mark’s Square.

Figure IV illustrates the timing of closure, that is, the timing of the Serrata. As society moves from point A to point B, international trade creates a rising group of very wealthy merchants. As trade continues to feed the joint wealth and power of this group, which is amplified by their control over the coercive power of the state, eventually the group becomes powerful enough that it can close the Great Council without a successful revolt by up-and-coming merchants. This happens when the wealth distribution crosses the line forumla at point S. (S stands for Serrata.) Before this point, if members of the Great Council had voted for the Serrata they would have faced a successful revolt. After point S, members of the Great Council are powerful enough as a group to vote for hereditary membership without facing a revolt. From that point on, a person becomes a member of the Great Council only if his father was a member: Membership, which until then had been associated with commercial wealth, becomes hereditary. A formal nobility is established and equated with membership in the Great Council. This prevents further erosion of political rents through mobility into the Great Council.

After the Great Council becomes hereditary, many nonmembers continue to accumulate high wealth through commercial success. As they do so, these commoners become sedentary merchants and compete economically with nobles (Great Council members), all of whom are sedentary merchants. This competition squeezes the profits of nobles and, in particular, at point B the expected profits of sedentary merchants drop discretely.27 Nobles then have a strong incentive to impose economic restrictions on commoners. A second restrictive measure voted by the Great Council can exclude nonmembers from investing in international trade. Restrictions on trade help nobles by preserving high profits for their commercial activities. The downside is again that this may trigger a violent revolt by those who are negatively affected. Excluding commoners from international trade harms a larger segment of Venetian society than did excluding commoners from the Great Council. Specifically, it reduces the expected earnings of commoners. At the time of the Serrata restrictions on trade would trigger a successful revolt.28 Thus, to restrict investment in trade, nobles need to co-opt some of the nouveau riche commoners who had recently gained high wealth through commercial success but had been excluded from the Great Council. They can do this by increasing membership in the Great Council to forumla so that βPN = 1 − PN, that is, so that a revolt would be defeated. As we shall see, this increased membership is referred to in Venetian history as the enlargement of the Great Council. With this influx of new members, the Great Council is tremendously powerful: membership defines nobility status, and commoners are excluded from the highly lucrative long-distance trade.

In Figure IV, as a result of the restrictions on investing in international trade, the wealth distribution moves rightward from point B to point D instead of moving from B to C. Despite the political and economic closure, while moving toward point D, Venice continues to engage in international trade. However, compared with the evolution toward C that Venice would have followed absent any restrictions, a smaller fraction of Venice’s population is involved in international trade, a larger fraction is involved in manufacturing, Venice’s wealth distribution is more polarized, and social and economic mobility is reduced to a minimum.

To summarize, our model features a Serrata-like event with four key characteristics. First, following a phase of substantial mobility into the Great Council, the council passes measures that implement both political and economic closure. Since restrictions on Great Council membership harm a smaller share of Venetians than do restrictions on participation in the most lucrative aspects of trade, the latter come about later in the process and are preceded by co-optation. This co-optation involves an enlargement of the Great Council that admits more wealthy merchant families. For those who are admitted, their descendants are ensured a seat in the Great Council and a share of the nobility rents, even if those descendants become impoverished. Second, participation in the most lucrative aspects of international trade and in politics become based on family lineage and not on individual merit or commercial success. Third, closure leads to social stratification (decreased social mobility). Fourth, there is a shift in economic activity away from long-distance trade and toward manufacturing.

V. The Oligarchs Triumphant

In this section we review three key events in Venetian history through the lens of our model. First, we provide new evidence from the period 1261–1296 that mobility into and out of the Great Council was eroding the power of many established families. Second, we argue that this erosion is essential for understanding the Serrata of 1297–1323, the most important constitutional event in Venetian history. Norwich (1977, p. 181) describes the Serrata as “The Oligarchs Triumphant.” Third, we show that toward the end of this period and culminating in the early 1330s, a series of laws were passed that severely restricted the ability of non-nobles to engage in long-distance trade. Furthermore, among nobles, it was the most powerful nobles who benefited most from these restrictions.

V.A. The Changing Membership of the Great Council

We start with novel evidence that in the period leading up to the Serrata: (i) there was a high degree of mobility into and out of the Great Council; (ii) a majority of seats in the Great Council were held by a relatively small number of powerful families; and (iii) some of these families were losing seat shares to merchants who had not previously participated in the Great Council. To this end, we constructed a database on representation in the Great Council. A Great Council session lasted for one year, starting in October. The council recorded the names of its members and these lists have survived for each of the sessions in 1261–1262, 1264–1271, 1275–1284, and 1293–1296. The handwritten lists, together with other surviving records of Great Council deliberations, have been transcribed in the Deliberazioni del Maggior Consiglio di Venezia (Cessi 1931–1950).29

The lists are complex. They contain 8,178 legible names. As is well known, Venetian society in general and Great Council elections in particular were organized along family (i.e., clan) lines. See, for example, Raines (2003). It is therefore important to group individuals’ names into families. Most family names have multiple variants, and standardizing these was a lengthy and meticulous process.30

Figure V graphically portrays the extent of mobility into and out of the Great Council and the erosion of seat shares of families who were initially represented in the Great Council. Consider the dashed line. To construct it, we first rank all families based on their initial seat shares, that is, on the average number of seats the family held during the first three available sessions (1261–1262 and 1264–1266). For example, the Dandolo family (1 on the horizontal axis) held the most seats, 4.7% of the total. This 4.7% appears on the vertical axis. The Contarini family (2 on the horizontal axis) held 4.6% of seats, so that the cumulative seat shares held by these two families was 9.3%. This 9.3% is displayed on the vertical axis. Moving rightward along the solid line, 50% of the seats were held by 21 families, 75% of the seats were held by 52 families, and 100% of the seats were held by 162 families. This gives meat and precision to a common observation in the literature that among Venetian families, “between 20 and 50 might be considered great families” (Lane 1973, p. 100).

Figure V

Share of Great Council Seats Held by Families Dating to 1261–1262, 1264–1266

The solid line in Figure V presents the cumulative seat shares at the end of our sample, during the last three available sessions (1293–1296). We retain the ordering of names from 1261–1262 and 1264–1266 so that 1 is still Dandolo, 2 is still Contarini, and so on. Families that did not appear in this initial period are ranked by seat shares in the 1293–1296 period. (This is the concave section at the right end of the solid line.) Three features of Figure V stand out.

First, at the point where the dashed line reaches 100%, the solid line only reaches 87%. Thus, 13% of the end-period seats were held by families that entered the Great Council after the initial period. There were 50 such new families.31 This implies considerable mobility into the council. This was not simply entry of a bunch of small-time players. The seat shares of new families were highly skewed, as can be seen from the concavity of the final portion of the solid line. For example, the new family with the most seats was the Caroso family, who went from no seats to being 28th in the seat-share rank of the end period. Furthermore, most of the new families were engaged in long-distance trade, as evidenced by their appearance in commercial contracts. For example, the new family with the second-most seats was the Caotorta, for whom the surviving records include settlements of accounts with Zaccaria Stagnario for trade between Venice and Constantinople. The new families with the third-most and fourth-most seats, the Nicola and the Barastro, also appear in commercial contracts.32 These four merchant families, with no seats in the initial period, were all in the top 50 by seat shares in the end period. Their rank placed them among Lane’s great families. Thus, new families were quickly growing wealthy and politically powerful from long-distance trade.

The second feature of the figure is mobility out of the council. The flat portions of the solid line are due to families who initially had seats but ended up with none. There are 47 such families among the initial 162. This implies that the exit rate from the Great Council was 1.2% a year. This was nine times higher than the exit rate after the Serrata. For example, in the initial period the Dauro family held 1% of the seats and was ranked 29th, yet the family was no longer in the Great Council by the end period.33

The third and most striking feature of the figure is that the solid line (1293–1296) is well below the dashed line (1261–1262 and 1264–1266). Established families—even some of the most powerful—were losing seat shares. For example, the Falier family, one of the founding families of Venice, who had given the commune two Doges, held 2.5% of the seats and was ranked 6th in the initial period but by the end period its rank had dropped to 17th. Similarly, the powerful Zane family saw their seat rank drop from 9th to 26th.

In summary, this discussion surrounding Figure V shows that there was a high degree of mobility into and out of the Great Council, that new members were engaged in long-distance trade, and that the power even of great families was being eroded by up-and-coming families.

V.B. The Serrata, the “Enlargement of the Great Council,” and State Capacity for Repression

Wealthy families did not take this mobility lying down. Their attack began in the Great Council, where they introduced a series of motions aimed at gaining permanent control. After the failure of four such motions during 1286–1296, a landmark vote on February 28, 1297, effectively handed control of Great Council elections to a small number of powerful families.34 In particular, control over elections passed into the hands of the Council of Forty, a government organ “which had never before claimed a leading role in the state” (Rösch 2000, p. 74), and was controlled by older, powerful families.35 The initial Serrata motion distinguished between those who had served in the Great Council in the previous four years and those who had not. The former group was reelected automatically, provided they were supported by 30% of the Council of Forty (12 votes out of 40). The latter group had to overcome significant obstacles to membership, unless they had sat in the Great Council recently. Measures approved in 1298, 1300, and 1307 substantially strengthened this asymmetry between Great Council insiders and outsiders. Membership in the Great Council had taken a major step toward being locked in. See Hazlitt (1966), Lane (1971, 1973), Todesco (1989), and Rösch (2000).

Political closure was tightened with a series of laws that created a Venetian nobility. In 1310 the concept of nobility was formally introduced for the first time: a nobiles was a man “who was or could be a member of the [Great] Council” (Ruggiero 1980, p. 9). In 1319, the process for electing new members was eliminated. Henceforth, the only route to entry involved proving that a paternal ancestor had sat in the Great Council. The last of the Serrata laws was passed in 1323. It unequivocally made membership in the Great Council a hereditary position. Only men whose fathers and grandfathers had been in the Great Council could hold seats.36

Likely as a reaction to the Serrata, the period 1300–1355 was the most internally violent period in Venetian history from 976 to the demise of the Serene Republic in 1797. In early 1300, a popular commoner named Boccono along with 11 of his associates forced their way into the Great Council chambers. Boccono appears to have been intent on murdering several council members and brow-beating the remainder into reenfranchising those excluded by the Serrata. Arms were not permitted in the chambers, so Boccono and his associates represented a real threat, all the more so because they were backed by a crowd of armed supporters waiting outside in St. Mark’s Square. By a stroke of luck, an overheard conversation revealed the plot: the 12 conspirators were disarmed in the council chambers and executed that night. Their bodies were left hanging in St. Mark’s Square, where they served as a warning. In addition, 40 other supporters of the conspiracy were exiled and had their properties confiscated. See Ruggiero (1980, chapter 1).

Violence boiled over again on the night of June 15, 1310, when Venice was rocked by an armed insurrection. By luck, the plot was revealed the night before by a defector, and even this may not have prevented the insurrection: a violent storm wreaked havoc with communications between the two groups of insurgents who were converging on St. Mark’s Square, and the miscoordinated attack was repulsed. A successful revolt was barely averted. Ruggiero (1980, chapter 1) emphasizes that the motivation for this revolt was opposition to the Serrata.37

A sense of panic began to grip the elite: they might not be so lucky next time. It was time for a new, two-pronged approach that involved the building of coercive capacity and co-optation. The revolt occurred on June 15, 1310. On June 30, the Great Council declared martial law, and on July 10 the first meeting of the infamous Council of Ten was convened. The Council of Ten was initially tasked with tracking down the supporters of the revolt, but it evolved into the Venetian state’s repressive apparatus. From its beginnings, the Ten’s authority within the state hierarchy was left intentionally ambiguous. For example, it was not appointed by the Great Council nor accountable to it.38 Over time, the Council of Ten arrogated to itself whatever powers it needed. For example, in 1319 it created its own police force. Even this was not enough to quell the uproar over the Serrata, for in 1328 the Ten executed the Barozzi brothers for leading a conspiracy against the nobility. By mid-century, the Council of Ten had the necessary resources and experience to repress internal dissent, as evidenced by their speedy handling of Doge Falier’s 1355 attempt to overthrow the Great Council. This brought an end to the stormy period of post-Serrata violence.39

In addition to building up the state’s coercive capacity, the elite co-opted key potential opponents to the Serrata by granting them membership in the Great Council.40 This one-off enlargement of the Great Council is famous in Venetian history and happened quickly, essentially between 1297 and 1310.41 By the time membership of the Great Council became fully hereditary in 1323, its size had more than doubled from 415 members on average in 1261–1296 to around 950 members.42 To get a sense of the scale of the co-optation we have examined all 257 families that were present in the Great Council in 1323. One hundred fifty of these families had seats in the Great Council in 1293–1296 and so were essentially guaranteed hereditary membership. Another 107 families had no seats in 1293–1296 but were co-opted. These co-opted families included 31 who had sat in the Great Council recently, lost all their seats as a result of the intense churning during 1261–1296, and were brought back in. The remaining 76 co-opted families had not been in the Great Council during the period for which records exist (1261–1296). Thus, the Serrata made Great Council membership hereditary, locked in those Great Council members who had seats just before the Serrata (1293–1296), and co-opted many families who, if excluded, could have threatened the stability of the new system.43

V.C. The Closure of Long-Distance Trade

In the decade following 1323, the newly defined nobility passed a series of laws whose consequence was to limit participation by commoners in the most lucrative aspects of long-distance trade. The most important of these was the reorganization of the galley trade, although wealth-based restrictions on who could trade also played a role.

