1 Introduction

Globalization and the Nation State

When the fruits of the Industrial Revolution began to be reaped in the course of the nineteenth century, a period of rapid economic expansion followed. Steamships and trains improved transportation, machines allowed for the mass production of goods, and the telegraph and telephone sped up communication. Global markets for commodities and manufactured goods alike became increasingly interconnected. This growing connectivity benefited from the fact that the world’s nations were cosmopolitan in a way unthinkable now, operating almost entirely without ID requirements, visas or other crossborder restrictions. In addition, this was a time of economic development largely unchecked by labor laws, unions or social safety nets. Finally, it was the height of European imperialism, with about half the planet’s populated surface governed by some form of imperial or colonial rule. The result of this potent technological, economic and political mix would be the so-called first global economy, already established by the 1870s.¹ In a famous passage in The Economic Consequences of the Peace, John Maynard Keynes lauded this “extraordinary episode in the economic progress of man”:

The inhabitant of London could order by telephone, sipping his morning tea in bed, the various products of the whole earth, in such quantity as he might see fit, and reasonably expect their early delivery upon his doorstep. … He could secure forthwith, if he wished it, cheap and comfortable means of transit to any country or climate without passport or other formality … and could then proceed abroad to foreign quarters, without knowledge of their religion, language, or customs, bearing coined wealth upon his person, and would consider himself greatly aggrieved and much surprised at the least interference. But, most important of all, he regarded this state of affairs as normal, certain, and permanent.²

The outbreak of war on July 28, 1914, revealed that the first global economy had been none of those things. “The projects and politics of militarism and imperialism, of racial and cultural rivalries, of monopolies, restrictions, and exclusion” were to play “the serpent to this paradise,” Keynes wrote. The Wall Street Crash of 1929 and the Great Depression of the 1930s finished off what was left of the first global economy. How paradisiacal it had all been much depended on whom you asked, I suppose. But in any case, an “extraordinary episode” it undeniably was.

After the Great Depression, World War II prevented economic globalization from reemerging. The process only began shifting gears again in the early 1950s, speeding up significantly after the 1970s following market deregulation and liberalization. The collapse of the Soviet Union reopened Russia and Eastern Europe to foreign investment, and by the turn of the century large emerging economies had firmly joined the global capitalist mainstream, most notably China. With these developments the world had entered a new phase of ever deepening and accelerating interconnectedness, the “second global economy.”³

But now a backlash against this seemingly unstoppable process is underway, almost exactly a century after World War I brought the first global economy to a violent halt. Among electorates in many Western democracies there is a growing sense that the payoffs of globalization have been too disappointing for too many, the burdens too unevenly distributed, the benefits too unequally shared. In the political and public discourse the question is increasingly being asked if globalization “works for everyone,” indeed if it is even capable of working for everyone. The merits of heightened trade barriers, curbed migration and strengthened borders or, in the case of the European Union, restored borders are discussed in a way that they have not been in decades. Nation-states have always exerted a strong influence over people’s sense of collective identity and economic self-determination. But that influence is growing again after a long period in which supranational organizations enjoyed broad and almost unquestioned support.

The sour mood gripping many societies at the moment seems largely to be expressed in a nihilistic desire to demolish the status quo, seen by large numbers of citizens as rigged against them. In Age of Anger, Pankaj Mishra has attempted to capture what is driving the current destructive temper. He notes that in the recent past, the shocks of modernity were “absorbed by inherited social structures of family and community, and the state’s welfare cushions,” whereas now “individuals are directly exposed to them in an age of accelerating competition on uneven playing fields, where it is easy to feel that there is no such thing as either society or state, and that there is only a war of all against all.” The results are not pretty: “An existential resentment of other people’s being, caused by an intense mix of envy and sense of humiliation and powerlessness … is presently making for a global turn to authoritarianism and toxic forms of chauvinism.”⁴

If nativist sentiments and zero-sum thinking are on the rise, industrialscale, interstate warfare of the kind that marred the twentieth century still seems only a remote possibility. All the same, the period of political turbulence we live in might see the end of the second global economy. Some of its pillars are being chipped away while others are being openly questioned, a slow erosion with an uncertain outcome. Developments are too young for predictions yet, and they are likely to remain so for some time to come. But one way or the other, changes are afoot in how the world economy is governed. The coming years will teach us to what degree national governments, under pressure from electorates demanding a retreat from globalization, will dismantle the interconnected economic world order constructed after World War II.

I note these things here because this book will be devoted to the role of states in the economic development of the ancient Mediterranean. State ideology, economic migration, commercial connectivity and social trust will be recurring themes, and just as I was writing about these matters the intensity of the current political discourse was reaching new heights. Of course as I deal with the ancient world, the states to be discussed were not nation-states with flags, passports and national anthems. Moreover, the ancient world was premodern in nature and did not have the benefits of mechanized transportation and lightning-fast communication that the Industrial Revolution would bring. The primary sector always remained the bedrock of the economy and overall output always remained low by modern standards.

Yet if the rise of the Greco-Roman world was not comparable to the first, let alone the second global economy, an “extraordinary episode in the economic progress of man” it was also. How much so is visible in long-term data suggesting that economic activity reached levels not seen again until the high Middle Ages or the start of the Early-Modern Period. The data also suggest that ancient economies in the aggregate followed a trend of growth and decline, an observation that is central to the discussion in this book.

The Mediterranean Economy in the Long Run

In the bar chart below (fig. 1.1), we see the number of known Mediterranean shipwrecks dated 1500 BCE–1500 CE, broken down by half century.⁵ Not every ship type is equally visible. Wooden hulls have mostly disintegrated, leaving behind only nonperishable material such as tiles, bricks, blocks of stone and marble, but most of all ceramic containers. We are therefore seeing predominantly transport vessels, not warships.⁶

This chart is not as straightforward to read as it might appear, as both Andrew Wilson and Michael McCormick have emphasized.⁷ For one thing, ships that carried perishable cargoes such as grain, timber and textiles have all but vanished from the seafloor. Secondly, in Roman times wooden barrels began to be used alongside the traditional ceramic amphorae, making a larger percentage of shipwrecks invisible archaeologically. Wilson argues that this change in large part accounts for the remarkable drop in the bars from the first to the second century CE. But incomplete data may also explain that effect, perhaps more so than a shift from amphorae to barrels. Recent work by McCormick on new shipwreck discoveries shows that the decline from the second to the eighth century CE was much more gradual than figure 1.1 would suggest.⁸

What matters for my purposes is the trend line ca. 700 BCE–ca. 700 CE, the shape of which is increasingly being confirmed by new data. Much scholarly work has focused on the Roman-era high point. In earlier representations of the dataset it appeared in the late-republican period, but it has now shifted to the time of the early empire with Wilson’s incorporation of new finds and his more sophisticated data-processing methodology. To explain the imperial-era peak, Philip Kay in Rome’s Economic Revolution concentrated on the phase just preceding it. He notes that the late second and early first centuries BCE show a significant rise, a phenomenon he attributes to the development of finance during the later Roman republic.⁹ But he largely limits his discussion to that timeframe. He acknowledges that “Rome acquired control over developed trade and communication networks throughout the Mediterranean” and that “some of the economic developments which we see in the second century represent the continuation of processes that had begun earlier, under the influence of contact with the Hellenistic world.”¹⁰ But he does not assign those older processes any particular significance in his explanatory model. Instead, he points to Roman-era monetary flows as the main driver of economic change:

During the second century [BCE], increased inflows of bullion combined with the contemporaneous expansion of the availability of credit to produce a large increase in monetary liquidity. This in turn resulted in a major upward inflection in Roman economic activity and the creation both of a more complex system of production and distribution and of an enormous material culture that was to reach its height under the Principate.¹¹

These multiplier effects could occur because large sections of the Mediterranean economy were still operating without coinage when the Roman increase in monetary liquidity began. In Kay’s estimate, Italy’s monetization level rose from 39 to 68 percent between 150 and 50 BCE, and levels are likely to have risen also elsewhere. Rather than produce inflation, as standard economic theory would predict, the expansion of the money supply therefore fueled economic activity. It increased the size and degree of commercialization of the nonagrarian sector and stimulated overseas trade.¹² Kay’s observations and explanations coincide with those of David Hollander. Several years earlier in Money in the Late Roman Republic, Hollander had already concluded that the “expansion of Roman banking and the creation of business networks throughout the Mediterranean in the late Republic were the two primary developments allowing for the growth of trade.”¹³

Kay’s and Hollander’s analyses seem compelling to me. However, they do not explain what in my view is the most remarkable aspect of the shipwreck graph, namely the steady climb starting around 700 BCE, to which Rome initially contributed nothing. The Romans were not a maritime people in the Iron Age, and when for military reasons they finally began plying the seas late in the fourth century BCE, this was only a limited phenomenon. In the subsequent century they were still entirely absorbed by warfare on the Italian peninsula and Sicily.¹⁴ Roman long-distance trade supported by advanced financial institutions would become a significant factor in the Mediterranean from the mid-second century BCE onward. But the smooth climb of the trend line centuries beforehand suggests that what was happening then continued a process initiated earlier.