Galleys had long handled the most lucrative traded goods, including cloth, silk, cash, bullion, and spices (Lane 1963, p. 181). Their speed allowed them to escape capture by pirates, their maneuverability allowed them to stay together in convoys, and their small cargo holds made them impractical for anything but valuable lightweight goods. They were also excellent war machines. Toward the end of the Serrata the Venetian state completely overhauled the organization of the galley trade: instead of convoys of primarily privately owned and operated galleys, Venice moved to a system of publicly owned galleys that were auctioned off to private operators. Under the new system, which evolved rapidly between 1321 and 1329, the state chose the destinations and sailing dates of convoys of galleys and then auctioned off the galleys for the duration of the trip (muda). Crucially, “only nobles were allowed to participate in this auction, an exclusive privilege that gave them control of the financial and commercial operations of the fleet” (Doumerc 2003, p. 157). In 1329, this system became a permanent feature of the muda to Greece, Constantinople, and the Black Sea. In 1331, it was extended to the rest of the western Mediterranean and a decade later to Flanders. The new system allowed a handful of powerful noble families to corner what was by far the most lucrative facet of long-distance trade.44

In addition to sewing up the galley trade, the noble-run Commune directly restricted who could trade on the most lucrative routes. In 1324, just one year after the completion of the Serrata, a law was introduced (the Capitulare Navigantium) that forbade any merchant from shipping wares with a value in excess of the merchant’s assessed wealth. Wealth assessments were used by the Commune to determine taxes and, because only the very wealthy paid taxes, the law excluded the poor from long-distance trade. Indeed, it ensured that only the very richest merchants (those with large assessments) could engage in large-scale long-distance trade. The Officium de Navigantibus was created to enforce the new law. It was initially active for less than a year, but was reinstated in 1331–1338 and again in 1361–1363 (Cessi 1952). Although there were a variety of reasons for the Capitulare Navigantium, restricting trade to nobles and wealthy citizens was an important one (e.g., Hocquet 1997, p. 595). This in turn reduced the economic and political mobility long promoted by Venetian trade. Thus, the Capitulare Navigantium “must have galled many ambitious merchants on the make” (Lane 1973, p. 140).45

To examine the impact of the reorganization of the galley trade in the 1320s and the 1324 Capitulare Navigantium, we look at the characteristics of merchants who used the colleganza before and after 1324 to see (i) whether non-nobles were excluded and (ii) whether, among nobles, usage shifted to those with greater political power (as measured by seat shares in the Great Council). We begin by examining colleganza contracts that have survived for the period 1073–1342. In particular, we examine all contracts that appear in Morozzo della Rocca and Lombardo (1940), Lombardo and Morozzo della Rocca (1953), Tiepolo (1970), and Sebellico (1973). These volumes are collections of all types of commercial contracts, such as dowries, wills, lease agreements, loans, and settlements. We first identify which of these commercial documents are colleganza or settlements of a colleganza. In some volumes, each contract is preceded by an editorial header giving the date, place, and type of contract; however, these headers are often vague or inaccurate, so we reviewed each of the 2,833 documents individually. Identification is tricky and requires a considerable time investment to learn how to distinguish colleganza from other related contracts.46 In all we identified 381 colleganza for the period 1073–1342. Some of these have also been coded by Kedar (1976) and González de Lara (2008): Kedar (1976) examines contracts dated 1240–1323 and González de Lara (2008) examines contracts dated 1073–1261. Although neither codes contracts dated after the Capitulare Navigantium, we have been deeply influenced by their work.

For each colleganza we identify the sedentary and traveling merchants and match their family names to the names of families with seats in the Great Council. This involves standardizing family names using the same procedure described earlier. We have data on Great Council membership and seat shares for 1261–1296. From Raines (2003, appendix 1), we also have Great Council membership (but not seat shares) for 1297–1323. We match the merchants’ family names in the colleganza with the 1261–1296 and 1297–1323 Great Council family names and the 1261–1296 seat shares. For the remainder of this section, we refer to merchants with family members in the Great Council in 1261–1323 as “nobles” and to all others as “commoners.”47

Table I presents the results. Column (1) displays the period. The reader will immediately notice one bit of historical irony—no colleganza have survived for 1262–1309, the period that includes Great Council membership records. This is not crucial because our primary interest is in comparing the pre- and post-1324 periods. The gray-shaded rows are the years in which the Officium de Navigantibus was in operation (1324 and 1331–1338). Recall that the Officium was in charge of enforcing the Capitulare Navigantium.

Table I

Commoner Involvement in Colleganza

graphic
graphic
Table I

Commoner Involvement in Colleganza

graphic
graphic

Column (2) reports the number of colleganza that have survived for each period. Column (3) reports the number of colleganza in which at least one of the merchants was a commoner, that is, a merchant with no family in the Great Council from 1261 onward. Column (4) reports these colleganza as a share of all colleganza in the period. Comparing 1310–1323 with all later periods, there is a sharp drop in commoner participation after the Capitulare Navigantium. During 1310–1323, commoners participated in 27% of all colleganza. After 1324 there is only a single colleganza with commoner participation.48

By 1310, Venice was already deep into the Serrata, so we might expect that an informal process of commoner exclusion may already have been under way. That is, a comparison of the 27% figure for 1310–1323 with the essentially 0% figure for 1324–1342 may understate the full extent of commoner exclusion. It is therefore useful to look further back, to 1241–1261 and even further. Indeed, commoner participation was higher in earlier years, making the 1324 break starker. Commoners were involved in 51% of all colleganza during 1241–1261.49 As far back as 1073–1200, commoners were involved in 42% of all colleganza.50

Column (5) moves the discussion away from commoner participation and to the power of the nobles who participated. In each period we draw up a list of all the merchants involved in a colleganza. We then assign each of these merchants a power score, which is simply his family’s number of seats in the Great Council. (The number of seats is the family’s average number of seats per session during 1261–1296.) We then examine how power scores of the median merchant evolved across periods. Similar results hold for averages. The first observation from column (5) is that the median number of seats is positive, that is, the median merchant had family members in the Great Council. Prior to 1310, the median merchant’s family presence in the Great Council was modest. For example, in 1241–1261, the median family had less than one seat, which signals that this median family alternated in and out of the council. During the Serrata but after the Enlargement (1310–1323), the median merchant’s family held almost three seats. After the Serrata (1325–1338), it jumped even higher, to about five seats. Thus, after the Serrata and, especially after the Capitulare Navigantium, use of the colleganza shifted to more and more powerful families.51

Note that after 1330, there is a very significant drop in the number of extant colleganza. This does not appear to be the result of changes in notarial contracts: we do not see similar trends for other types of contracts. The most convincing explanation has to do with the reorganization of the galley trade. As we discuss next, this led to a shift in financing away from the colleganza and toward financing through family and marriage alliances.

VI. Economic Inequality, Social Stratification, and Resource Reallocation

The political and economic Serrata had very significant long-run implications for economic inequality, social stratification, and resource reallocation. We turn to the three of these in the following subsections.

VI.A. Economic Inequality

The capital requirements of the galley trade were huge. A winning bid in the galley auction averaged 793 ducats during 1332–1345. This was only a minor item in the total cost of chartering the galley, which reached 9,200 ducats (equivalent to 33 kg of fine gold) by the late 1400s. In addition to the charter costs, it was necessary to cover salaries and provisions for a crew in excess of 150 men for a period of 5 to 11 months. All of these costs were in turn dwarfed by the cost of the freight, often valued at over 150,000 ducats in the early 1400s.52 In an earlier, pre-Serrata age, these huge up-front fixed costs would have been shared by many merchants, both noble and commoner, and financed with a large number of colleganza. In the post-Serrata age, we initially see broad-based noble participation in the galley trade. However, over the next 150 years an ever-narrower group of nobles came to monopolize the galley trade.

To understand how this happened, one must understand the details of how the state-run galley trade was financed. Each galley in its entirety was auctioned off to a single noble bidder (the patrono), who in turn divided the galley into 24 shares and up to 24 shareholders.53 In the years immediately following the Serrata there was widespread noble participation in the galley trade. This was necessary because not even the richest families could afford the high up-front capital costs of a successful bid.

The mid-1300s were a difficult time for trade, with the plague of 1348 and ongoing wars with Genoa until 1380. After 1380, however, Venice began to recover and with this recovery a slow process of concentration in the galley trade began. At the start of the recovery, participation was still widespread: two thirds of noble families participated in the galley trade and more than a a third provided patroni (Doumerc and Stöckly 1995, p. 143). By the mid-1400s, evidence of increasing concentration in the galley trade was inescapable. In Doumerc and Stöckly’s (1995, pp. 140–142) analysis of 121 galleys during 1445–1452, the patrono held the majority of shares, either alone or with his brothers and sons, in 60% of galleys. On average, the patrono’s family held 56% of the shares. More and more often the patrono was from a particularly prominent family.

To become a majority shareholder in a galley, a noble typically had to give up the advantages of risk diversification and concentrate his investment in that galley. During 1445–1452, 85% of shareholders were invested in a single galley (Doumerc and Stöckly 1995, p. 146). By 1500, it was common for a single family to hold all of the shares in a galley. One even begins to see instances where all the galleys in a muda (convoy) are controlled by a family or small group of families. For instance, the brothers Alvise, Andrea, and Pietro Marcello held all 24 shares in a galley of the muda to Trafego in 1496. They raised their stake with 12, and then 18, and then 23 shares in a second galley in 1497, 1498, and 1499, and finally held all 48 shares in two of the three galleys of the muda to Trafego in 1500 (Doumerc and Stöckly 1995, p. 147; Judde de Larivière 2008, p. 181).

The greatest advantage of cartelizing a muda came from price fixing. Michiel da Lezze, son-in-law of Pietro Marcello, left detailed evidence of this practice in his business correspondence (Braudel and Tenenti 1966, p. 62). In 1506 he instructed his son Luca, patrono of a galley of the muda to the Barbarie Coast, to collude with the other patroni as monopsony buyers to drive down the price of wool in Valencia. Upon returning home, they colluded again as monopolists to drive up the sale price in Venice. These and other anticompetitive practices begin to appear frequently in court cases from 1450 on. “The abuses are more and more frequent as financial concentration increases” (Doumerc and Stöckly 1995, p. 147, our translation).

Controlling an entire muda required vast financial resources, and during the course of the 1400s a uniquely un-Venetian solution emerged. Family members, typically brothers, raised capital within a family. Because even this was rarely enough to control one or several galleys, marriage alliances were established with other powerful families and additional capital was raised within the alliance. This in part explains the decline of the colleganza documented above. It is a step backward from impersonal relationships to kin-based relationships as the basis for Venice’s long-distance trade. It is also a step backward from the era of pre-Serrata economic mobility. In its place a period of spectacular inequality at the top end of the income distribution was ushered in.

The use of marriage alliances had a profound effect on financing and hence on concentration in the galley trade. The 46 galleys sent in the muda to the Levant between 1519 and 1528 had an average of just two shareholders, despite a cargo value worth between 150,000 and 200,000 ducats per galley (Doumerc and Stöckly 1995, p. 152). Furthermore, the lists of shareholders after 1500 are dominated by the Contarini, Garzoni, Marcello, Loredan, Pisani, Priuli, Michiel, Morosini, and a very small handful of other rich families. During the period 1495–1529, 30 individuals from just 17 noble families owned 38% of all shares in the galleys of the different muda (Judde de Larivière 2008, table 8, p. 140). Over the same period, the families of the shareholders were linked by marriage in almost every single galley (Judde de Larivière 2008, p. 144).

The problems of monopolization—both the anticompetitive costs and the implications for extreme inequality—were decried by contemporary chroniclers such as Sanudo and members of the Great Council. Ideally, we would like to track the extreme inequality associated with the rise of this ultra-rich elite, especially after the post-1380 Venetian expansion. One certainly sees it visually in the ornate palazzos that began lining the Grand Canal in this period (Goy 1992, p. 10). Unfortunately, there are no systematic data that would allow us to track economic inequality.54 Nevertheless, we have been able to exploit a source that has not previously been systematically examined: records of Venetian noble marriages.55 Since these marriages were intimately connected with cartelization, as we have seen, this will give us a systematic portrait of post-Serrata Venetian economic polarization.