The shipwreck graph is not the only image telling us to extend our gaze backward beyond the time of the Roman republic. Another image doing so is presented by figure 1.2. It shows the data on levels of Pb (lead) pollution obtained by coring the Greenland ice sheet.¹⁵ The mining of the metal itself released substantial amounts into the atmosphere, but the Pb pollution we see in figure 1.2 was caused mainly by the extraction of silver. Airborne particles were carried to Greenland by a strong south-to-north atmospheric transport, were captured by snowfall and then trapped in the compacted strata of ice. Research on the stable isotope signatures of the lead in the ice cores has shown a heavy contribution from silver mines in the Spanish Rio Tinto region.¹⁶ Future analyses promise to give us a much more detailed picture of the geographical source distribution, especially as the study of anthropogenic lead pollution in lake sediments and peat bogs in Spain, Sweden, Britain and Switzerland is also intensifying.¹⁷

The data on atmospheric Pb pollution have a clear significance for ancient economic history, especially in their relation to silver mining. From early on in Mediterranean history, silver was extracted to manufacture decorative ornaments and to produce bullion. Moreover, from the mid-sixth century BCE onward it was used to mint coins.¹⁸ Pollution levels can thus be taken as a proxy variable for economic activity.¹⁹ However, as is the case with the shipwreck graph, the picture presented by the Pb pollution data is not free of interpretation problems, as Hannah Friedman cautions.²⁰ A shift from silver to gold in minting new coins might, for instance, have produced a drop in the trend line not necessarily indicative of a decline in economic activity. Friedman also notes that because of the prevailing winds in the troposphere, the sample of Pb pollution in the Greenland ice sheet may not be representative of all silver mining occurring in the ancient world. Some production areas such as the Rio Tinto may be overrepresented, while others such as the Kosmaj (Serbia) may be underrepresented. But problems of sample collection and interpretation notwithstanding, the value of the data as a general gauge of long-term economic development is not in doubt.

Figure 1.2 shows some notable differences from figure 1.1. The Pb diagram follows a jagged trajectory, the result mainly of a paucity of data points.²¹ But most conspicuous is the sustained rise after the late-antique or early-medieval low point, a rise not seen in the shipwreck graph. The latter effect should be attributed to the decreased visibility of wrecks due to a shift from amphorae to wooden barrels, which was all but complete in the Middle Ages.²² But despite these differences, two entirely different datasets display comparable trend lines and seem to tell a broadly similar story ca. 700 BCE–ca. 700 CE. If the data reflect general, long-term economic developments, as they seem to, then the question of what those might have been is highly pertinent.

One long-term process largely agreeing with the image in both graphs is demographic growth and decline. The end of the Bronze Age saw a massive population contraction in Greece, but from the late second millennium BCE onward there was steady, aggregate demographic growth around the Mediterranean. From the available data, Walter Scheidel concluded that with all due caution,

we may assume that between the twelfth century BC and the second century AD, the population of the part of Europe that was eventually taken over by the Roman empire approximately quadrupled in size, at a long-term average annual growth rate of around 0.1 percent. … After the depression of population numbers following the disintegration of the western and much of the eastern Roman empire in the fifth and sixth centuries, the formerly Roman part of Europe (with the exception of Greece) generally re-attained peak Roman population levels by the twelfth or thirteenth centuries, and after another slump caused by the Black Death consistently exceeded them from the mid-fifteenth century onwards.²³

An increase in population size generated rising levels of consumption and production and thus at least aggregate economic growth. Whether certain areas and periods also experienced per capita growth is an open question, although a large and growing number of scholars, including Scheidel, argue that the answer should be yes.²⁴ But one way or the other, demography only partially explains long-term economic trends. As Willem Jongman has pointed out, shipping and metal extraction “had obviously increased by much more than could be expected from just population growth. Similarly, decline was much steeper than could be expected from just demographic contraction.”²⁵

State Formation and the Mediterranean Economy

Another phenomenon was also occurring: the trend lines in both graphs broadly track the formation, growth and disintegration of Mediterranean states. The period between ca. 700 BCE and ca. 50 CE witnessed a process of pronounced state consolidation. After the collapse of the Aegean Bronze-Age societies around 1200 BCE, city-states formed during the Iron Age in the eastern and central Mediterranean.²⁶ During the late sixth and early fifth centuries BCE, the Achaemenid empire reached its maximum extent, tying together the East Mediterranean from Egypt to Thrace.²⁷ Its retreat and conquest by Alexander in the second half of the fourth century BCE resulted in the emergence of three large kingdoms. Those would in turn all fall to Rome, which had meanwhile steadily been consolidating its hold on the Italian peninsula.²⁸ In the west as well, a city-state, Carthage, was on the rise. Between the sixth and third centuries BCE it would bring Sardinia, Sicily, Corsica and large parts of North Africa and Spain under its control, in effect building an overseas empire.²⁹ Just as the Hellenistic kingdoms, it would ultimately be defeated and taken over by Rome.

This process of ongoing territorial unification in the west, center and east would culminate in the emergence of the largest state the ancient Mediterranean would ever know: the Roman empire. After its disintegration, state formation occurred again in the early Middle Ages, a process that would grow stronger especially after ca. 1000 CE, and that continued into the Early-Modern Period with the birth of the European nation-state.³⁰

A knotty problem needs to be addressed at this point in the discussion. There are “as many definitions of the state as there are social theorists.”³¹ The Roman empire and the Hellenistic kingdoms may be clear-cut cases, but especially for earlier societies the question of what counts as a state is not an easy one to answer. Chris Wickham was confronted with the same problem in his study of the early Middle Ages ca. 400 CE to ca. 800 CE with its many emerging states and statelets. As a solution he offered a set of five parameters to conceptualize an “ideal type” of the state:

the centralization of legitimate enforceable authority (justice and the army); the specialization of governmental roles, with an official hierarchy which outlasted the people who held official position at any one time; the concept of a public power, that is, of a ruling system ideologically separable from the ruled population and from the individual rulers themselves; independent and stable resources for rulers; and a class-based system of surplus extraction and stratification.³²

No definition of the state will satisfy everyone, and Wickham’s as well has its drawbacks. For one thing, it does not include control over a specific territory, an aspect all but universally associated with states. Problematic for my purposes is that one can question if cities such as Tyre, Sidon and Byblos ca. 900 BCE fall under the definition, as at that time their monarchies may still have been more personalized and less institutionalized than Wickham’s ideal type would call for. It is also debatable how well his definition covers Greek city-states, especially if they had a democratic form of government such as classical Athens. Of course that is a broader problem, as the statehood of Greek city-states under any definition is a topic of debate.³³

However, Wickham’s “ideal type” does not represent a checklist of prerequisites. Even if in a particular society one or more of the five parameters were present to only a lesser degree, it might still qualify as a state. Admittedly, some purity is sacrificed in adopting such a flexible approach. But for my purposes, ignoring definitional imperfections for the sake of argument brings major benefits. Being able to put the Roman empire, classical Athens and eighth-century BCE Tyre on the same plane allows me to analyze how their establishment of public authority affected economic conditions. Obviously those three societies varied widely in their governmental structure. But by calling them all states, I do not mean to argue that they were comparable that way. What I argue instead is that their emergence as public entities influenced the economic landscape beyond the capabilities of private individuals or groups of private individuals.