After the Serrata, given the patrilineal hereditary nature of Great Council membership, it became increasingly important for the state to keep a record of marriages of sons of male nobles. The process went hand in hand with the closing of any gray areas surrounding eligibility for Great Council membership, for example, sons born out of wedlock. It is a process that was not completed until the early 1400s (see Chojnacki 1994, 2000, and the discussion later in this section). We are therefore able to examine records of marriages starting in 1400. A handwritten list kept at the Archivio di Stato di Venezia records Venetian marriages involving a noble husband. This was compiled in the late nineteenth century by archivist Giuseppe Giomo from multiple sources, including the records of Avogaria di comun and an earlier compilation by Marco Barbaro. We have digitized this list and use all 6,959 marriages among noble families for the period 1400–1599 in our analysis.56 We have performed the same standardization of family names as used for the Great Council seat data of the previous section and merged both data sets. This allows us to track the evolution of power within Venice’s nobility.

Figure VI shows the network of marriages among noble families in 1400–1499. Each circle (node of the network) represents a family, and each arc is a marriage connecting two families. All marriages for this period are plotted, with thicker and darker arcs representing more marriages between two families. The heaviest line is between Contarini and Morosini (52 marriages). There are also thick lines connecting Michiel to Contarini, Corner to both Contarini and Morosini, and so on. Roughly speaking, families that appear closer to each other in the graph are more strongly connected. So, for instance, Michiel is more closely related to Contarini (17 marriages) than to Morosini (6 marriages).57 The size of the circles increases with the family’s importance in the network as measured by eigenvector centrality (Bonacich 1972). This assigns relative scores to all nodes in the network based on the idea that a node is more important when it is better connected to other important nodes.58

Figure VI

Marriage Network among Noble Families in 1400s

Given that marriage alliances were used to mobilize the vast financial resources required to control entire galleys or even muda, eigenvector centrality gives a measure of families’ ability to monopolize routes in the galley trade and how this evolved over time. Consider Figure VII. Each point is a noble family’s eigenvector centrality in the marriage network of 1400–1499 plotted against the family’s percentile in the Great Council seat share distribution of 1261–1296. Families with extremely high centrality in the 1400s had extremely high pre-Serrata seat shares. This strong persistence is in contrast with the remarkable mobility we documented for the pre-Serrata period. More specifically, in Figure VII only families in the top quartile of the pre-Serrata seat-share distribution appear in the top decile of marriage centrality in the 1400s.

Figure VII

Centrality in the Marriage Network in 1400s against Great Council Seat Share in 1261–1296

To externally validate the relevance of our centrality measure, we compare high-centrality families with those who are prominent in the galley trade. In the galley trade data from 1495–1529 there are 12 families with at least 10 participants and at least 100 galley shares. As expected, these 12 families are all in the the top decile of marriage centrality in the 1400s. Together, they accounted for half of all of the galley trade in terms of shares invested. This externally validates our eigenvector centrality measure. It also shows that the most powerful families in the Great Council in the period leading up to the Serrata strategically used marriage alliances to monopolize the galley trade, diverting the public resources devoted to the muda to their own private benefit.59

From Figure VII, high pre-Serrata seat shares did not guarantee high post-Serrata marriage centrality (e.g., Tron and Venier). These two families did not manage to intermarry as well as others and were increasingly excluded from the galley trade. Contemporary chronicles and debates in the Senate, the body that regulated international trade, reflect growing tensions in the 1500s about the capture of the galley trade by a handful of families. Summarizing these tensions, Doumerc and Stöckly (1995, p. 156, our translation) describe how the families “exploiting the galley trade defend inch by inch the advantages gained, often through abuses against the law, to preserve their privileges. On the other side, fiercely opposing the ongoing activities of the muda, are the families whose names no longer appear as shareholders. First among them one should mention Tron and Venier.” In short, a high pre-Serrata seat share was necessary but not sufficient for success in the post-Serrata era.

Further to the left of Figure VII, it is interesting to look at Priuli and Pesaro. Priuli’s seat share only just placed it in the top quartile of the pre-Serrata seat-share distribution. Thus, the family would have required very strategic marriage alliances to penetrate the cartelized galley trade. This sometimes meant marrying with top families and, during the 1400s, the Priuli established marriage links on three occasions with each of the two most central families, Contarini and Morosini. However, sometimes it meant entering into exceedingly complex alliances with other families of intermediate centrality so that together they could control a muda. This is illustrated by the 1504 muda to Flanders where all three galleys were jointly controlled by the Priuli and Pesaro families. The families were linked in 1499 by the marriage of Alvise Priuli’s daughter to one of Nicolo Pesaro’s sons, and even though the son died shortly afterward, the marriage link was quickly reestablished in 1502 by having the Priuli widow marry her late husband’s brother Pietro Pesaro. Now consider the three galleys of 1504. In the first galley, Alvise Priuli held 8 of the 24 shares jointly with his son-in-law Pietro Pesaro and another 4 jointly with his brothers Bernardo and Giovanni Priuli. Pietro Pesaro held another four shares and his father, Nicolo Pesaro, held the remaining eight shares (8 + 4 + 4 + 8 = 24). In the second galley, Alvise Priuli held ten shares, Pietro Pesaro held six shares, and the family of Alvise’s wife held the remaining eight shares (10 + 6 + 8 = 24). In the third and final galley, the Priuli family (including Alvise and his brothers) held 18 of the 24 shares. This Priuli–Pesaro example illustrates how complex it was for all but the richest families to break into the galley trade. The Priuli family was far from poor—it ranked 18th among noble families by wealth in the 1379 assessment—yet it still required complex marriage alliances to succeed. This example also illustrates the huge payoffs from monopolizing a muda. By the 1500s, Priuli had already become the fourth most central family in the noble marriage network and two Priuli brothers held the Doge position consecutively after 1556.60

In all our figures, the Contarini are the most prominent family. Interestingly, five Contarini held the Dogeship between 1623 and 1684, an uncomfortable regression toward an earlier era were the Dogeship was passed on from father to son.

At the very left of Figure VII we see that 24% of noble families had no seats in the Great Council in 1261–1296. These are the Enlargement families, co-opted into the Great Council during the Serrata. None of them became important in the noble marriage network, as shown by their low eigenvector centrality in the 1400s. They also had little participation in the galley trade, and less so as time went by. As we shall see, being in the nobility was no guarantee against poverty.61

We have demonstrated that the distribution of power was remarkably persistent from 1261–1296 to 1400–1499. We next show that this persistence was even more pronounced between the 1400–1499 and 1500–1599 periods. Figure VIII tracks the evolution of families’ eigenvector centrality in the marriage network between the 1400s and the 1500s. The intertemporal correlation is very high, which shows that the same families dominated over these two different centuries. This extreme persistence over such an extended period is in stark contrast to the high mobility and permeability that characterized Venice before the Serrata.

Figure VIII

Persistence in Centrality in the Marriage Network

To summarize, restrictions to commoner participation in long-distance trade and the reorganization of the galley trade dramatically reduced the economic competition faced by the most powerful families. The result was a tremendous strengthening of the economic and social position of those families who held the most seats in the Great Council in 1261–1296 and who stood to gain the most from the Serrata.

VI.B. Social Stratification

The Serrata transformed Venetian politics and economics. This in turn led to a fundamental shift in society away from one characterized by political, economic, and social mobility and toward one of political immobility, economic polarization, and social stratification. In the words of Romano (1996, p. xv), the “values of the early fourteenth century gradually gave way in the last quarter of the fourteenth century and the early fifteenth century to a new emphasis on rank and hierarchy.” Furthermore, these changes “signalled a major transformation of the social foundations of the Venetian Renaissance state” (Romano 1987, p. 38). This transformation has been famously emphasized and documented by Chojnacki (1973, 1985, 1997). We briefly relate it to the Serrata and the consequent reorganization of the galley trade.

The starting point of the great social transformation was the implicit bargain contained in the Enlargement of the Great Council. The Enlargement guaranteed that patrilineal descendants of co-opted members would also be nobles, but did not guarantee that these descendants would be successful in long-distance trade. During the difficult economic years from the plague (1348) to the Battle of Chiogga (1379–1380), a very large number of nobles became impoverished: “many, probably most, nobles were poor” (Queller 1986, p. ix).62 This could not be ignored by rich nobles: poor nobles could potentially organize a revolt, which was precisely what their co-optation during the Enlargement was intended to prevent. To prevent revolt, the old system of patronage (grazie) that had been dominated by rich families was reshaped during the second half of the 1300s to meet the demands of poor nobles. (Queller 1986, chapter 2) famously describes the new patronage system as “welfare jobs for poor nobles.”63

Immediately after the Serrata the most important government jobs were reserved for the nobility. Because these jobs paid well and because there were increasingly many more poor nobles than government jobs, a great deal of attention and rent-seeking behavior was focused on job allocation. Traditionally, the allocation was done through the old patronage system: rich noble patrons handed out jobs to their clients rather than to needy nobles. Poorer nobles objected strenuously to this favoritism. They were particularly frustrated by the fact that not enough of these jobs went to poor nobles, that the scarce jobs were not spread fairly among these poor nobles (these were typically annual jobs that could easily be rotated), and that too many jobs went to non-noble relations of the richest nobles. Such non-noble relations included underaged children of nobles, illegitimate children of nobles, and non-noble branches of noble families.

Poor nobles fought back using their voting power in the Great Council. In 1367, legislation was passed requiring Great Council approval of all candidates for government jobs. This legislation was strongly opposed by the richest families and only passed on the fifth vote (Chojnacki 1997, p. 677). Then, during 1380–1420, a series of new laws began formalizing precisely which jobs were reserved for poor nobles.64

Of course, none of this mattered unless there was a clear definition of nobility. We thus see in this same period a series of initiatives aimed at eliminating all gray areas in the definition of nobility. In 1367, a law was passed that excluded illegitimate sons from retaining noble rights. In 1414, a requirement was passed to create written lists of all children who were eligible to become nobles, thus excluding the underaged from government jobs and creating proper recordkeeping of noble status. Finally, since marriage alliances had become so central to the economic workings of the nobility, in 1422 the status of marriageable woman was also clarified: sons born to mothers of “low or vile condition” could not inherit their father’s noble status. As a result, it became necessary in the patrilineal system to document the parentage and premarital behavior of mothers to claim noble status (Chojnacki 1994).

These developments led directly to the social stratification that was the hallmark of Venetian society after 1400. The consolidation of social stratification during the period 1380–1420 has been richly documented by Chojnacki (1973, 1985, 1994, 1997, 2000) and Romano (1987) and is widely recognized by social historians as a key event in Venice’s social history.65

While the richest nobles would have preferred a fluid definition of nobility, poor nobles wanted clarity so that the richest nobles could not abuse the new state-controlled patronage system. The process of social stratification must thus be understood in the context of the Enlargement of the Great Council, which laid the groundwork for the appearance of so many poor nobles. In short, the Serrata reduced social mobility and replaced it with a stratified system emphasizing rank and hierarchy.

VI.C. The Relative Shift Away from Maritime Activities

A prediction of our model is that oligarchization and the resulting entry barriers to maritime trade cause a shift out of maritime occupations and into craft-based (industrial) occupations.66 More generally, entry barriers lead to a shift into nonmaritime activities. In the early 1400s Venice expanded dramatically into the Italian mainland and, as well, began a steady expansion of industry (woollen cloth, chemicals, glass, leather work, and printing). By the mid-1500s, industrial activities had overtaken maritime activities as the primary driver of the Venetian economy. For example, Venice was a leading center in the production of woollens, which in turn was one of the biggest sectors within European manufacturing (Lane 1973, pp. 309–312; Braudel 1984, pp. 135–136). Stöckly (1995, p. 345) argues that the monopolization of the galley trade directly explains why nobles increasingly turned away from the sea. In short, the Serrata marked the beginning of the end of Venice’s maritime economy. “Never again was Venice so largely a maritime nation as it had been in the thirteenth century” (Lane 1973, p. 170).

VII. Conclusion

Beginning in 800, Venice was launched on a path of political independence for more than 1,000 years. This allowed us to examine the Venetian response to Europe’s great wave of medieval globalization, the Commercial Revolution. Venice’s unique geographic and cultural location between Byzantium and Western Europe benefited its merchants, especially after the opening of Mediterranean shipping lanes to Christian shipping in 976 and the preferential trade arrangements with Byzantium after 1082. The Venetian (Fourth Crusade) conquest of Constantinople in 1204 created a colonial system in the eastern Mediterranean that massively expanded Venetian trade. Between 976 and 1297, rapidly rising long-distance trade empowered Venetian merchants, who used their clout to push for novel institutional arrangements. In 1032 they reined in the Dogeship (it stopped being hereditary), and in 1172 they created a parliament (the Great Council). That is, they successfully pushed for improved constraints on the executive. Furthermore, they showed remarkable dynamism in developing new contracting institutions. We examined one of these in detail, the colleganza, and showed how it promoted income mobility and, with it, political mobility.