None of this is to say that private individuals could not profoundly alter the economic landscape. They could and did, as can be seen clearly in their role in the early Iron Age, when long-distance cultural and mercantile links were reestablished following the collapse of Mediterranean Bronze Age societies. That process of civilizational renewal was initially driven by individual traders unconnected to any emerging public institutions. As Tamar Hodos noted, it “is particularly evidence for trade, in its most general sense, that characterizes the beginning of a new impetus in the Mediterranean in the early first millennium, initially conducted by individuals working in an independent rather than state capacity.”³⁴

But figures 1.1 and 1.2 suggest that from ca. 700 BCE onward, state formation began to have a positive influence on the economic activities of those pioneering individuals. A practical reason why that might have been the case readily comes to mind: states provided a transportation infrastructure, aiding traders in their mercantile endeavors. The construction of maritime harbors formed an important part of that process, which is relevant especially for understanding figure 1.1. Harbors started to appear in the Aegean in the eighth century BCE; witness for instance the 100-meter-long mole dated to that time on the island of Delos. Literary evidence tells us that seaports began to be built on the initiative of powerful Greek monarchs such as Polycrates of Samos (Hdt. 3.60).³⁵ In the context of developments in Archaic Greece, the evidence shown in figure 2.2 is relevant as well. The sixth-century BCE introduction into the Greek world of coinage on an official weight standard stimulated the building process by making it easier to finance major public works.³⁶ In later periods as well, rulers decided on the creation of Mediterranean harbors, including some of the largest and most famous ones, such as Alex-andria’s.³⁷ In the Roman empire, funding harbor construction remained “something like an imperial—or at least public—privilege,” falling to either emperors or cities.³⁸

Port infrastructure was complicated to build, requiring specialized engineering skills that had to accumulate over time. The process of learning by doing that started in the Iron Age continued until, in the words of Lionel Casson, “the essential elements of a harbor had been worked out by the fifth century B.C. The successive centuries saw chiefly elaboration of facilities and increase in size.” But he adds that later developments still included some major advances:

The Hellenistic Age brought to harbor construction the vastness of size and the layout according to an integrated plan that characterized the architecture of the times. In addition, it contributed a feature of the highest practical importance, the lighthouse. … The Romans introduced a significant innovation, the use of concrete that would set under water. This powerful and flexible material enabled them to strike out boldly and plant harbors where nature had nothing at all to offer.³⁹

Apart from facilitating overseas shipping, the construction of harbors had a stimulating economic effect by integrating public and private monetary flows. Especially if public works went hand in hand with rising levels of monetization, multiplier effects of the kind identified by Kay will have followed. Such effects likely were at work well before the time of the later Roman republic. They are almost certain to have been felt, for instance, in Ptolemaic Egypt, where taxation levels were high and where in the course of the third century BCE the economy went from being largely unmonetized to largely monetized.⁴⁰

Because of its geographical size the Roman empire is especially significant for the effect of tax spending on economic integration. The impact of one on the other has been studied by Keith Hopkins in what is now a classic article.⁴¹ Hopkins’ fundamental assumption in his “taxes and trade” model was that a number of rich areas—including Spain, southern Gaul, North Africa, Asia Minor, Syria and Egypt—were revenue exporters, paying more in imperial taxes than they received in public spending. The central government partly invested the collected money at its source, but sent most of it out it to pay for the legions on the frontiers in provinces that were tax importers. The amount of revenue going to the defensive armies was considerable. As the largest item in the state budget, the military probably accounted for over 50 percent of public spending.⁴² But for this process of monetary redirection to continue, the net-contributing provinces had to find a way to fill their deficits. Without a counterbalancing revenue stream, they would not have had the money to pay their fiscal dues the following year. Engaging in trade was the answer, according to Hopkins. By producing and exporting agricultural surplus and manufactured goods, tax-exporting provinces earned back cash, thereby balancing out their losses.

In a follow-up paper Hopkins incorporated rent-taking by Roman aristocratic landowners into his model. Members of the Roman senatorial elite possessed large tracts of provincial land, from which collectively they received rent income on a par with the state’s annual revenue net of army costs. This rent sustained the elite’s ostentatious lifestyle, and like the tax revenue received by the state, it was spent at some distance from its place of origin, chiefly in Rome but also in local and regional capitals around the empire. Just like taxes, rent contributions were predominantly paid in money, not in kind. To allow provincial laborers to pay their dues in subsequent years, these monetary flows therefore needed reverse ones, and here as well the solution was to engage in surplus production and exports to earn back cash.

Taxation and rent payments counterbalanced by trade thus produced what Hopkins proposed was a “mildly developmental” economic effect, creating a “thin veneer of monetary and economic integration” over an essentially agrarian economy.⁴³ This model has not been universally accepted but, as Scheidel observed, “the debate is primarily concerned with the question of whether the central state and its associated élites played an absolutely crucial or merely a very significant rôle in commercial development: the overall importance of state formation is not in doubt.”⁴⁴

Enforcement and the State

The topic of the economic role of states brings me to the main theme of this book. Mention of Douglass North, the father of New Institutional Economics, is inescapable here. North’s work has had a profound impact on the field of ancient economic history, and I think it is fair to say that Neo-Institutionalism has now established itself as the default paradigm in both Greek and Roman scholarship.⁴⁵ Examples abound. Elio Lo Cascio argued that “the theoretical framework proposed by North can allow a better insight into the performance of the Roman Empire as a unified political organization.”⁴⁶ Joseph Manning, in a monograph on the Ptolemaic economy, wrote: “understanding how local economies were linked to the central state requires a model of the state. I adopt in this book North’s neo-classical theory of the state.”⁴⁷ In a recent work on classical Greek history, Josiah Ober specified that Neo-Institutionalism’s “insistence that institutions … and organizations (including, but not only, states), along with markets and networks, are fundamental determinants of economic change, grounds the arguments of this book.”⁴⁸

But difficulties arise with this wholesale adoption of Neo-Institutionalism in ancient economic history, specifically with regard to its concept of the state. In Structure and Change in Economic History, North defined a state as “an organization with a comparative advantage in violence, extending over a geographic area whose boundaries are determined by its power to tax constituents,” continuing that “an organization which has a comparative advantage in violence is in the position to specify and enforce property rights.”⁴⁹ The latter part of that definition is problematic when applied to Greco-Roman history, and not just there.⁵⁰ First of all, in nonstate-level societies, including chiefdoms and tribes and arguably even bands, basic property rights were also specified.⁵¹ Their complexity could increase along with growing sociopolitical complexity, but the state was not a conditio sine qua non for their creation. More importantly, if states could in theory enforce private property rights, they did not necessarily do so in practice.

In the abstract, the idea of casting the state in the role of third-party enforcer is an attractive one. Already in 1651 Thomas Hobbes explained the logic of that notion in the Leviathan:

If a covenant be made, wherein neither of the parties perform presently, but trust one another; in the condition of mere nature … upon any reasonable suspicion, it is void: but if there be a common power set over them both, with right and force sufficient to compel performance, it is not void. For he that performeth first, has no assurance the other will perform after; because the bonds of words are too weak to bridle men’s ambition, avarice, anger, and other passions, without the fear of some coercive power. (14.18)

But this line of reasoning has a serious deficiency, as Peter Leeson points out in a deliberately provocative book, Anarchy Unbound: “Hobbes overlooked the possibility of self-governance: privately created social rules and institutions of their enforcement.”⁵² Leeson offers theoretically grounded historical case studies showing that transactions with delayed performance under “anarchy,” loosely defined as the absence of formal government, are perfectly possible. Moreover, he posits that if a state is sufficiently rapacious, its dissolution and replacement by private order can be positive for economic performance. In support of that claim he cites the case of Somalia, arguing that the complete disappearance of government there in 1991 had by the early 2000s produced a marked improvement in the country’s economy.⁵³ These studies lead him flatly to declare that Hobbes had it all wrong.

In so doing he is giving the intellectual achievement of the Leviathan too short shrift, but he is correct in treating its disregard of nonlegal enforcement as a significant gap. Even in well-functioning states governed by the rule of law, much economic interaction still depends on private order, including bargaining “in the shadow of the law.”⁵⁴ That socioeconomic reality should come as no surprise. As Eric Posner has argued, courts have trouble determining contractual obligations and understanding business relationships because in both repeat and one-shot deals the number of unpredictable contingencies is large. Courts, therefore, “are not good at deterring opportunistic behavior in contractual relationships, but parties are. This is why so much contractual behavior depends on reputation, ethnic and family connections, and other elements of nonlegal regulation, and not on detailed and carefully written contracts enforced by disinterested courts.”⁵⁵

Still, the basic notion, as formulated by Hobbes, that a central task of the state is to act as society’s ultimate enforcer is appealing. The idea was introduced into modern economic thinking well before North. An expression of it can be found in the work of Keynes, who wrote in A Treatise on Money that

it is a peculiar characteristic of money contracts that it is the State or Community not only which enforces delivery, but also which decides what it is that must be delivered as a lawful or customary discharge of a contract. … The State, therefore, comes in first of all as the authority of law which enforces the payment of the thing which corresponds to the name or description in the contract.⁵⁶

It should be noted here that Keynes makes a critical distinction between third-party enforcement and third-party adjudication. But contrary to what he seems to imply, states do not need to involve themselves in both tasks. They can assume some variation of adjudication without necessarily assuming enforcement, operating along a continuum from least forceful to most forceful. That continuum is cogently captured in a framework of “dispute settlement mechanisms” developed by Beth Yarbrough and Robert Yarbrough.⁵⁷ Though designed to study modern international trade relations, the framework serves my present purposes equally well. It consists of four main categories, conceptualized as fluid rather than fixed and exclusionary:

1. Third-party information gathering on alleged violations, and dissemination of that information, followed by group retaliation, potentially ending in ostracism.