Over time, a group of powerful merchants emerged and, starting in 1297, they used their resources to block political and economic competition. In particular, they made parliamentary participation hereditary and erected barriers to participation in the most lucrative segment of long-distance trade (the galley trade). We documented this rise and fall of political and economic competition using a unique database on the names of 8,178 parliamentarians and their families’ use of the colleganza. We also linked this database to data on the galley trade and 6,959 marriages to track the rise of an ultra-rich and socially stratified oligarchy. These data demonstrate Venice’s fundamental shift after 1297 from a society characterized by political, economic, and social mobility and toward one of political immobility, economic polarization, and social stratification.

Appendix: A Formal Dynamic Model of Wealth and Politics in Venice

This appendix provides a formal presentation of the model on which Section IV is based. Venice has a continuum population of constant measure 1. Each person inherits wealth as a bequest from his parent. He is also endowed with one unit of labor from which he earns additional income. Time is continuous, individuals reach the age of maturity at a rate λ, and when maturity is reached the following sequence of events occurs instantaneously.

  • Mature citizens with inherited wealth above wN become members of the Great Council. In the council they vote on two issues (votes are decided by simple majority):

    • whether membership in the Great Council is hereditary, that is, passed on from father to son.

    • whether commoners can invest in international trade. (“Commoners” are citizens with inherited wealth less than wN.)

  • Mature citizens who are unhappy with the Great Council’s decisions choose whether to revolt. Great Council members and their supporters choose whether to fight the revolt.

  • Mature citizens choose one of four occupations (detailed later), which determines how they invest their labor and capital. They then realize investment outcomes.

  • If Great Council membership is hereditary (the vote in i.a) and if a mature citizen’s wealth exceeds wN after investment outcomes are realized, then the citizen enters the Great Council.

  • Members of the Great Council enjoy a nonpecuniary benefit that accrues to Great Council members. This benefit corresponds to the “nobility rents.”

  • Each mature citizen consumes, dies, and leaves a bequest to his only child.

There is a single physical good in the economy that may be consumed or used as capital. All citizens have identical preferences:
(1)
where c is the individual’s consumption, b is the bequest to his offspring, l is the labor effort he exerts, and g is the rents from Great Council membership. For nonmembers, g = 0. For members, g is positive and decreasing in the number of Council members.

Turning to d in equation (1), citizens whose interests are harmed by Great Council votes may choose to revolt. d in the utility function captures the ex post cost of a revolt: d = 0 if there is no revolt or if there is a revolt and the individual is on the winning side. d > 0 if there is a revolt and the individual is on the losing side. We assume that d is so large that no one participates in a revolt they cannot win (the individual is hanged in St. Mark’s Square). Details of the revolt technology appear below.

Turning to occupations, a Venetian’s capital and labor may be employed in one of four occupations:

  • Craftsman: he uses his labor effort to produce v units of output, where v > 1 is fixed by technology.

  • Traveling merchant: he signs a colleganza contract with a sedentary merchant. The sedentary merchant puts up the capital; the traveling merchant contribute his labor (effort) but needs to be monitored. The voyage yields a high rate of return ρ1 with probability σ and a low rate of return ρ0 > 0 with probability 1 – σ. The expected rate of return is forumla The traveling merchant receives an endogenous share of profits.

  • Self-financed merchant: he makes a fixed indivisible investment I and uses his labour effort on a ship to foreign ports. The voyage yields a high rate of return forumla with probability σ and a low rate of return forumla with probability 1 – σ. The expected rate of return is forumla.

  • Sedentary merchant: he makes a fixed indivisible investment μI and uses his labor effort to monitor μ traveling merchants. Monitoring is an indivisible activity and one cannot monitor another monitor, so that as in Banerjee and Newman (1993), μ > 1 is an exogenously given constant. Returns are assumed to be perfectly correlated across a sedentary merchant’s μ colleganza, and he receives an endogenous share of profits.

There is also a safe divisible asset that yields a fixed rate of return r, where forumla and γ is defined in equation (1).67 We also assume forumla and forumla so that investing in colleganza is always preferred to investing in a safe asset.

Following Banerjee and Newman (1993), we assume that due to capital market imperfections, people can borrow only limited amounts. As a result, occupations that require high levels of investment are beyond the reach of individuals with lower wealth levels.68

Let wM = I (M for middle wealth) denote the minimum wealth level required to become a self-financed merchant and wH = μI (H for high wealth) denote the minimum wealth level required to become a sedentary merchant. To capture the historical evidence, we assume that being a sedentary merchant opens the doors to participation in the Great Council. That is, wH = wN. (See point iv above.) Let PL be the share of the population with wealth w < wM, let PM be the share with wMw < wH, and let PH be the share with wwH. Because wealth w evolves endogenously, the Ps evolve endogenously.

As in Banerjee and Newman (1993), this type of model is tractable if one focuses on parameter configurations such that children’s occupational choices depend on their parents’ wealth bracket (low or L, medium or M, and high or H) and the success or failure of their parents’ projects, but not on the parents’ specific wealth levels within each bracket. Rather than analyzing every possibility, we focus on a case that captures key elements of the evolution of international trade and political institutions in Venice. A second case is also discussed shortly.

Consider the market for traveling merchants. Suppose there are very few high-wealth people (all of them are sedentary merchants). Then very few traveling merchants are needed and the returns to being a traveling merchant are low—so low that they are no better off than craftsmen. This implies that the traveling merchant’s profit share is:69
(2)
If there are many high-wealth people (all of them are sedentary merchants) then they will demand so many traveling merchants that there will be no more craftsmen. At this point the returns to being a traveling merchant rise, which squeezes the profits of sedentary merchants. In equilibrium, high-wealth people (w > wH) are so squeezed that they are indifferent between being sedentary merchants and self-financed merchants. This implies the following profit share for traveling merchants:70
(3)
Recalling that there are PH sedentary merchants, each hiring μ traveling merchants, equation (2) applies when there are relatively few high-wealth people (μPHPL) and equation (3) applies when there are relatively many high-wealth people (μPH > PL).71
Few Rich Merchants (μPHPL)

Panel A in Figure A.1 presents wealth dynamics when the demand for travelling merchants is low (μPH < PL). It plots the size of an individual’s bequest b against his initial (inherited) wealth level w. The size of a bequest depends on the individual’s choice of occupation and the success or failure of his projects. In every case, the bequest is simply a share γ of the wealth of the individual at the end of his life. The choice of occupation in our credit-constrained economy depends on initial wealth: sedentary merchants require wealth forumla; self-financed merchants require wealth forumla; while travelling merchants and craftsmen do not require any wealth.

Figure A.1

Occupations and Bequests by Individual Wealth

Consider first the three upward-sloping lines to the left of w = wM. A low-wealth individual who inherits w < wM can choose between being a craftsman and being a traveling merchant. If working as a craftsman, his labor income is v (middle line). If working as a traveling merchant, his labor income is forumla with probability σ (top line) and forumla with probability 1 – σ (bottom line). Whether a craftsman or a traveling merchant, he also earns nonlabor income rw from placing his inherited wealth in the safe asset. Note that γ(r + 1) is the slope of all three lines, which reflects the fact that those with higher initial wealth w end up with higher nonlabor income rw. In Panel A, forumla adjusts to keep an individual indifferent between being a craftsman and a traveling merchant. As drawn in the region w < wM, initial wealth w is not enough for the child to become a self-financed merchant (i.e., b < wM). This will be what matters for the wealth dynamics of the poor (we discuss later an alternative scenario with upward mobility for low-wealth individuals).

Staying with Panel A, consider an individual in the middle-wealth region wMw < wH. He is wealthy enough to be a self-financed merchant (wMw), but not a sedentary merchant (w < wH). Furthermore, expected returns are higher for self-financed merchants than for craftsmen or traveling merchants. Hence he chooses to be a self-financed merchant. He thus makes a fixed indivisible investment I and the voyage yields profit forumla if successful (probability σ) and profit forumla if unsuccessful. He also receives a return r(wI) from placing his remaining wealth in the safe asset. The involvement in international trade of middle-wealth individuals creates mobility in the wealth distribution: a successful self-financed merchant bequests b < wG and so his offspring begins life as a high-wealth individual, whereas an unsuccessful self-financed merchant bequests b < wM and his offspring begins life as a low-wealth individual.

Looking finally at high-wealth individuals, someone with inherited wealth wwH always becomes a sedentary merchant (it gives higher expected returns than any other occupation). He makes a fixed indivisible investment μI in colleganza. His profit net of payments to traveling merchants is forumla if successful and forumla if unsuccessful. He also receives a return r(w – μI) from placing his remaining wealth in the safe asset. When there are relatively few high-wealth individuals and relatively many low-wealth individuals, the returns for a sedentary merchant are always high enough that he bequests b > wH and his child also begins life with high wealth.

Stepping back from the details of Panel A, all wealth mobility comes from middle-wealth individuals. Depending on their success as self-financed merchants, their children are either upwardly mobile (with probability σ) or downwardly mobile (with probability 1 – σ). The children of all others (low- and high-wealth) remain in the same wealth groups as their parents. This immediately implies the following dynamics:
(4)
where a dot indicates a time derivative. (Recall that λ is the share of the population that is active.)
Many Rich Merchants (μPH > PL)
Panel B presents the case where the relative number of high-wealth individuals increases sufficiently (μPH > PL) that the returns to traveling merchants are given by equation (3) instead of equation (2). Low-wealth individuals now all become traveling merchants and, if successful, bequest to their children b > wM, so that these begin mature life as middle-wealth individuals. For middle-wealth individuals, nothing changes. For high-wealth individuals, there is now indifference between being self-financed or sedentary. The children of successful high-wealth individuals inherit high wealth. The children of unsuccessful high-wealth individuals inherit middle wealth if their parents were sedentary merchants and low wealth if their parents were self-financed merchants. This implies the following dynamics:72
(5)
Wealth Dynamics
Because total population is fixed, we can express the dynamics of equations (4) and (5) in terms of PL and PH only, with PM = 1 – PLPH:
(6)
(7)

These dynamics are represented on a simplex in Figure IV in the main text.

Wealth Dynamics with Upward Mobility by Low-Wealth Individuals

In the parameter configuration discussed so far, chosen for its expositional simplicity, all wealth mobility comes from middle-wealth individuals. The experience of Zaccaria Stagnario (discussed in the text) is an example of upward mobility starting from low wealth instead of middle wealth. It is straightforward to incorporate this into our model with a single parameter change: we simply increase v, the output of craftsmen, which improves outside opportunities for traveling merchants and allows them to obtain a greater profit share forumla when μPHPL.

Relative to the individual wealth dynamics represented in Figure A.1, this implies a single change. In Panel A, which presents wealth dynamics when the demand for traveling merchants is low (μPH < PL), the wealth attained by a low-wealth individual who succeeds as a traveling merchant (forumla) is now enough for his child to become a self-financed merchant, that is, the top line in the low-wealth range now lies above the wM horizontal threshold instead of below it. This implies that the children of the μPH low-wealth individuals working as traveling merchants begin mature life as middle-wealth individuals with probability σ. Thus, compared with the population wealth dynamics when μPHPL given by equation (4), the element in the first row and first column of the matrix is now –σμPH instead of 0, and the element in the second row and first column of the matrix is now σμPH instead of 0. The population wealth dynamics when μPH > PL are still given by equation (5). Expressing the population wealth dynamics in terms of PL and PH only, with PM = 1 − PLPH yields:
(8)
with forumla still given by equation (7). Figure A.2 plots the evolution of the wealth distribution according to these alternative wealth dynamics. This features a Serrata-like event with the same key characteristics as those discussed in the main text. The only difference is that in the transition from point A to B in Figure A.2 there is upward mobility out of the low-wealth group, whereas in Figure IV all mobility comes from the middle-wealth group.
Figure A.2

Alternative Evolution of the Wealth Distribution

*Dedicated to John Munro: inspiring teacher, inspired colleague. We are thankful to Lisa Chen, Jennifer Konieczny, Isabela Manelici, and Luca Repetto for excellent research assistance. We are grateful to our colleagues in the Institutions, Organizations and Growth Program at the Canadian Institute for Advanced Research (CIFAR), including Daron Acemoglu, Mauricio Drelichman, Elhanan Helpman, Joel Mokyr, Jim Robinson, and especially Avner Greif. We have benefited from conversations with Abhijit Banerjee, John Munro, Dorit Raines, Nathan Sussman, Aloysius Siow, Bob Staiger, Jeff Williamson, and patient participants at many workshops. Yadira González de Lara provided us with detailed, repeated, and essential feedback. Comments by the editor (Elhanan Helpman) and the anonymous referees greatly improved the article. Trefler thanks CIFAR and the Social Sciences and Humanities Research Council of Canada (SSHRC) for financial support. Puga thanks the European Commission’s Seventh Research Framework Programme for the European Research Council’s Advanced Grant “Spatial Spikes” (contract number 269868) and for the Collaborative Project HI-POD (contract number 225343), the Banco de España Excellence Programme, the Comunidad de Madrid (grant S2007/HUM/0448 PROCIUDAD-CM) and the IMDEA Ciencias Sociales and Madrimasd Foundations for financial support. The data and replication files for this article are available from http://diegopuga.org/data/venice/.