2. Nonbinding third-party adjudication resulting in a recommended remedy, potentially followed by multilateral retaliation; no third-party enforcement.

3. Binding third-party adjudication without third-party enforcement; retaliation by the aggrieved party, permitted only if a violation has been confirmed.

4. Binding third-party adjudication and enforcement; no private retaliation allowed.

The potential benefits of the state assuming the task of both arbiter and enforcer in a reliable, unbiased way are obvious. Yet public adjudication without public enforcement can still be beneficial to economic performance, depending on the state’s level of impartiality, the quality of its arbitration and the sophistication of its legal framework. Conversely, if the state’s legal institutions are poor, adjudication with full enforcement can be harmful to economic development. A wide range of possible outcomes is imaginable, and historical ones depend on the specific circumstances of individual societies. But in all cases, what matters is that for a proper understanding of the state as an authority of law, distinguishing between adjudication and enforcement is imperative.

Getting the two concepts confused is all too easy. It is tempting to take evidence for disputants submitting to the jurisdiction of official courts of law as confirmation of the existence of third-party enforcement. That assumption seems to have informed the work of, for instance, Sheilagh Ogilvie. In Institutions and European Trade, she sets out to portray medieval merchant guilds as nothing but rent-seeking monopolists interested only in creating barriers to entry and bending the rules of trade to their advantage. The argument that such associations might have been critical for contract enforcement is incorrect, she contends. Traders could turn to the state to perform that task, an option they often preferred over internal dispute settlement within their own organizations. In a typical expression of that view, she asserts that “Wherever European merchants traded during the medieval Commercial Revolution … state enforcement was available and long-distance merchants voluntarily used it to enforce contracts.”⁵⁸

However, what she is referring to in making such claims is not state enforcement but state adjudication. To explain medieval traders’ preferences, she repeatedly emphasizes the quality and impartiality of official courts of law. But no matter how informative such alleged legal superiority may be about public institutions of adjudication, it tells us little about enforcement. A verdict by a medieval judge still had to be effectuated. The private-order mechanisms necessary to do so operated within and between medieval trading organizations, which by implication seem to have had the central role in contract enforcement that Ogilvie categorically denies they had.

Avner Greif, by contrast, does take the view that medieval trading coalitions were key to private-order enforcement, providing an essential service to the functioning of markets.⁵⁹ Much in his analysis is specific to medieval Europe, but his observation that Ogilvie wrongly equates courts with third-party enforcement has general validity: “The premise here is that all historical courts were the same and similar to a modern, effective court. … The issue is not whether … courts existed …, but whether they differed substantially from each other and from their modern equivalents.”⁶⁰ In other words, if we see public institutions of adjudication operating in any given historical society, we still need to establish what mechanisms of enforcement underpinned them.

In the study of Roman law, a position close to Ogilvie’s is taken by what David Ratzan has called the “legal centralists.” Ratzan rightly dismisses their denial of the importance of nonlegal sanctions as “historically incorrect and theoretically confused.” But he equally thinks that their counterparts, whom he labels “legal skeptics,” are at a loss to explain, for instance, official petitions to Roman state authorities: “If the government did not care, why bother complaining to it? … At the very least, we must imagine … Roman subjects as buying something of value in their contracts, petitions, and trials. What was it?”⁶¹ That question might mutatis mutandis be asked of Ptolemaic subjects, classical Athenian subjects, or for that matter the subjects of all ancient states.

But is the problem as intractable as Ratzan presents it as being? I suggest that the solution is in fact quite simple, hiding as it were in plain sight. What a plaintiff bought, or hoped to buy, with an investment in legal action was an official recognition of his position. Such an imprimatur by an authoritative public body enhanced the chances that collective action would be taken against his adversary. The latter’s reputation would be diminished, as would his social room to maneuver, especially in the long run if he became the object of repeated official censure. Scholars of ancient law tend not to think in such terms, and when they do they tend to see reputational mechanisms and collective action as alternatives to official legal procedures. But private order and public institutions were not working in opposition as competitors in a “market for enforcement.” On the contrary, they worked together. The more seamlessly they did so, the more effective enforcement would be.

The outcome of public institutions reinforcing private order does not equate to third-party enforcement in the sense of North, Keynes or Yarbrough and Yarbrough. North fully acknowledged that such enforcement has historically been rare. It became more widespread only after the Industrial Revolution and even then remained limited to a small number of societies. That historical argument is elaborated in his Violence and Social Orders, written with John Wallis and Barry Weingast.

The authors distinguish between two social orders based on how societies limit and control violence: open-access and closed-access societies. In the first, physical force is consolidated into official military and police organizations that are subordinate to the political system. Citizen identity in the open-access order is defined as a set of impersonal characteristics and rights possessed by all citizens. This shared set of impersonal rights at the heart of civil society prevents the executive from abusing its power. All open-access societies conform to Max Weber’s definition of a state as “a human community that (successfully) claims the monopoly of the legitimate use of physical force within a given territory.”⁶² Only a handful of states in history have come close to the Weberian ideal type. Today only about twenty-five countries containing about 15 percent of the world’s population fall into that category.⁶³

By contrast, the much more prevalent second type of social order, the closed-access society or “natural state,”

reduces the problem of endemic violence through the formation of a dominant coalition whose members possess special privileges. The logic of the natural state follows from how it solves the problem of violence. Elites—members of the dominant coalition—agree to respect each other’s privileges, including property rights and access to resources and activities. By limiting access to these privileges to members of the dominant coalition, elites create credible incentives to cooperate rather than fight among themselves.⁶⁴

Natural states were the only ones in existence before the early nineteenth century and still form the majority today. They do not achieve a monopoly on violence, nor do they need to for their survival. Their assertion of physical force is a continuous rather than a discrete variable. The more force they manage to appropriate, the more effective they will be as governing organizations. Relationships within and between the elite coalitions that rule them are strongly personal, and status and hierarchy tend to be defined by individuals’ social personas. Yet public institutions can grow in complexity, allowing them to become more impersonal and less dependent on the individuals controlling them at any given moment. Through that process an enhanced stability and durability of the state and its ruling coalition can be achieved. But even if increasingly stable and durable social arrangements are to the advantage of governing elites, nothing about the push toward complexity is teleological in nature. Natural states are as capable of regression as progression in this regard, depending on ever-changing external circumstances and internal power struggles.

Those ideas are thought-provoking and contain much of interest to the social theory–minded historian. Nevertheless, squeezing all recorded human history into a binary framework of open-versus closed-access societies is reductive in the extreme. The dichotomy inevitably creates friction when applied to some historical societies. In addition, the framework’s simplicity limits its usefulness for a student of the premodern world, who is left with only the concept of the natural state. North, Wallis and Weingast acknowledge that natural states differ widely, a problem they address by distinguishing between “fragile,” “basic” and “mature” ones. However, those three manifestations differ primarily in “the sophistication of the organizations they can support.”⁶⁵ Such sophistication is a sliding scale, and the classificatory distinctions are more idiomatic than substantive analytical tools. But the shortcomings of Violence and Social Orders notwithstanding, it offers useful insights into premodern states’ use of force, interference in trade and stability of governance. I will be referring to it repeatedly in this book, and will engage with it more fully in chapter 2.

From early on in North’s work, his thinking about what makes states successful was predicated on the economic development of medieval Europe and the subsequent rise of the Western world.⁶⁶ The period of change culminating in the Industrial Revolution and the “Great Enrichment,” to use Deirdre McCloskey’s evocative phrase, has been extensively studied by economic historians.⁶⁷ But identifying the “rule of law” as a critical factor in that change may be erroneous. Seeing the economic expansion of industrializing Britain as the result of well-defined property rights enforced by the state is in any case a “gross oversimplification.”⁶⁸ Though at the epicenter of the Industrial Revolution, British society in the 1700s and early 1800s still relied on private-order institutions, including reputation mechanisms and social ostracism of serial cheaters. Rather than economic development resulting from third-party enforcement, it would seem that causation worked the other way around. British society was transformed by the Industrial Revolution. As its economy became more urbanized and anonymous, reputation mechanisms became untenable and enforcement was consequently transferred to the public sector.

Janet Landa has proposed an abstract microeconomic theory that is much in line with these historical events. She enumerates the options open to middlemen to cope with “contract uncertainty,” the term she employs for a lack of third-party enforcement. All options bear costs, which middlemen seek to reduce by moving from individual to collective action. As trade increases in volume and middlemen increase in number, unanimous collective action becomes ever harder to achieve and ever less cost effective. Landa posits that traders’ cost-optimizing calculus thus leads to the emergence of the “protective state,” a phrase borrowed from the work of James Buchanan and meaning “that part of government which acts as the enforcing institution of society.”⁶⁹ She concludes that the resulting social order ensures a more Pareto-efficient equilibrium.

The existence of social order (1) reduces unnecessary transaction costs arising from a trader’s breach of contract …; (2) facilitates the impersonal process of exchange by encouraging the trader to trade with outsiders, thus ensuring all opportunities for trading are exhausted; and (3) gives the trader the opportunity to shift some of the resources tied up in the protection of contracts into trade or capital accumulation.⁷⁰

Landa does not invoke historical examples, but does not seem to see this shift toward a Pareto-superior social order as contingent on industrialization. Instead she presents her ideas as a general theory on the emergence of an “exchange economy with legally binding contract.” Yet if the emergence of the “protective state” is such a natural outcome of economic actors’ cost-optimizing behavior as she makes it seem, the question might well be asked why it has materialized so rarely.

To return to the ancient states that will be the subject of this book, they were non-Weberian in nature, not claiming a monopoly on violence. As powerful organizations they did have a “comparative advantage in violence,” but they did not employ it to “enforce agreements such that the offending party always had to compensate the injured party to a degree that made it costly to violate the contract.” Of course, expecting them to do so would be setting the bar too high. As already noted, North admitted that the conditions for such state behavior “are seldom, if ever, met in the real world,” including the postindustrial one.⁷¹ We should, thus, not measure the states of the ancient world against modern nation-states with fully developed institutions of public enforcement. But effective enforcement is indispensable for commercial exchange, and the question of how and to what extent it involved the state is relevant for our understanding of ancient economic history. An exhaustive treatment of all ancient societies is a near impossibility, and even a partial discussion would require a separate book. But a brief survey of some selected examples will suffice to show the limits of state enforcement.

In Greek city-states some degree of public physical force in defense of justice was available through city representatives, and such force was well understood to be necessary. Aristotle, in the Politics (6.1321b40–22a7), showed awareness that law enforcement occasionally required publicly administered violence. But in the small, face-to-face societies of Greek cities, laws were respected out of civic obedience and concern for social order as much as out of fear of sanctions, as Mogens Hansen has pointed out.⁷²

Greek city-states did not maintain a body of bailiffs or civil enforcement officers, even if Athens had a public-order apparatus of market overseers, jailors and Scythian archers.⁷³ The effect of that lack of public support on commercial disputes becomes visible especially when we scale up the canvas. What of enforcement between rather than within communities? To resolve intercommunity disputes over business affairs, Greek city-states maintained a system of seizure of goods by private individuals. A citizen of one city could request court approval to seize goods from any of the community members of his adversary to satisfy his claim. City-states could choose to supersede this principle of collective responsibility by agreeing on rules laid down in interstate treaties. But ultimately, this system was a form of private justice, albeit subject to customary law and court oversight.⁷⁴ Such a system was not unique to Greek city-states or for that matter antiquity. Greif has shown that a similar “community responsibility system” governed relations between medieval trading centers in Italy, Germany and England.⁷⁵

With the formation of the Hellenistic kingdoms in the third century BCE, large states commanding impressive powers of physical force entered the scene. In the relatively well documented Ptolemaic kingdom, a new legal order was established that combined central state control with the autonomy of local communities. Papyrological evidence shows that legal proceedings could be consequential in land disputes, as we learn from the records of a second-century BCE family conflict over the inheritance of an estate. Manning, in his analysis of the case, concludes that the “authority of the Egyptian court had enormous influence upon the local community. … The complaints addressed in petitions to local and regional Ptolemaic officials were resolved, in theory, through the authority of a local Egyptian court.”⁷⁶ Nevertheless, whatever respect the word of the court may have enjoyed in the adjudication of community disputes over real estate, for a full assessment of the Ptolemaic legal system we need to consider how the state employed its coercion machinery.

In Law and Enforcement in Ptolemaic Egypt, John Bauschatz has recently argued that the Ptolemaic kingdom had an extensive organization of police officials charged with maintaining public order at the village level.⁷⁷ According to his study, this coercion apparatus targeted not only crime but also debt. Yet the debt for which one could land in prison was predominantly public in nature, in particular tax arrears. Bauschatz argues that detention for private debt also occurred, but the evidence he cites is extremely meager and ambiguous.⁷⁸ Even if it did occur occasionally, there was nothing systematic or institutionalized about it. Ptolemaic creditors could not count on public officials to enforce their claims.

Under the formulary procedure of Roman law, considered to be Rome’s classical legal system, the execution of court verdicts was a form of state-sanctioned self-help. In the republican era and well into the empire, a successful plaintiff had the right to seize the person of the debtor and take him into custody, forcing him to work off his debt.⁷⁹ In later times he was also allowed to confiscate and auction off his debtor’s entire property in a process resembling bankruptcy proceedings. The rules were relaxed in the course of the imperial period to allow for the seizure of individual assets, making for a more efficient way to satisfy debts.⁸⁰ But state-provided force did not form part of the system. “Judgment given, the duty of the judge was over. If the convicted party did not obey the judgment it was up to the plaintiff to take further steps. He was not given physical help by the authorities.”⁸¹

In the course of the imperial period, various innovations of civil procedure were introduced, eventually to result in a comprehensive new system in the mid-fourth century CE.⁸² It is clear that this development accompanied a pronounced hierarchization of the Roman legal process, characterized by a strict division between an upper and lower social class.⁸³ Some measure of public enforcement also emerged as part of the reform. A handful of legal texts mention bailiffs charged by special fiat with the execution of certain verdicts.⁸⁴ Much about these men remains unknown, including how numerous they were and what verdicts they might have been commissioned to enforce. But we do know that, as other court functionaries, they charged fees for their services, a practice inviting corruption.⁸⁵ A mandate of Justinian (Nov. 96 pr., 539 CE) aimed to stop them from colluding with plaintiffs in initiating spurious lawsuits and splitting the gains, apparently a widespread practice.

What all this amounts to is that ancient states fell somewhere in category 2 or 3 in the adjudication and enforcement framework of Yarbrough and Yarbrough, cited above.⁸⁶ Arguably, the Roman empire late in its history took some steps in the direction of category 4. But those steps were imperfect and incomplete and in part even counterproductive. None of this is to say that legal institutions did not matter for the economic performance of the ancient world. A large part of this book will in fact be devoted to arguing that they did matter. But I reiterate that in evaluating their impact, we should be careful to distinguish between public adjudication and public enforcement.

The Argument of This Book

The two central tenets of this book have now been made clear. First, state formation and consolidation had an aggregate positive effect on the economy of the ancient Mediterranean, starting in the Late Iron Age and peaking sometime in the Roman imperial period. Second, we should not ascribe that effect to ancient states acting as third-party enforcers of private property rights. Two questions follow. First, what did states’ positive influence consist of? Second, what was the state’s role subsequently, when economic decline set in?

The idea that state behavior can have both positive and negative effects on economic performance is at the heart of North’s historical analysis. In The Rise of the Western World, he juxtaposed successful states such as the Dutch Republic and England in the 1500s and 1600s with less successful ones such as contemporary France and Spain. The latter, he proposed, “failed to create a set of property rights that promoted economic efficiency.”⁸⁷ For the reasons discussed above, I will not follow North’s ideas about the role of states in defining and enforcing property rights. However, that states’ actions can be both helpful and harmful is a key issue to address.

As for harmful action, one might suppose warfare to have been the predominant type, because of its destructive potential and because Greco-Roman history had no shortage of it. Yet armed conflict seems not to have been the disruptive force one might have expected.⁸⁸ Figures 1.1 and 1.2 suggest that the ancient Mediterranean experienced at least aggregate growth during a time when interstate warfare was frequent. Of course, not all military action is detrimental to economic performance. But even if things turned severely negative sum, as doubtless happened in several major conflicts, it seems that in the long run the effect was offset and more by positive factors. That observation brings me to one of the main arguments of this book: ancient states were beneficial to growth because they created institutions conducive to economic development, even if that was not necessarily the intention.