1. This success was not a given. As in Greif (2005, 2006b, chapter 6), Acemoglu (2008), and Acemoglu and Robinson (2012), we argue that constraints on the executive are not enough to guarantee long-term success.

2. This point is part of Greif’s notion of coercion-constraining institutions (Greif 2005, 2008; González de Lara, Greif, and Jha 2008). See also North’s (1990) related critique of North and Thomas (1973). This point is also part of Acemoglu’s (2003) discussion of why there is no political Coase theorem.

3. See Acemoglu (2008), Acemoglu et al. (2008), and especially Acemoglu and Robinson (2006, chapter 10), who examine how changes in economic and political inequality affect economic development and transitions to democracy. See also Greif and Laitin (2004) and Greif (2006b), who study the role of changing income distribution for self-enforcing cooperation and institutional change. Do and Levchenko (2009) and Levchenko (2013) develop theoretical models of the impact of trade, via rent creation and the power it confers, on institutions.

4. The importance of the interaction between politics and economics in credit-constrained economies has been investigated by de Soto (1989) and Besley, Burchardi, and Ghatak (2012).

5. This aspect of our article is related to Jha’s (2010) analysis of the English Long Parliament (1640–1660). He uses detailed data on parliamentarians to examine the formation of the coalition that supported stronger constraints on the Crown. Using data on the investments of about 500 parliamentarians, Jha finds that a particular financial innovation—shares in overseas companies—allowed broader investor participation in overseas trade and thus was key in aligning interests against royal discretion over foreign economic affairs. This is similar to the longer term role the colleganza played in Venice before (but not after) the Serrata.

6. Greif’s point is best illustrated by his cross-cultural analysis of medieval trade in Western Europe versus the Islamic world (including the Maghreb in North Africa). At the start of the medieval period both regions organized trade in kin-based (and community-based) networks. However, as trade expanded, its volume reached a level that was not supportable by kin-based organizations. The Venetian response was to develop business organizations that allowed unrelated individuals to work together and the Venetian parliament passed civil statutes that legalized these business organizations (e.g., the colleganza). Finally, Venice could claim legitimacy for its parliament and civil statutes only because of the political vacuum left by weak European states. In contrast, such institutional responses were not possible in the Islamic world. See Greif (1992, 1993, 1994, 2005, 2006a).

7. This second conjecture is related to the issue of Venetian stability that has preoccupied students of Venetian history. See the summaries by Queller (1986, pp. 251–252) and especially Romano (1987, pp. 6–10).

8. See Lopez’s (1971) seminal book, The Commercial Revolution of the Middle Ages, 950–1350. See also North and Thomas (1973), De Long and Shleifer (1993, table 1), and Findlay and O’Rourke (2008, p. 83).

9. The number of legible names in 960, 971, and 982 was 48, 72, and 128, respectively. The number of new names as a percentage of legible names was 50%, 81%, and 51%, respectively. (For 960, we conservatively assumed that the percentage was 50% based on Castagnetti’s statement that the new names were in the majority.) See Castagnetti (1992a, pp. 624–628).

10. Very little is know about dogal elections; indeed, the earliest account dates from 1071, well after the 1032 constraints on the executive were put in place. In 1071, there was no slate of candidates and the Doge was “elected” by acclamation.

11. The information underlying this figure and much of this subsection is available from many sources, for example, Castagnetti (1992a). The only contemporary source for most of this period is Chronicon Venetum by John the Deacon, circa 1008. (We use the edition by Monticolo 1890). Since modern dogal histories are frequently wrong, we often resort directly to the Chronicon.

12. The first two boxes are Orso Participazio and his son Giovanni Participazio ii. The second two boxes are Orso Participazio ii and his son Pietro Participazio.

13. One obtains a deeper understanding of Venice’s 1032 constraints on the executive by comparing them to those in Genoa and England. According to Acemoglu, Johnson, and Robinson (2005) and Acemoglu and Robinson (2012, chapter 7), England’s initial constraints prevented the Crown from expropriating merchants’ assets, thus removing the fetters to international trade. According to Greif (1995, 2006b) and Greif and Laitin (2004), Genoese constraints on the executive fostered cooperation between clans, thus facilitating the mobilization of resources needed to engage in further trade. Put crudely, English merchants needed to get the Crown off their backs, whereas Genoese merchants needed an executive to keep merchant clans off each other’s backs. In both cases the resulting institutions involved a strengthened parliament, but the resulting long-term growth outcomes were very different. In the period leading up to the 1032 reforms, Venice lay somewhere between the English and Genoese cases. As in England, the ambitions of key figures (e.g., the feudal ambitions of Doge Pietro Candiano iv and Doge-pretender Stephano Coloprino) actively threatened the merchant economy. As in Genoa, the nepotism of Doges exacerbated interclan rivalry and weakened Venice’s ability to collectively mobilize naval and mercantile resources. Thus, the 1032 Venetian constraints on the executive were intended to eliminate expropriation by the executive (as in England) and promote interclan cooperation (as in Genoa).

14. Note that we are not selectively looking at jumps in the number of new endorsers: jumps occur only twice (in the second half of the tenth century and in the 1090s), and we have reviewed both instances.

15. The Michiel Doges were Vitale Michiel i (1096–1102), his grandson Domenico Michiel (1117–1130), Domenico’s son-in-law Pietro Polani (1130–1148), and Vitale Michiel ii (1156–1172). Vitale Michiel ii is the only case in Figure II where the relationship is not certain: he is either Domenico’s son or belonged to a different branch of the Michiel family.

16. These events are described by the contemporary Byzantine historian Choniates. See Choniates (1984, pp. 50–51, 97–98). The figure of 10,000 is from the early thirteenth-century Historia Ducum Veneticorum, which is reprinted in Berto (1999, chapter 18).

17. A feature of this change was that it reduced the arrengo’s (popular assembly’s) role in the election of the Doge. It was thus a victory of the leading merchants over the popolo. A fuller treatment of our subject would deal with the interactions between merchants and the popolo and the rents transferred to the popolo to maintain their cooperation. On this topic see Pullan (1971) and Romano (1987) and our brief discussion of the Venetian patronage system in Section VI.

18. While the documentary evidence about the election of the Doge in this period is scarce, the Great Council gradually gained control over the election process as the rules evolved toward the election protocol of 1268. This is famous for its alternating use of randomizations and nominations, but behind this neat fact hides something more fundamental: from that date the Doge was elected by the members of the Great Council.

20. This observation is further supported by the dating of the Law Merchant. Berman and Kaufman (2004) date its inception to the growth of trade in the eleventh century and cite the 1095 Amalphitan maritime code as its earliest written form. Documents that refer to substantive merchant law appear shortly after 1100 and by 1200 formal commercial courts appear (Kadens 2004).

21. For Venice, this is apparent from the collection of the earliest commercial contracts (Morozzo della Rocca and Lombardo 1940). See Wickham (2003) and Radding (1988) for Tuscan and Lombard evidence, respectively.

22. The penalty is double the amount of capital and profit plus a 20% annual interest (6/5 – 1).

23. We find multiple references to Zaccaria Stagnario and his family in original commercial contracts of the period. The paragraph is based on documents 49, 72, 290, 415, 444, 478, and 479 in the collection edited by Morozzo della Rocca and Lombardo (1940), on the original documents from the Archivio di Stato di Venezia with catalog numbers UD81001346 and UD02000581, and on Borsari (1988) and Robbert (1999). The salary of the Duke of Crete in 1224 was 350 hyperpera (Robbert 1994, table 7).

24. These included the Sanudos in Naxos, the Ghisi in Tinos and Mykonos, Marco Dandolo in Andros, Leonardo Foscolo in Anaphe, Marco Venier in Cerigo, Iacopo Barozzi in Santorini, Iacopo Viear in Cerigotto, and of course, the Corners in Crete. See O’Connell (2009, p. 18). Note that we are not distinguishing between formal and informal colonial possessions.

25. Nobility rents are famously documented by Queller (1986), who argues that the nobility used the Great Council corruptly and to great personal advantage. Other examples of nobility rents include a welfare system that transferred public funds to impoverished members of noble families, the use of influence peddling, and the subsidization of the nobility-dominated galley trade.

26. The movement from A to B implies a hollowing out of the middle-wealth group, but this plays no role for our conclusions. See the Appendix, in particular Figure A.2, for details.

27. There are many ways of modeling the drop in profits. Here it is caused by competition for inputs (traveling merchants) which drives up costs. Alternatively, it could have been modeled as reduced revenue from greater competition in product markets.

28. Recall that at the time of the Serrata, the noble population share was frozen at forumla where forumla is PL valued at point S. Rearranging this implies that the power of nobles, forumla, is less than the power of commoners (forumla).

29. There are very partial membership lists for 1296–1297. No other years of data are available. Portions of these data have been used by Cracco (1967), Chojnacki (1973), and Rösch (1989). Like us, Rösch (1989) emphasizes that a majority of seats in the Great Council were held by a relatively small number of powerful families. However, his analysis stops in 1282. For us, it is essential to carry the analysis to 1296 so that we can also show that there was a high degree of mobility into and out of the Great Council between the 1260s and the 1290s and that some established families were seeing their seat shares eroded. In addition, Great Council membership in 1293–1296 served as the basis for the hereditary nobility established during the Serrata.

30. We use an extended definition of family (casata). Sometimes a casata was made up of multiple branches (rami), but separating these systematically is not feasible. See Raines (2003, pp. 23–25), who uses a comparable definition for the post-1297 period. The standardization of family names presents many difficulties. The same family name appears sometimes in Latin and other times in Italian (e.g., Mauroceno or Morosini). There are multiple patronymic prefixes (d’, da, de, di, dalla, della, de ca’, de cha, de Casa, da Casa, etc.). There are also many spelling variants (up to 13 in the case of the Sesendillo family), the doubling of the n or of the l being the most common variants. Other variations reflect differences between Venetian and Italian, such as the alternative spellings ç, z, or zh for the interdental voiceless fricative (a sound that is used in Venetian but not in Italian). The standardization was done with careful assistance from Lisa Chen and Jennifer Konieczny, doctoral students in the Medieval History Department at the University of Toronto. Lisa Chen specializes in Venetian literary texts. Jennifer Konieczny specializes in Florentine legal texts.

31. Of these 50 new families, only 3 appear in either Rösch’s (1989) list of families that were politically prominent in 960–1141 or in Castagnetti’s (1995) list of families providing high-office holders in 1142–1204.

32. The settlement of Caotorta with Zaccaria Stagnario is in document number 75 in Lombardo and Morozzo della Rocca (1953). The Nicola were investors in the colleganza of document number 811 in Morozzo della Rocca and Lombardo (1940). The Barastro appear in the colleganza of documents numbers 749, 751, 794, and 834 in Morozzo della Rocca and Lombardo (1940).

33. The compounded average annual exit rate between 1265 and 1293 is the γ that solves 162(1 – γ)(1293–1265) = 162 – 47, which implies an annual exit rate γ = 1.2%. We can compare this to the exit rate after the Serrata using exit dates for families from Raines (2003, appendix 1). Of the 165 families with seats in 1293–1296, only 36 exited the Great Council between 1323 and 1500, which implies an exit rate after the Serrata of 0.14%. Clearly, the decline in the exit rate by a factor of nine after the Serrata cannot be explained by demographics, for example, the death of an entire family. If anything mortality rates substantially increased during and after the plague of 1348. The point here is that prior to the Serrata there was a high probability of exit from politics for nonbiological reasons.

34. The relevant motions are in Cessi (1931–1950, volume 1), p. 156, no. 118 (October 3, 1286), pp. 156–157, no. 120 (October 5, 1286), p. 157, no. 123 (October 17, 1286), and p. 396, no. 6 (March 6, 1296), and in Cessi (1931–1950, volume 3), pp. 417–418, no. 104 (“The Last Day” of February 1297).

35. In 1298, membership of the Council of Forty was formally restricted to families who had been in the Great Council for at least two generations.

36. Men took their seats at age 25 unless they won the Balla D’Oro lottery, in which case they took their seats at age 18.

37. Rösch (2000) disagrees, arguing that the revolts were personally motivated, while Hazlitt (1966, chapter 20) points to interclan rivalry.