The last point deserves emphasis. In this book I am nowhere arguing for a “public economic policy” designed and executed by designated state representatives. By now it is a truism that no such policy existed in antiquity, an argument made already by Moses Finley.⁸⁹ But if the basic idea remains generally accepted, modifications to it have begun to emerge in the scholarly literature. Darel Engen, for instance, in a book on Athenian commerce, suggested that in “granting honors and privileges for trade-related services, Athens adapted and manipulated traditional institutions to formulate a practice that was flexible enough to acknowledge and exploit the dual desires for honor and profit that existed in the Greek economy, thereby fulfilling its trade policy.”⁹⁰ Those are interesting new ideas worth exploring (see chapter 2).

By accounting for a desire for honors, Engen goes a long way toward incorporating ideology into his analysis. I propose that a much greater emphasis on the force of ideology is warranted for our understanding of ancient economic development. Rulers had to legitimize their power, an essential dynamic of natural states that North, Wallis and Weingast acknowledge but spend little time discussing.⁹¹ Carles Boix has recently argued (contra Weber) that in what he calls “monarchical regimes,” any ideological beliefs about the ruler’s legitimacy cannot be disentangled from subjects’ fear of the ruler’s coercive capabilities and their instrumental calculations about the benefits of social order.⁹² But regardless of whether one sees rule-legitimizing ideology as an independent or an ancillary source of support for monarchical regimes, its political reality had economic effects. In the Hellenistic kingdoms it led state representatives and the social classes wishing to emulate them to engage in behavior with expansionary effects, such as conspicuous consumption. The same held true for the Roman empire, even if unlike the Hellenistic kingdoms it was not in direct competition with peer polities, which might have reduced the need for rule-legitimizing consumptive patterns.

Ideology had not just macroeconomic but also microeconomic effects, and to study those we can build on North’s extensive work on social norms. North has repeatedly emphasized the significance of normative rules that structure society by providing informal constraints on individual behavior. Collectively those constraints can be considered a public good because they create social order and reduce economic uncertainty. Concretely, norms, beliefs and ideologies are ways of solving collective-action problems such as freeriding.⁹³ To those ideas I add that the social impact of ideologies related to or promoted by states increased if states were larger and wielded greater power. That effect is particularly relevant for the case of the Roman empire as the largest and most powerful state in the ancient Mediterranean.

Nevertheless, I do not mean to imply here that a bigger state size was always preferable. City-states especially had many advantages, and their economic success in premodern European history, from classical Athens to medieval Venice, suggests that the efficiencies they produced could be considerable. To date, no comprehensive theory exists to explain why this might have been the case. But it is not difficult to see how city-states may have hit something of a “sweet spot” in their geographical extent and population size. A variegated literature has pointed to the benefits of their limited dimensions for public finance, information exchange, technological innovation and intracommunity cooperation.

John Hicks, in A Theory of Economic History, was attracted to city-states because of their historical prevalence, success and persistence. He posited that their small scale was “favourable to the growth of a diversified trade.” However, the structure of his argument required him to introduce legal institutions: “It is here that the city-state form of organization shows its superiority. The possibility of having recourse to regular legal institutions within the individual city state makes it easier for new kinds of trade to be carried on securely.”⁹⁴ But he inserts this element into his theory without any explanation of how such institutions might have emerged. As the discussion above has shown, his assumption also requires qualification in light of current Neo-Institutional work. Although Hicks sees the absence of formal legal institutions as an “insuperable obstacle” to intergroup commercial diversification, Leeson in his already-cited Anarchy Unbound makes a good case for the possibility of trade between heterogeneous groups based on private order.⁹⁵

Just as Hicks, Joel Mokyr presents city-states as “optimally sized” units of organization, but for a different reason. Although he, too, suggests that they might have had an advantage in contract enforcement, he proposes that they were well equipped predominantly for “the information-processing needed for trade,” adding that they “played, from the later Middle Ages on, an increasingly pivotal role in the generation of new useful knowledge and innovations.”⁹⁶ The role of any single medieval or Early-Modern city-state as an engine of creativity and innovation was usually just short lived. But as David Stasavage has pointed out, “While city-states were innovators for only a relatively brief span of time, all the evidence suggests that they retained their advantage with regard to public credit.”⁹⁷ City-states’ superior ability to borrow compared to territorial states was a product mainly of their political structure. The greater representativeness of their public bodies allowed them to make more credible repayment commitments to lenders, giving them easier access to long-term credit.

We are far less informed about the fiscal regimes of ancient city-states, but we do know that public borrowing was common in Greek cities, which may well have enjoyed a credible-commitment advantage similar to the one described by Stasavage.⁹⁸ Their small scale and political structure also spurred information exchange and institutional innovation, elements adduced by Ober to explain their “efflorescence,” defined mostly in economic terms. In addition, their nature as a collection of small, autonomous states allowed them to be agile in their collective-action response to outside pressure: “The endemic risk posed to Greek communities by would-be predatory states, Greek and non-Greek alike, fostered decentralized cooperation by rewarding the high levels of mobilization that were facilitated by federalism and democracy.”⁹⁹

But whatever advantages Greek and other ancient city-states may have drawn from their limited scale, in the long run they could not compete with bigger polities because of their restricted access to resources. In the ancient Mediterranean they ended up either being incorporated into larger states, as happened with Tyre and Athens, or growing into dominant empires themselves, as happened with Carthage and Rome. The incorporation of city-states into larger empires did not necessarily mean that they lost all the advantages of their small size. Much of the political and civic culture and decision-making process of Greek cities, for instance, remained the same under Hellenistic rule.¹⁰⁰ But obviously with the arrival of kings and, later, emperors as powerful hegemons, cities’ self-governance diminished and their political ecology changed, both individually and collectively.¹⁰¹ Nonetheless, in the aggregate their integration into larger state units seems to have been beneficial for economic development, as suggested by the upward economic trends in the proxy data presented in figures 1.1 and 1.2.

The downward trends past the Roman-era inflection point are equally significant.¹⁰² Those trends have recently been studied from a Neo-Institutional perspective by Daron Acemoglu and James Robinson, but unfortunately their analysis fails to convince on all counts. In the second, more subsidiary part they assert that Roman emperors feared Schumpeter-style creative destruction and sought to suppress technological progress. In support of that claim, they cite two infamous stories: one of a man who was beheaded after showing his invention of unbreakable glass to Tiberius (Pliny NH 36.195; Petr. Sat. 51; Dio 57.21.7), the second of a man who was dismissed after showing Vespasian an efficient way of moving heavy columns (Suet. Vesp. 18).¹⁰³ But to take these two almost certainly fictional tales as proof of a multi-reign, empire-wide policy aimed at stamping out innovation is patently absurd. (For my own thoughts on the matter of Roman innovational development, see chapter 7.)

The first and more institutional-analytical part of the discussion concerns state predation. Acemoglu and Robinson argue that the seed for the fall of Rome had been planted when the republic turned into the empire. At that time, growth ground to a halt because Rome’s “extractive institutions” had reached their limit. Citing the shipwreck and lead pollution graphs as evidence, they conclude that the “experience of economic growth during the Roman Republic was impressive, as were other examples of growth under extractive institutions, such as the Soviet Union.”¹⁰⁴ The regrettable comparison to the USSR aside, one reason why this line of thinking is fallacious has already been indicated. Both the shipwreck and the lead pollution graph show a steady rise centuries before Roman influence began. Acemoglu and Robinson’s argument would imply that all Mediterranean states from the Iron Age onward were characterized by “extractive institutions,” in which case there would be nothing exceptional about the Roman republic. That argument would be interesting, but it is not the one offered.