38. Its membership was secret and drawn exclusively from the most powerful families. As few as 20 families controlled the Council of Ten in the fourteenth century (Ruggiero 1980, p.16).

39. Note that in our model, the state’s coercive capacity is exogenous and captured by the parameter β, which measures the fighting effectiveness of Great Council members (nobles) relative to nonmembers. The build-up of state capacity could be introduced into our model by having β be a concave function of the costly effort devoted by nobles to building coercive capacity. There would then be some optimal combination of co-optation (i.e., Enlargement) and coercion that would be used by families in the Great Council to accomplish the Serrata.

40. This included some who had displayed their loyalty by fighting to defeat revolts.

41. Chojnacki (1973) argued that Great Council membership was open for up to seven or eight decades after 1297, but his argument has since been shown to be inaccurate. See in particular the criticism by Ruggiero (1979, pp. 248–249): “A yet more troubling aspect of Chojnacki’s statistics is his failure to look for differential rates of change in the additions of new families to the Major Council in the period 1298–1379. Most scholars agree that there was a very rapid addition of families in the years immediately following the Serrata… . Considering as a whole the period from the 1290s to 1379, thus, gravely distorts the real picture of elite stability following 1310.” Todesco (1989, p. 8, our translation) similarly notes that “a restrictive policy was instead implemented from 1310, when the admission of new members was made increasingly selective. While initially twelve votes in the Council of Forty were sufficient for their approval, after various changes, from 29 June 1310, thirty votes in the Council of Forty and a two-thirds majority of the Great Council were required.” Todesco also notes that “the final closure occurred around 1320. … Any further admissions were limited to cadet branches of old families that returned from the colonies, or the lords of terra firma or the governors whose appointment, however, only had an honorary meaning” (Todesco 1989, pp.126–127, our translation). The only exception was the admission in 1381 of families who had made significant contributions to the Battle of Chiogga.

42. The average size in 1261–1296 is calculated from our Great Council membership data. No systematic records of Great Council membership have survived for the period immediately afterward, so all counts after 1297 are estimates. The approximate size circa 1323 is calculated on the basis of the 670 and 664 votes cast in the Great Council for the 1320 and 1324 elections of the procurator of Saint Mark and a typical absentee rate of 30% during such elections (Todesco 1989, p. 120 and appendix 1).

43. Raines (2003) provides the most careful systematic analysis of the presence of different families in the Great Council from 1297 onward, recording the first and last known presence of each family. Combining the data in Raines (2003, appendix 1) with our data for 1261–1296 results in (i) 76 families that she records as present in the Great Council in 1323 but who had no seats in 1261–1296 and (ii) another 31 families present in the Great Council in 1323 with no seats in 1293–1296, but who had seats in earlier years. Of the 76 new families, only 6 appear in either Rösch’s (1989) list of families that were politically prominent in 960–1141 or in Castagnetti’s (1995) list of families providing high-office holders in 1142–1204.

44. See Doumerc and Stöckly (1995), Doumerc (1997), Hocquet (1997), Judde de Larivière and Doumerc (1998), and Stöckly (1995) for detailed discussions of the overhaul of the galley trade.

45. Other reasons for the Capitulare included prevention of tax evasion and exclusion of noncitizens from long-distance trade, for example, Cessi (1952). The Capitulare also addressed periodic inventory build-ups in Venetian warehouses (Hocquet 1997, pp. 595–597).

46. We have benefited enormously from numerous discussions with Yadira González de Lara on the coding of colleganza. Colleganza contracts are most clearly identified by the statement that, in the event of profits, the traveling merchant receives a share of these profits, for example, “Reliquam quartam partem in me retinere debeam.” See the colleganza in Figure III. Settlements of colleganza require a careful reading to understand what type of contract is being settled. The most common difficulty is distinguishing between a colleganza and a sea loan. Where no other distinguishing features are available, we follow González de Lara (2008) in classifying a contract as a colleganza if the traveling merchant rendered accounts under oath. Sometimes both the original colleganza and its settlement have survived, in which case we count only the original contract; however, most often only one of the two has survived, in which case we date the contract to the date of the original colleganza (which is almost always specified in the settlement). There are a small number of other related contracts dealing with the transfer of colleganza obligations, and we include these as well. There are two settlements of colleganza for which only a fragment of the original parchment has survived. Since the names of the merchants are missing, we exclude these two.

47. Recall that 1323 marks the completion of the Serrata of the Great Council.

48. This 1326 colleganza is a bizarre “coals to Newcastle” colleganza. The non-noble traveling merchant is carrying mink fur to Tana on the Black Sea on behalf of a nobleman and his partner, despite the fact that Tana was a center for fur exports.

49. In his excellent book, Kedar (1976) also documents this decline in commoner participation in colleganza between 1240–1261 and 1310–1323. However, he does not look at commoner participation after the Capitulare Navigantium. He is also unclear about how he defines nobility—he certainly does not define it by reference to Great Council participation.

50. The Table I numbers for 1310–1342 are accurate and establish our point. The pre-1261 numbers are less accurate. We explain the inaccuracy by way of two examples. (i) If a merchant in 1240 whose family had never sat in the Great Council entered the Great Council in 1261 he is classed as a noble, thus leading us to understate commoner participation. (ii) If a merchant in 1240 was in the Great Council in 1240 but not after 1261 he is incorrectly classified as a commoner, thus leading us to overstate commoner participation. It is hard to see how such misclassifications could account for the drop in commoner participation after 1323. However, as a robustness check we also classified a merchant as noble if his family name appears in Rösch’s (1989) list of families that were politically prominent in 960–1141 or in Castagnetti’s (1995) list of families providing high-office holders in 1142–1204. This reduces commoner participation rates in 1073–1261 by just 6 percentage points, from 46% to 40%; it reduces commoner participation rates in 1310–1323 from 27% to 25%; and it still leaves a single colleganza involving a commoner after 1324. In short, it is hard to make a case that pre-1261 misclassification of nobles and commoners explains away our result. A second potential source of concern is that data for different periods comes from different notaries who may have worked with different types of customers. If we restrict ourselves to the period 1317–1342, we can compare commoner participation in colleganza contracts from a single notary who worked both before and after 1324, the priest Felice de Merlis. We also know that he worked for all sorts of customers, for example, see the many references to his social contacts in Romano (1996). Looking only at this notary’s contracts we see commoner participation dropping from 16% in 1317–1323 (commoners were involved in 7 contracts out of 44) to 3% in 1324–1342 (a commoner was involved in a single contract out of 34). Restated, even looking at contracts from the same notary and starting in a late period when the Serrata was almost complete, there is a sharp drop after 1324. Thus, it is highly unlikely that our results can be explained by differences across notaries.

51. Our results should not be misconstrued to mean that only the rich traded. There continued to be trading by commoners in less profitable routes, mostly on bulk commodities (e.g., Apellániz 2013). Our results mean that it was now much more difficult for commoners to break into the most lucrative segments of long-distance trade.

52. Numbers for the cost of the charter and freight of galleys are from Doumerc (2003, p. 158), who also notes that the five galleys in the 1409 muda to Flanders were carrying merchandise worth 460,000 ducats, the equivalent of one and a half tons of gold. Data on winning bids are from Hocquet (1997, table 1, p. 595).

53. Ships had long been divided into shares, and these shares appear as investments in colleganza. Indeed, they are referenced in the earliest extant colleganza dating from 1073 (Lopez and Raymond 1967, document 82).

54. The only systematic wealth data source is the tax register (estimo) of 1379, prepared to finance the war with Genoa. Unfortunately this comes just before the period of recovery and expansion in Venetian trade and provides no time dimension.

55. Marriage networks have been studied in other contexts. Perhaps the best-known example is the study by Padgett and Ansell (1993) documenting the rise of the Medici family in Florence through the strategic use of marriage alliances. Although they do not explicitly link their study of high marriage inequality to high economic inequality, if these two were as correlated in Florence, as we show they were in Venice, then Padgett and Ansell’s work suggests that high inequality was not unique to Venice.

56. The handwritten list also contains a few earlier marriages (23 in 1398 and another 33 scattered over the period 1348–1397), but these are far too few to be useful. The list only contains marriages involving a noble husband. Interestingly, the frequency of noble husbands marrying non-noble wives falls over time, from 11.8% in 1400–1499 to 7.7% in 1500–1599. This is another way in which non-noble families were increasingly excluded.

57. Nodes are arranged following the Fruchterman and Reingold (1991) algorithm, which applies stronger attractive forces to more heavily connected nodes and repulsive forces to all nodes.

58. Let M be a matrix where each element mij is the number of marriages between family i and family j. The centrality of family i, denoted by ci, is proportional to the weighted sum of the centrality of the families it has married, with weights given by the respective number of marriages: forumla. In matrix notation, λc = Mc. Thus, c is an eigenvector of M and λ its corresponding eigenvalue. To obtain a measure with non-negative values, eigenvector centrality is defined as the eigenvector c associated with the largest eigenvalue. Note that c is unique only up to a scaling factor, so it allows for comparisons within a network, but not across networks or across periods of a given network.

59. A second possible external validation of our marriage centrality measure is to compare it with the tax assessment of 1379. The Spearman rank correlation between eigenvector centrality in the marriage network and wealth in the 1379 assessment for individual families is 0.74 (the rank in the 1379 assessment is based on the transcription in Luzzatto 1929, pp. 139–195). This is another strong indication that our marriage centrality measure is informative of economic status.

60. This example is pieced together from information in Tucci (1981, p. 183), Judde de Larivière and Doumerc (1998, pp. 16–17), Judde de Larivière (2008, p. 181), Chojnacki (1973, appendix ii), and our marriage database.

61. The only family with no pre-Serrata seats to play a significant role in the muda was the Garzoni. However, this is a case of geographical mobility, not wealth mobility. When Bandin Garzoni arrived in Venice from Lucca, he was already immensely wealthy from the grain trade, so much so that he ranked in the top four in the 1379 wealth assessment. He was granted citizenship in 1374, and his two sons were ennobled in 1381 as an exceptional reward for his enormous contribution to the Battle of Chiogga.

62. See also Chojnacki (1985, p. 245) and citations therein.

63. Grazie are also documented by Romano (1987, chapter 6) and, in the context of stato da mar, by O’Connell (2009, chapter 5). O’Connell’s view of the new system is somewhat less cynical than Queller’s (O’Connell 2009, p. 43). On the role of grazie in preventing revolt see Romano (1987, p. 125) and Chojnacki (1997, pp. 676–677). On the poverty of nobles, revolt and grazie in the fifteenth and sixteenth centuries see Pullan (1971). Grazie had many other dimensions beside government jobs; however, for simplicity we focus only on this one element. Venetian historians refer to grazie when referring to the entire period, both pre- and post-Serrata. We depart from this by referring to the system that emerged after grazie came under the control of the Great Council as “the new patronage system.” This is not common usage and ignores the many elements of continuity in patronage before and after the Serrata.

64. See Chojnacki (1985, p. 246) and especially Queller (1986, chapter 2).

65. See the Romano (1996) quote at the start of this subsection. Note that as socially stratified as Venice was, there remained interactions between nobles and non-nobles; see Martin and Romano (2000). Note also that our discussion of the formalization of the definition of the nobility is quite similar to Chojnacki (1997) (though as already documented it is not similar to the argument in Chojnacki 1973).

66. The monopolization of the galley trade, besides restricting entry, also led to regulatory capture. By decree, galleys could always be requisitioned for state service, such as war. However, after 1500 the families that controlled the galleys increasingly subverted such state demands when they conflicted with the families’ mercantile needs. See Stöckly (1995), Doumerc and Stöckly (1995), Judde de Larivière and Doumerc (1998), Doumerc (2003), and Judde de Larivière (2008). For other negative effects of oligarchization see Acemoglu (2008). These impacts include rent-seeking behavior, misallocation of talent, and a decline in entrepreneurship, all of which are apparent in Venice by 1500.

67. forumla ensures that if someone invests all his wealth in the safe asset and gets no labour income, then his offspring will be poorer than he is: forumla or forumla.

68. More specifically, the ability of a borrower to default on a loan, subject to a fixed nonmonetary punishment if caught, leads to credit rationing. As in Ghatak and Jiang (2002), we let the probability of being caught be 0 so that only those with enough wealth can invest. (If we let the probability of being caught be strictly positive, all wealth thresholds described below are simply raised by a constant).

69. An individual with inherited wealth w < wM = I ends up with wealth v + (r + 1)w if he chooses to become a craftsman and with expected wealth forumla if he chooses to become a traveling merchant. Equating the two yields forumla. (Recall that forumla is the expected return on investment I and forumla is the traveling merchant’s share of colleganza profits, so that forumla is a traveling merchant’s expected income from a colleganza.)

70. Consider an individual with inherited wealth w < wH = μI. If he chooses to become a self-financed merchant, he expects to end up with wealth forumla. If he chooses to become a sedentary merchant he expects to end up with wealth forumla, where forumla is his expected profits after payments to traveling merchants, μI is the capital invested in colleganza, and (r + 1)(w – μI) is the wealth invested in the safe asset plus returns on this. Equating the expected returns to self-financed and sedentary merchants yields forumla.

71. A reader whose knowledge of colleganze is from secondary sources may think that profit shares are fixed at one fourth. This is not the case: we observe many colleganza with alternative profit shares. Venetian law was explicit that profit shares need not be one fourth. In the Gli Statuti (Besta and Predelli 1901) in the chapter on Observing Contracts, we have “We decree also concerning contracts that it ought to be observed that he who receives money from anyone, so that he may make a profit with it, disregards nothing in the contract, but profiting with the received money up to the time stated, holds for himself from the success the fourth part or as much as is contained in the contract. Moreover, in the accustomed manner he simultaneously ought to give to the creditor the remaining parts with the capital” (our translation, emphasis added).

72. The explanation is as follows. Consider the first row, which deals with the change in the measure of low-wealth people. This can be written as forumla. Among those who initially have low wealth, a fraction σ are successful as traveling merchants and so their children will start life as middle-wealth individuals. This accounts for the term –σPL. Among those who initially have middle wealth, a fraction (1 – σ) are unsuccessful as self-financed merchants, and so their children start life as low-wealth people. This accounts for the term (1 − σ)PM. Among those who initially have high wealth, there is one sedentary merchant for every μ traveling merchants, that is, there are forumla sedentary merchants. The rest, forumla, are self-financed merchants. A fraction (1 – σ) of these are unsuccessful and their children begin life as low-wealth people. This accounts for the term forumla. Finally, individuals mature at a rate λ and the dynamics apply only to them.

References

Acemoglu
Daron
.
“Why Not a Political Coase Theorem? Social Conflict, Commitment and Politics”
.
Journal of Comparative Economics
.
2003
;
31
:
620
652
.

Acemoglu
Daron
.
“Oligarchic vs. Democratic Societies”
.
Journal of the European Economic Association
.
2008
;
6
:
1
44
.

Acemoglu
Daron
,
Angélica Bautista
María
,
Querubín
Pablo
,
Robinson
James A
.
Helpman
Elhanan
.
“Economic and Political Inequality in Development: The Case of Cundinamarca, Colombia”
.
Institutions and Economic Performance
.
2008
;
Cambridge, MA
:
Harvard University Press
:
181
245
.

Acemoglu
Daron
,
Johnson
Simon
,
Robinson
James A
.
“The Rise of Europe: Atlantic Trade, Institutional Change and Economic Growth”
.
American Economic Review
.
2005
;
95
:
546
579
.

Acemoglu
Daron
,
Robinson
James A
.
Economic Origins of Dictatorship and Democracy
.
2006
;
Cambridge
:
Cambridge University Press
.

Acemoglu
Daron
,
Robinson
James A
.
Why Nations Fail: The Origins of Power, Prosperity, and Poverty
.
2012
;
New York
:
Crown Publishing
.

Apellániz
Francisco
.
“Venetian Trading Networks in the Medieval Mediterranean”
.
Journal of Interdisciplinary History
.
2013
;
44
:
157
179
.

Banerjee
Abhijit V
,
Newman
Andrew F
.
“Occupational Choice and the Process of Development”
.
Journal of Political Economy
.
1993
;
101
:
274
298
.

Berman
Harold J
.
Law and Revolution: The Formation of the Western Legal Tradition
.
1983
;
Cambridge, MA
:
Harvard University Press
.

Berman
Harold J
,
Kaufman
Colin
.
“Law of International Commercial Transactions (Lex Mercatoria)”
.
Harvard International Law Journal
.
2004
;
19
:
221
277
.

Berto
Luigi Andrea
.
“Historia ducum Veneticorum”
.
Testi storici veneziani (xixiii secolo)
.
1999
;
Padova, Italy
:
CLEUP
:
1
83
.

Besley
Timothy
,
Burchardi
Konrad
,
Ghatak
Maitreesh
.
“Incentives and the de Soto Effect”
.
Quarterly Journal of Economics
.
2012
;
127
:
237
282
.

Besta, Enrico, and Riccardo Predelli, eds., Gli statuti civili di Venezia anteriori al 1242 (Venice: Federico Visentini Editore, 1901)
.

Bonacich
Phillip
.
“Factoring and Weighting Approaches to Clique Identification”
.
Journal of Mathematical Sociology
.
1972
;
2
:
113
120
.

Borsari
Silvano
.
Venezia e Bisanzio nel xii secolo: i rapporti economici
.
1988
;
Venice
:
Deputazione Editrice
.

Braudel
Fernand
.
Civilization and Capitalism, 15th–18th Century, Volume iii: The Perspective of the World
.
1984
;
New York
:
Harper & Row
.

Braudel
Fernand
,
Tenenti
Alberto
.
Abel
Wilhem
,
Borchardt
Knut
,
Kellenbenz
Hermann
,
Zorn
Wolfgang
.
“Michiel da Lezze, Marchand Venitien (1497–1514)”
.
Wirtschaft, Geschichte und Wirtschaftsgeschichte. Festschrift zum 65 Geburtstag von Friedrich Lütge
.
1966
;
Stuttgart
:
Gustav Fischer
:
38
73
.

Brown
Horatio F
.
“The Venetians and the Venetian Quarter in Constantinople to the Close of the Twelfth Century”
.
Journal of Hellenic Studies
.
1920
;
40
:
68
88
.

Castagnetti
Andrea
.
Cracco Ruggini
Lellia
,
Pavan
Mario
,
Cracco
Giovanni
,
Ortalli
Gherardo
.
“Famiglie e affermazione politica”
.
Storia di Venezia dalle origini alla caduta della Serenissima, Volume i: Origini-età ducale
.
1992a
;
Rome
:
Istituto dell’Enciclopedia Italiana
:
613
644
.

Castagnetti
Andrea
.
La Societá Veneziana nel Medioevo i, Dai Tribuni ai Giudici
.
1992b
;
Verona
:
Libreria Universitaria Editrice
.

Castagnetti
Andrea
.
Cracco
Giovanni
,
Ortalli
Gherardo
.
“Il primo comune”
.
Storia di Venezia dalle origini alla caduta della Serenissima, Volume ii: L’età del comune
.
1995
;
Rome
:
Istituto dell’Enciclopedia Italiana
:
81
130
.

Cessi, Roberto, ed., Deliberazioni del Maggior Consiglio di Venezia, 3 vols. (Bologna: Nicola Zanichelli, 1931–1950)
.

Cessi
Roberto
.
Cessi
Roberto
.
“L’‘officium de navigantibus’ ed i sistemi della politica commerciale veneziana nel secolo xiv
.
Politica ed Economica di Venezia nel Trecento
.
1952
;
Rome
:
Edizioni di Storia e Letteratura
:
23
61
.

Cessi
Roberto
.
Hussey
Joan M
.
“Venice to the Eve of the Fourth Crusade”
.
The Cambridge Medieval History, Volume iv. The Byzantine Empire, Part i: Byzantium and Its Neighbours
.
1966
;
Cambridge
:
Cambrige University Press
:
250
274
.

Chojnacki
Stanley
.
Hale
John R
.
“In Search of the Venetian Patriciate: Families and Factions in the Fourteenth Century”
.
Renaissance Venice
.
1973
;
London
:
Faber
:
47
90
.

Chojnacki
Stanley
.
“Kinship Ties and Young Patricians in Fifteenth-Century Venice”
.
Renaissance Quarterly
.
1985
;
38
:
240
270
.

Chojnacki
Stanley
.
“Social Identity in Renaissance Venice: The Second Serrata
.
Renaissance Studies
.
1994
;
8
:
341
358
.

Chojnacki
Stanley
.
Arnaldi
Girolamo
,
Cracco
Giorgio
,
Tenenti
Alberto
.
“La formazione della nobilità dopo la Serrata”
.
Storia di Venezia dalle origini alla caduta della Serenissima, Volume iii: La Formazione dello Stato Patrizio
.
1997
;
Rome
:
Istituto dell’Enciclopedia Italiana
:
641
725
.

Chojnacki
Stanley
.
Martin
John
,
Romano
Dennis
.
“Identity and Ideology in Renaissance Venice: The Third Serrata
.
Venice Reconsidered, the History and Civilization of an Italian City-State 1297–1797
.
2000
;
Baltimore
:
Johns Hopkins University Press
:
263
294
.

Choniates, Niketas, O City of Byzantium: Annals of Niketas Choniates (Detroit: Wayne State University Press, 1984), translated by Harry J. Magoulias
.

Cracco
Giorgio
.
Società e stato nel medioevo veneziano (secoli xiixiv)
.
1967
;
Florence
:
Leo S. Okchki
.

De Long
J Bradford
,
Shleifer
Andrei
.
“Princes and Merchants: European City Growth before the Industrial Revolution”
.
Journal of Law and Economics
.
1993
;
36
:
671
702
.

de Roover
Raymond
.
Postan
M M
,
Rich
E E
,
Miller
Edward
.
“The Organization of Trade”
.
The Cambridge Economic History of Europe, Volume 3: Economic Organization and Policies in the Middle Ages
.
1965
;
Cambridge
:
Cambridge University Press
:
42
118
.

de Soto
Hernando
.
The Other Path: The Invisible Revolution in the Third World
.
1989
;
New York
:
HarperCollins
.

Do
Quy-Toan
,
Levchenko
Andrei A
.
“Trade, Inequality, and the Political Economy of Institutions”
.
Journal of Economic Theory
.
2009
;
144
:
1489
1520
.

Doumerc
Bernard
.
Arnaldi
Girolamo
,
Cracco
Giorgio
,
Tenenti
Alberto
.
“Gli armamenti marittimi”
.
Storia di Venezia dalle origini alla caduta della Serenissima, Volume iii: La Formazione dello Stato Patrizio
.
1997
;
Rome
:
Istituto dell’Enciclopedia Italiana
:
617
640
.

Doumerc
Bernard
.
Hattendorf
John B
,
Unger
Richard W
.
“An Exemplary Maritime Republic: Venice at the End of the Middle Ages”
.
War at Sea in the Middle Ages and the Renaissance
.
2003
;
London
:
Boydell Press
:
151
165
.

Doumerc
Bernard
,
Stöckly
Doris
.
“L’évolution du capitalisme marchand à Venise: le financement des galere da mercato à la fin du xve siècle”
.
Annales Histoire, Sciences Sociales
.
1995
;
50
:
133
157
.

Findlay
Ronald
,
O’Rourke
Kevin H
.
Power and Plenty: Trade, War, and the World Economy in the Second Millennium
.
2008
;
Princeton, NJ
:
Princeton University Press
.

Fruchterman
Thomas M J
,
Reingold
Edward M
.
“Graph Drawing by Force-Directed Placement”
.
Software—Practice and Experience
.
1991
;
21
:
1129
1164
.

Galor
Oded
,
Zeira
Joseph
.
“Income Distribution and Macroeconomics”
.
Review of Economic Studies
.
1993
;
60
:
35
52
.

Ghatak
Maitreesh
,
NienHuei Jiang
Neville
.
“A Simple Model of Inequality, Occupational Choice, and Development”
.
Journal of Development Economics
.
2002
;
69
:
205
226
.

Giomo
Giuseppe
.
“Indice per nome di donna dei matrimoni dei patrizi Venetie”
.
archivio di Stato di Venezia, indice 82 ter i/ii, date unknown
.

González de Lara
Yadira
.
“The Secret of Venetian Success: A Public-Order, Reputation-Based Institution”
.
European Review of Economic History
.
2008
;
12
:
247
285
.

González de Lara
Yadira
.
Koppell
Jonathan G S
.
“Regulation and Litigation: The Protection of Investor Rights in Late-Medieval Venice”
.
Origins of Shareholder Advocacy
.
2010
;
New York
:
Macmillan
:
101
122
.

González de Lara
Yadira
.
Schofield
Norman
,
Caballero
Gonzalo
.
“Self-Enforcing, Public-Order Institutions for Contract Enforcement: Litigation, Regulation, and Limited Government in Venice, 1050–1350”
.
Political Economy of Institutions, Democracy and Voting
.
2011
;
Berlin
:
Springer
:
95
116
.

González de Lara
Yadira
,
Greif
Avner
,
Jha
Saumitra
.
“The Administrative Foundations of Self-Enforcing Constitutions”
.
American Economic Review Papers and Proceedings
.
2008
;
98
:
105
109
.

Goy
Richard J
.
The House of Gold: Building a Palace in Medieval Venice
.
1992
;
Cambridge
:
Cambridge University Press
.

Greif
Avner
.
“Institutions and International Trade: Lessons from the Commercial Revolution”
.
American Economic Review
.
1992
;
82
:
128
133
.