More fundamentally, it is generally agreed that fiscal pressure went down after the republic collapsed because Rome’s institutions became less, not more extractive (see the introduction and concluding remarks to chapter 3 for some details). In addition, the work by Hopkins discussed above shows that imperial levels of taxation were generally low. In his estimate, total state spending required no more than 10 percent of minimum Roman GDP, meaning that actual taxation levels were lower than that, perhaps between 5 and 7 percent.¹⁰⁵

Finally, Acemoglu and Robinson make no attempt to differentiate between either the early and later empire or between east and west, presenting things as on a universally downward trajectory after the end of the republic. This will not do. Much current work by Roman economic historians has focused on per capita growth, which seems to have begun under the republic and continued well into the empire. Both Jongman and Paul Erdkamp, for instance, have recently argued that real incomes continued to rise into the second or perhaps even the third century CE.¹⁰⁶ After that point we need to be careful to distinguish between the western part of the empire, where universal decline set in, and the eastern part, where some areas continued to do well, as Bryan Ward-Perkins has shown (see further, chapter 6).¹⁰⁷

Where Acemoglu and Robinson fail, Ward-Perkins makes the correct assessment: “A prerequisite for understanding the decline of the Roman economy is … the acceptance that it was a closely interlocked structure in which commerce and regional specialization, as well as the redistributive power of the state, all had an important role to play.”¹⁰⁸ The economy of the empire was sustained by a complex but delicate equilibrium that was eventually tipped off balance, in part by stochastic events.¹⁰⁹ It was robust enough to withstand one or more manageable shocks such as the severe but brief period of civil war during the “year of the four emperors,” 69 CE. But ultimately the imperial economy could not withstand several prolonged and simultaneous shocks, as happened in late antiquity. “Rome’s empire was always poised uncertainly between fragility and resilience, and in the end the forces of dissolution prevailed.”¹¹⁰ For my purposes there is no need to discuss all the endogenous and exogenous factors that have been proposed to explain the decline and ultimate demise of the Roman world. Staying with my main theme, I will limit myself to a study of one aspect of state behavior in late antiquity that I propose produced an institutional shock, adversely affecting the Roman system of intercommunity trade.

In the chapters to follow I will make these ideas concrete, focusing as much as I can not on what “the state” did but on what people did: people in their capacity as Greek city councilors, Ptolemaic officials or Roman magistrates; people in their capacity as long-distance traders, financiers or consumers of goods. This approach is intended to avoid speaking in abstractions where possible and to keep the discussion grounded in the sources. As for the book’s structure, it does not attempt a narrative history from the birth of ancient states to the ultimate collapse of the Roman empire. Instead, the separate chapters present targeted studies, organized loosely chronologically, into the ways in which ancient states influenced economic performance. Each study discusses a selected set of sources, but all aim to elucidate the broader economic developments occurring around the Mediterranean at the time.

In chapter 2 I will discuss the trade diaspora, an institution that was all but obligatory for regular, long-distance trade in the absence of third-party enforcement. Traders could do business in a foreign community because people from their homeland had moved there permanently and could vouch for them. If a promise to pay or deliver according to agreement was not fulfilled, foreign settlers could be held accountable for any debt of their fellow citizens. All members of a diaspora network, itinerant and stationary alike, ultimately faced expulsion if they behaved opportunistically. To reduce the vast amount of source material to a workable level, I will take the Phoenicians as a leading thread. Because of their centuries-long Mediterranean mercantile tradition, we can trace their interaction with public institutions through much of Greco-Roman history.

In studying that interaction, I will engage with the social-order theory proposed by North, Wallis and Weingast, specifically their postulate that natural states control trade, which is useful if applied flexibly.¹¹¹ I will approach the topic from two mirrored angles to argue that states’ role in trade had a beneficial effect on Mediterranean economic development: the public authority that Phoenician civic bodies claimed over their native diasporas, and the public authority asserted by host societies over those diasporas. As for the first angle, Phoenician ruling elites promoted trade from early on. The political institutions of some Phoenician cities still did so in the Hellenistic and Roman periods, involving themselves in the operation of their native trading associations overseas. As for the second angle, I will show how the institutions of Greek “public friendship” (proxenia) and subsequently Roman imperial ideology supported private order by facilitating the coordination of collective action.

The chapter is diachronic, treating a longer time span than chapters 3 to 5. It thereby sets the stage for the subsequent treatment of the Mediterranean economy, addressing long-term institutional developments connected to state formation, and touching on topics to be discussed later. An appendix will examine developments at Carthage, a city-state that grew into an empire at approximately the same time as the Hellenistic states, including the Ptolemaic kingdom on which chapter 3 centers. The trade diaspora will be a theme again in chapter 5, while Roman state ideology will recur there and in chapter 4.

In chapter 3 I will shift the discussion to mid-third-century BCE Syria-Palestine, an area ruled by the Ptolemies at the time. As in all Hellenistic states the power structure of the Ptolemaic kingdom was strongly personalized. A small number of state actors surrounding the king were allowed access to agricultural surplus, making them influential men politically and economically. The process by which the ruling elite negotiated power through conspicuous consumption, diplomatic tokens and gift-giving drove the wheels of an expanding economy of long-distance trade in both luxury goods and agricultural staples.

To investigate these matters, I will study the activities of a certain Zenon, who was an agent of the Ptolemaic finance minister Apollonios. We know of both men through a large collection of papyri containing correspondence between Zenon and his principal, and between members of a larger network of agents. By tracing the activities of these men we can analyze how as a public official Apollonios managed the Ptolemaic overseas territory of Syria-Palestine, at the same time allowing us to see how as a businessman he managed his private affairs there. Following Mancur Olson’s ideas about how state actors provide public goods if doing so allows them to reap the benefits, I will argue that the activities of Apollonios and his agents fit the model of the “stationary bandit.”¹¹² Apollonios was incentivized to keep Syria-Palestine socially stable and economically productive and thus to prevent its overexploitation. As we will also see, the Ptolemaic state that Apollonios served provided public goods, including a banking system and a public-order apparatus.

Contract enforcement and transaction costs under the Roman empire will be the topic of chapter 4. The unification of the Mediterranean Basin by a single state ameliorated economic conditions in practical ways: monetary and metrological systems were standardized, removing costly barriers to trade. Legal rules were standardized as well, which held the potential for an even greater transaction-cost-reducing effect. However, without third-party enforcement, private means still had to be employed to enforce contracts. It is therefore not immediately clear why transacting parties would adopt the Roman legal system.

I will argue that contracts drawn up in accordance with imperial law and in the presence of witnesses were “publicly embedded,” which increased their enforceability and reduced enforcement costs. Witnesses were listed following a status hierarchy largely determined by the civic order created by the state. Political officeholding conferred the highest social standing, while priesthoods of the imperial cult and the possession of freeborn citizenship also provided status. Acting as a witness to formal contracts gave members of Roman society an opportunity publicly to reaffirm their personal status and to endorse the civic order from which it derived. By allowing their social standing to be tied to contracts, witnesses provided transacting parties with incentives either to meet their contractual obligations or, should it come to that, cooperate in official litigation. I will argue that this enforcement-enhancing effect was an “emergent property,” the result of legal, social and ideological factors interacting in an undesigned and unintended fashion.¹¹³ Legal documents recorded on wax tablets from Pompeii and Transylvania will provide the source material for the discussion. They give insight into contracting practices in two different socioeconomic settings, having been found respectively in the old imperial heartland and a young province.

Chapter 5 will discuss the changing world of the fourth century CE, a time of social ferment heightened by the Roman emperors’ adoption of Christianity as a religion of state. Although this shift followed a turn toward forced religious centralization initiated by the emperors during the crisis of the third century, the choice for Christianity represented a momentous departure from Roman tradition. I will argue that the intolerance and violence it engendered upset the equilibrium of Mediterranean diaspora trade, producing an institutional shock.

Religion played a prominent role in how diaspora groups operated. Through the worship of their native gods, group members remained distinct from their hosts and connected to their place of origin, both necessary ingredients for successful intercommunity trade. Equally important, acts of religious devotion signaled commitment and loyalty to the group, encouraged collective action against defectors and fostered economic trust and collaborative behavior. Religious “honest signaling” was also helpful at the intercommunity level. It allowed diasporas to connect to other groups and to their hosts. By setting up inscriptions with religiously inspired wishes of well-being for reigning emperors, they displayed adherence to a shared imperial ideology, thereby lowering intergroup social barriers and signaling their trustworthiness as commercial partners.

But this complex system of socioeconomic interaction came under pressure when emperors began legislating against pagan cults. Groups involved in long-distance trade could no longer employ their native religious practices to engage in internal “honest signaling,” nor could they insert their religious practices into an imperial ideology as external “honest signals.”¹¹⁴ To investigate the institutional upset that these changes produced, the chapter will examine two well-documented instances of religious violence against pagan sanctuaries in Gaza and Alexandria, and one lesser known one in the city of Rome.

The Epilogue, chapter 6, will briefly discuss a topic mentioned in chapter 5 but not fully explored there: the weakening of the Roman state. In the third century CE the Roman empire began having trouble maintaining its geographical integrity, a problem that would grow noticeably worse thereafter. The split between an eastern and western half in 395 CE was the most dramatic manifestation of that decreasing ability. After the empire split in two, especially the west in the course of the fifth century saw the abandonment of peripheral areas, although signs of declining state power appeared in the east as well. But as the western half eventually disintegrated, the eastern half recovered. In the sixth century it managed to extend its rule over parts of the west, including the Italian heartland. But even with this westward expansion, and even allowing for healthy economic activity in some eastern regions, as a military and economic organization the Roman empire was nothing like the mighty state it once had been. A separate section will discuss the effects of the empire’s disintegration on human welfare, the topic of current research based on archaeological data including skeletal remains.