Greif
Avner
.
“Contract Enforceability and Economic Institutions in Early Trade: The Maghribi Traders Coalition”
.
American Economic Review
.
1993
;
83
:
525
548
.

Greif
Avner
.
“Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on collectivist and Individualist Societies”
.
Journal of Political Economy
.
1994
;
102
:
912
950
.

Greif
Avner
.
“Political Organizations, Social Structure, and Institutional Success: Reflections from Genoa and Venice During the Commercial Revolution”
.
Journal of Institutional and Theoretical Economics
.
1995
;
151
:
734
740
.

Greif
Avner
.
“Commitment, Coercion, and Markets: The Nature and Dynamics of Institutions Supporting Exchange”
.
Handbook for New Institutional Economics
.
Claude Menard and MaryM. Shirley, eds. (Norwell, MA: Kluwer, 2005), chapter 28
.

Greif
Avner
.
“The Birth of Impersonal Exchange: The Community Responsibility System and Impartial Justice”
.
Journal of Economic Perspectives
.
2006a
;
20
:
221
236
.

Greif
Avner
.
Institutions and the Path to the Modern Economy: Lessons from Medieval Trade
.
2006b
;
Cambridge
:
Cambridge University Press
.

Greif
Avner
.
Helpman
Elhanan
.
“The Impact of Administrative Power on Political and Economic Developments: Toward a Political Economy of Implementation”
.
Institutions and Economic Performance
.
2008
;
Cambridge, MA
:
Harvard University Press
:
17
63
.

Greif
Avner
,
Laitin
David D
.
“A Theory of Endogenous Institutional Change”
.
American Political Science Review
.
2004
;
98
:
633
652
.

Greif
Avner
,
Milgrom
Paul
,
Weingast
Barry R
.
“Coordination, Commitment, and Enforcement: The Case of the Merchant Guild”
.
Journal of Political Economy
.
1994
;
102
:
745
776
.

Grubb
James S
.
“When Myths Lose Power: Four Decades of Venetian Historiography”
.
Journal of Modern History
.
1986
;
58
:
43
94
.

Hazlitt
W Carew
.
The Venetian Republic, Its Rise, Its Growth, and Its Fall A.D. 409–1797
.
1966
;
New York
:
AMS Press
.

Hocquet
Jean-Claude
.
Arnaldi
Girolamo
,
Cracco
Giorgio
,
Tenenti
Alberto
.
“I Meccanismi dei Traffici”
.
Storia di Venezia dalle origini alla caduta della Serenissima, Volume iii: La Formazione dello Stato Patrizio
.
1997
;
Rome
:
Istituto dell’Enciclopedia Italiana
:
529
616
.

Jackson
Matthew O
.
Social and Economic Networks
.
2008
;
Princeton, NJ
:
Princeton University Press
.

Jha
Saumitra
.
“Financial Innovations and Political Development: Evidence from Revolutionary England”
.
Research Paper 2005, Stanford University Graduate School of Business, 2010
.

Jones
Benjamin F
,
Olken
Benjamin A
.
“Hit or Miss? The Effect of Assassinations on Institutions and War”
.
American Economic Journal: Macroeconomics
.
2009
;
1
:
55
87
.

Judde de Larivière
Claire
.
Naviguer, Commercer, Gouverner: Économie Maritime et Pouvoirs à Venise (XV–XVI Siecles)
.
2008
;
Leiden
:
Brill
.

Judde de Larivière
Claire
,
Doumerc
Bernard
.
“Le rôle du patriciat dans la gestion des galères marchandes à Venise au début du seizième siècle”
.
Studi Veneziani, n.s.
1998
;
36
:
57
84
.

Kadens
Emily
.
“Order within Law, Variety within Custom: The Character of the Medieval Merchant Law”
.
Chicago Journal of International Law
.
2004
;
5
:
39
66
.

Kedar
Benjamin Z
.
Merchants in Crisis: Genoese and Venetian Men of Affairs and the Fourteenth-Century Depression
.
1976
;
New Haven, CT
:
Yale University Press
.

Landau
Peter
.
Luscombe
David
,
Riley-Smith
Jonathan
.
“The Development of Law”
.
New Cambridge Medieval History, Volume 4: c.1024–c.1198 Part i
.
2004
;
Cambridge
:
Cambridge University Press
:
113
147
.

Lane
Frederic C
.
“Venetian Merchant Galleys, 1300–1334: Private and Communal Operation”
.
Speculum
.
1963
;
38
:
179
205
.

Lane
Frederic C
.
“Rhythm and Rapidity of Turnover in Venetian Trade of the Fifteenth Century”
.
Venice and History: The Collected Papers of Frederic C. Lane
.
1966
;
Baltimore, MD
:
John Hopkins University Press
:
109
127
.

Lane
Frederic C
.
“Società e stato nel medioevo veneziano by Giorgio Cracco”
.
Speculum
.
1968
;
43
:
497
501
.

Lane
Frederic C
.
Rowe
J G
,
Stockdale
W H
.
“The Enlargement of the Great Council of Venice”
.
Florilegium Historiale: Essays Presented to Wallace K. Ferguson
.
1971
;
Toronto
:
University of Toronto Press
:
236
274
.

Lane
Frederic C
.
Venice: A Maritime Republic
.
1973
;
Baltimore, MD
:
John Hopkins University Press
.

Levchenko
Andrei A
.
“International Trade and Institutional Change”
.
Journal of Law, Economics, and Organizations
.
2013
;
29
:
1145
1181
.

Lombardo, Antonino, and Raimondo Morozzo della Rocca, eds., Nuovi documenti del commercio veneto dei sec. XI–XIII (Venice: Deputazione di storia patria per le Venezie, 1953)
.

Lopez
Robert S
.
The Commercial Revolution of the Middle Ages, 950–1350
.
1971
;
Cambridge
:
Cambridge University Press
.

Lopez
Robert S
,
Raymond
Irving W
.
Medieval Trade in the Mediterranean World
.
1967
;
New York
:
Columbia University Press
.

Luzzatto
Gino
.
I prestiti della Repubblica di Venezia. (Sec. xiii–xiv)
.
1929
;
Padova
:
Lireria Editrice A. Draghi
.

Madden
Thomas F
.
Enrico Dandolo and the Rise of Venice
.
2003
;
Baltimore, MD
:
Johns Hopkins University Press
.

Martin
John
,
Romano
Dennis
.
Martin
John
,
Romano
Dennis
.
“Reconsidering Venice”
.
Venice Reconsidered, The History and Civilization of an Italian City-State 1297–1797
.
2000
;
Baltimore, MD
:
Johns Hopkins University Press
:
1
38
.

McCormick
Michael
.
Origins of the European Economy
.
2001
;
Cambridge
:
Cambridge University Press
.

Milgrom
Paul R
,
North
Douglass C
,
Weingast
Barry R
.
“The Role of Institutions in the Revival of Trade: The Law Merchant, Private Judges, and the Champagne Fairs”
.
Economics and Politics
.
1990
;
2
:
1
23
.

Monticolo, Giovanni, ed., Cronache Veneziane Antichissime (Rome: Istituto Storico Italiano, 1890)
.

Morozzo della Rocca, Raimondo, and Antonino Lombardo, eds., Documenti del commercio veneziano nei secoli XI–XIII, 2 vols. (Turin: Istituto storico per il Medio Evo, 1940)
.

Mueller
Reinhold C
.
Money and Banking in Medieval and Renaissance Venice, Volume ii: Banks, Panics, and the Public Debt 1200–1500
.
1997
;
Baltimore, MD
:
Johns Hopkins University Press
.

Nicol
Donald M
.
Byzantium and Venice: A Study in Diplomatic and Cultural Relations
.
1988
;
Cambridge
:
Cambridge University Press
.

North
Douglass C
.
Institutions, Institutional Change and Economic Performance
.
1990
;
Cambridge
:
Cambridge University Press
.

North
Douglass C
,
Paul Thomas
Robert
.
The Rise of the Western World: A New Economic History
.
1973
;
Cambridge
:
Cambridge University Press
.

Norwich
John Julius
.
A History of Venice
.
1977
;
London
:
Penguin Books
.

O’Connell
Monique
.
Men of Empire: Power and Negotiation in Venice’s Maritime State
.
2009
;
Baltimore, MD
:
Johns Hopkins University Press
.

Padgett
John F
,
Ansell
Christopher K
.
“Robust Action and the Rise of the Medici, 1400–1434”
.
American Journal of Sociology
.
1993
;
98
:
1259
1319
.

Pryor
John H
.
Business Contracts of Medieval Provence
.
1981
;
Toronto
:
Political Institute of Mediaeval Studies
.

Pryor
John H
.
Commerce, Shipping and Naval Warfare in the Medieval Mediterranean
.
1987
;
London
:
Variorum Reprints
.

Pryor
John H
.
Geography, Technology and War: Studies in the Maritime History of the Mediterranean, 649–1571
.
1988
;
Cambridge
:
Cambridge University Press
.

Pullan
Brian
.
Rich and Poor in Renaissance Venice: The Social Institutions of a Catholic State, to 1620
.
1971
;
Cambridge, MA
:
Harvard University Press
.

Queller
Donald E
.
The Venetian Patriciate: Reality versus Myth
.
1986
;
Urbana
:
University of Illinois Press
.

Radding
Charles M
.
The Origins of Medieval Jurisprudence: Pavia and Bologna, 850–1150
.
1988
;
New Haven, CT
:
Yale University Press
.

Radding
Charles M
,
Ciaralli
Antonio
.
The Corpus Iuris Civilis in the Middle Ages: Manuscripts and Transmission from the Sixth Century to the Juristic Revival
.
2006
;
Leiden
:
Brill
.

Raines
Dorit
.
“Cooptazione, Aggregazione e Presenza al Maggior Consiglio: le Casate del Patriziato Veneziano, 1297–1797”
.
Storia di Venezia, Rivista
.
2003
;
1
:
1
35
.

Robbert
Louise Buenger
.
“Money and Prices in Thirteenth-Century Venice”
.
Journal of Medieval History
.
1994
;
20
:
373
390
.

Robbert
Louise Buenger
.
Queller
Donald E
,
Kittell
Ellen E
,
Madden
Thomas F
.
“Domenico Gradenigo: A Thirteenth-Century Venetian Merchant”
.
Medieval and Renaissance Venice
.
1999
;
Urbana
:
University of Illinois Press
:
27
48
.

Romano
Dennis
.
Patricians and Popolani: The Social Foundations of the Venetian Renaissance State
.
1987
;
Baltimore, MD
:
Johns Hopkins University Press
.

Romano
Dennis
.
Housecraft and Statecraft: Domestic Service in Renaissance Venice, 1400–1600
.
1996
;
Baltimore, MD
:
Johns Hopkins University Press
.

Rösch
Gerhard
.
Der Venezianische Adel bis zur Schliessung des Grossen Rates: zur Genese einer Führungsschicht
.
1989
;
Sigmaringen
:
Thorbecke
.

Rösch
Gerhard
.
Martin
John
,
Romano
Dennis
.
“The Serrata of the Great Council and Venetian Society, 1286–1323”
.
Venice Reconsidered, the History and Civilization of an Italian City-State 1297–1797
.
2000
;
Baltimore, MD
:
Johns Hopkins University Press
:
67
88
.

Ruggiero
Guido
.
“Modernization and the Mythic State in Early Renaissance Venice: The Serrata Revisited”
.
Viator
.
1979
;
10
:
245
256
.

Ruggiero
Guido
.
Violence in Early Renaissance Venice
.
1980
;
New Brunswick, NJ
:
Rutgers University Press
.

Sebellico, AndreinaBondi, ed., Felice de Merlis prete e notaio in Venezia ed Ayas 1315–1348, 2 vols. (Venice: Comitato per la Pubblicazione delle Fonti Relative alla Storia di Venezia, 1973)
.

Stöckly
Doris
.
Le système de l’Incanto des galées du marché à Venise (fin XIIIe–milieu XVe siècle)
.
1995
;
Leiden
:
Brill
.

Tiepolo, MariaFrancesca, ed., Domenico prete di S. Maurizio, notaio in Venezia, 1309–1316 (Venice: Comitato per la Pubblicazione delle Fonti Relative alla Storia di Venezia, 1970)
.

Todesco
Maria-Teresa
.
“Andamento demografico della nobiltà veneziana allo specchio delle votazioni nel Maggior Consiglio (1297–1797)”
.
Ateneo Veneto
.
1989
;
160
:
119
164
.

Tucci
Ugo
.
Tucci
Ugo
.
“Costi e ricavi di una galera veneziana ai primi del Cinquecento”
.
Mercanti, navi, monete nel Cinquecento veneziano
.
1981
;
Bologna
:
Il mulino
:
161
230
.

Wickham
Chris
.
Courts and Conflict in Twelfth-Century Tuscany
.
2003
;
Oxford
:
Oxford University Press
.

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com