Finally in the Concluding Remarks, chapter 7, I will offer some tentative thoughts on the impact of Roman imperial rule on the impetus toward innovation. I propose that although the political and economic unification of the Mediterranean reduced transaction costs, it may also have reduced the drive toward innovation in both a technological and an institutional sense.

. Jones 2007: 143–47.

. Keynes 1920: 9–10.

. Jones 2007: 150–52.

. Mishra 2017: 13–14.

. Wilson 2011: fig. 2.5, building on the fundamental work by Parker 1992.

. Iron cannon increase the visibility of Early-Modern warships, though: Wilson 2011: 37.

. Wilson 2011: 33–39; 2014: 150–54; McCormick 2012: 89–97.

. McCormick 2012: 84, fig. 3.12.

. Kay 2014: 143–44.

. Kay 2014: 1, 327.

. Kay 2014: 327.

. Kay 2014: 314–18, 329. Kay notes that there probably was some inflationary pressure, but that it seems to have remained mild.

. Hollander 2007: 104–11. Quotation on p. 111.

. Hopkins 1978; Harris 1979; 2017; Kay 2014: 21–42.

. Hong, Candelone, Patterson and Boutron 1994.

. Some 70 percent for the period 366 BCE to 36 CE: Rosman, Chisholm, Hong, Candelone and Boutron 1997: 3416.

. Renberg, Bindler and Brännvall 2001; Hillman et al. 2017; McConnell et al. 2018; Friedman forthcoming.

. Introduction of coinage: Von Reden 2010: 20–25.

. McCormick 2001: 53, 703; Wilson 2002; De Callataÿ 2005; Jongman 2007a: 188–89; Kehoe 2007: 547–48; McConnell et al. 2018.

. Friedman forthcoming.

. Wilson 2014: 157. See McConnell et al. 2018 for more fine-grained data. I thank Andrew Wilson for allowing me a preview of this article, which unfortunately appeared too late for me to take full advantage of it.

. Wilson 2011: 37.

. Scheidel 2007: 42–43.

. Hopkins 2002; Saller 2002; Morris 2004: 259–60; Scheidel 2007: 43–44; Jongman 2006; 2007a; 2007b; 2014; Ober 2015: 81–84; Erdkamp 2016; Harper 2017: 29–38. See also chapter 6.

. Jongman 2007a: 191.

. Osborne 2009; Terrenato and Haggis 2011; Hansen 2013.

. Wiesehöfer 2013: 199–202.

. Hellenistic kingdoms: Ma 2013. Roman Republic: Harris 1979; Mouritsen 2013.

. Ameling 2013; Pilkington 2013.

. Tilly 1990; Wickham 2005.

. Wickham 2005: 57 n. 2.

. Wickham 2005: 57 with 303–06.

. See Zuiderhoek 2017: 149–59 for a discussion with references. For Athens as a “natural state,” see also chapter 2.

. Hodos 2006: 3–4. Bronze Age and Iron Age trade, see also Meyer 2006.

. Casson 1971: 362; Blackman 1982a: 93.

. See Von Reden 2010: 35–41 for Greek monetization and the financing of public works and services.

. See chapter 5 for details on Alexandria’s harbor. See also chapter 2, Appendix, for the harbors of Carthage.

. Arnaud 2014: 172. Roman harbors, see also Blackman 1982a; 1982b.

. Casson 1971: 365–67.

. Von Reden 2011. See also chapter 3.

. Hopkins 1980.

. Hopkins 2002: 199–200.

. Hopkins 2002: 219, 224.

. Scheidel 2011: 22.

. To the point where some have begun to question its usefulness: Verboven 2015.

. Lo Cascio 2006: 221.

. Manning 2003: 10.

. Ober 2015: 5.

. North 1981: 21.

. E.g., Boix 2015: 119–21 for the period of earliest state formation in Mesopotamia and the Near East.

. Earle 1991: 71–74.

. Leeson 2014: 1, italics in original.

. Leeson 2014: 15–31, 170–96.

. Cooter, Marks and Mnookin 1982. See also Posner 2000: 153–61.

. Posner 2000: 153.

. Keynes 1930: vol. 1, 4.

. Yarbrough and Yarbrough 1997: 139–48.

. Ogilvie 2011: 302–03.

. Greif 1989; 2006. For the debate between Ogilvie and Greif, see Edwards and Ogilvie 2012; Greif 2012.

. Greif 2012: 448.

. Ratzan 2015: 215–16, italics in original.

. Weber 1965: 2, italics in original (including the original German).

. North, Wallis and Weingast 2009: xii, 21–22, 73, 110.

. North, Wallis and Weingast 2009: 2–6, 18. Quotation on p. 18.

. North, Wallis and Weingast 2009: 41–49. Quotation on p. 41.

. North and Thomas 1973. On the evolution of North’s thinking, see also Wallis 2014.

. McCloskey 2016.

. Mokyr 2009: 368–88. Quotation on p. 378.

. Buchanan 1975: 95.

. Landa 1994: 62.

. North 1990: 58.

. Hansen 2002: 30–32.

. Herman 2006: 216–38; Bresson 2016: 239–50.

. Bravo 1980; Bresson 2007: 45; 2016: 318.

. Greif 2006.

. Manning forthcoming. On this dispute, see also Manning 2004: 763; 2018: 212–15.

. Bauschatz 2013.

. Bauschatz 2007: 7–9.

. In a law promulgated by Zeno (486 CE) and another by Justinian (529 CE) we still hear of private prisons, which the emperors attempted to abolish: Cod. Just. 9.5.1–2.

. Kaser and Hackl 1996: 382–407; Du Plessis 2015: 70–71, 77–78; Rüfner 2016: 265.

. Crook 1967: 82.

. Rüfner 2016.

. Garnsey 1970; Kaser and Hackl 1996: 519–21.

. Exsecutores dati: Litewski 1974: 227–29, 242–43; Von der Fecht 1999: 56; Díaz-Bautista Cremades 2013: 87–88.

. Corruption: MacMullen 1988: 137–70. Legal fees: Di Segni, Patrich and Holum 2003; Haensch 2015.

. See Ratzan 2015: 215–18 for a good discussion of the Roman empire as a category 3 state (not a term Ratzan uses) based on Egyptian papyri.

. North and Thomas 1973: 120. The point is repeated in North 1990: 113–17.

. But cf. McConnell et al. 2018: wars produced temporary setbacks in silver mining.

. Finley 1999: 150–76. For a recent discussion of the topic, see Gabrielsen 2011.

. Engen 2010: 5–6.

. North, Wallis and Weingast 2009: 53–54, 57.

. Boix 2015: 69–70.

. North 1981: 36–45; 2005: 103–08. On Roman civic ideology and social norms, see Verboven, 2015: 55–56.

. Hicks 1969: 42–59. Quotations on p. 46.

. Leeson 2014.

. Mokyr 2002: 280–81.

. Stasavage 2011: 43–46. Quotation on p. 44.

. Migeotte 1984; Mackil 2015: 481–84.

. Ober 2015: 70. On Greek institutional innovation, see Ober 2015: 117–20.

. See Strootman 2011 for a good discussion.

. Ober 2015: 293–315.

. Acemoglu and Robinson 2012: 158–75.

. Infamous since Finley 1965. For a different evaluation of the same stories, see Greene 2000: 46–50; Wilson 2002: 4.

. Acemoglu and Robinson 2012: 170. Cf. Hopkins 1978 for a widely accepted model of socioeconomic change from republic to empire.

. Hopkins 2002: 201. For Roman taxation, see also Hopkins 1980; Potter 2014: 51–61; Scheidel 2015; Tan 2015.

. Jongman 2006; 2007a; 2007b; Erdkamp 2016.

. Ward-Perkins 2000b; 2005.

. Ward-Perkins 2000b: 382.

. Verboven 2015: 56–57; Harper 2017.

. Harper 2017: 286.

. North, Wallis and Weingast 2009: 38.

. Olson 1991; 2000.

. On emergent properties, see Harré 1985: 145–46; Checkland 1999: 50; Ober 2015: 45.

. On signaling, see Posner 2000: 18–27; Bulbulia and Sosis 2011; Leeson 2014.