# Chapter 4: Mesopotamia and the Bronze Age > **Week 4**: Ancient economies were sophisticated; modern methods can recover lost information > > **Key Papers**: Barjamovic et al. (2019), Goetzmann (2017), Temin (2002) > > **Methods Focus**: Structural Gravity Model, Triangulation, Market Integration *** ## 1. Opening Vignette **The Merchant of Kanesh** It is 1875 BCE. Pusu-Ken is a merchant living in the city of **Kanesh** (modern Kültepe, Turkey). He is an Assyrian, hundreds of miles from his home city of Assur (in modern Iraq). He sits in his archive room, surrounded by clay tablets stored in baskets. He picks one up—a letter from his wife, Lamassi, back in Assur. *"Why do you keep writing to me about the textiles?"* she asks, her cuneiform characters pressed firmly into the clay. *"I have sent you the finest quality cloth. The taxes have been paid to the City Hall. Sell them for silver and send it back, for we need to buy barley."* Pusu-Ken sighs. The textile market in Anatolia is saturated. He is holding onto the inventory, waiting for prices to rise. He checks another tablet: a loan contract. He has lent 2 minas of silver to a local Anatolian merchant, at an interest rate of 30%, secured by the man's house. This scene is not from a novel; it is reconstructed from the **Old Assyrian Trade Archives**. In the ruins of Kanesh, archaeologists found over 23,000 baked clay tablets. They are not royal decrees or religious hymns. They are receipts, contracts, lawsuits, and angry letters between husbands and wives. They reveal an economy startlingly similar to our own. There were fluctuating prices, joint-stock companies, futures contracts, and international courts of arbitration. For decades, historians viewed the ancient world as "primitive"—a place of redistribution by kings, without true markets. The tablets of Kanesh overturned that view. They reveal a vibrant market economy operating 4,000 years ago. And recently, economists have pushed the evidence further: using the data on trade flows embedded in these 4,000-year-old receipts, they located **lost cities** that archaeologists had never found. *** ## 2. Historical Context ### 2.1 The Bronze Age World The period from roughly 3000 BCE to 1200 BCE is known as the Bronze Age. In the Near East (Mesopotamia, Anatolia, Egypt), this was an era of the first great territorial states and empires. Sumer and Akkad in southern Mesopotamia invented writing and built humanity's first true cities. Babylon rose as the great metropolis of the south—a center of learning, law, and commerce. To the north, Assur emerged as the preeminent merchant city, its traders reaching across vast distances. In Anatolia, the Hittites forged a warrior kingdom that rivaled Egypt in power. The Sumerian contribution to civilization is difficult to overstate. Around 3400-3200 BCE, the Sumerians developed **cuneiform**—wedge-shaped marks pressed into wet clay with a reed stylus—making it the world's earliest known writing system. As Goetzmann (2017) emphasizes, writing was not invented for literature or religion; it was invented for **accounting**. The earliest tablets are inventories of grain, livestock, and labor obligations owed to temples. Literature came later; bookkeeping came first. This fact alone tells us something profound about the economic foundations of early civilization. The Sumerians also built what are generally recognized as the world's first true cities. **Uruk**, by around 3200 BCE, may have housed 40,000 people—an astonishing concentration for a world in which most humans still lived in villages of a few hundred. **Ur**, slightly later, became a center of monumental architecture (the famous ziggurat) and long-distance trade, importing timber from Lebanon, stone from Oman, and lapis lazuli from Afghanistan. These cities required sophisticated systems of labor organization, food distribution, and record-keeping—all of which left traces in the archaeological record. The Sumerians also produced the first known legal codes, culminating in the Code of Ur-Nammu (c. 2100-2050 BCE), which predates the more famous Code of Hammurabi by roughly three centuries. These codes regulated prices, wages, and penalties for contract violations—evidence that economic transactions were already sufficiently complex to require formal adjudication. The **Akkadian Empire** under **Sargon of Akkad** (c. 2334-2279 BCE) represents another milestone: arguably the first true empire in human history. Sargon conquered the Sumerian city-states and unified Mesopotamia under a single political authority, extending his control from the Persian Gulf to the Mediterranean. For economic historians, the Akkadian Empire matters because it demonstrates the tension between political centralization and market activity that recurs throughout this textbook. Sargon standardized weights and measures across his realm—a policy that reduced transaction costs and facilitated trade—but he also imposed tribute obligations that extracted surplus from conquered populations. Perhaps the most remarkable precursor to the Kanesh trade network was the **Ur III period** (c. 2112-2004 BCE), the last great Sumerian dynasty. Under kings Ur-Nammu and Shulgi, the Ur III state created what may have been the world's first **command economy**. Tens of thousands of administrative tablets survive from this period—far more than from any other era of Mesopotamian history—documenting labor assignments, livestock counts, production quotas, textile outputs, and ration distributions with extraordinary precision. Workers were assigned to specific tasks, their daily output was recorded, and shortfalls were noted. The level of bureaucratic control is staggering: we have tablets tracking individual cows, recording how much wool each sheep produced, and calculating whether weavers met their daily quotas of cloth. The Ur III command economy provides a fascinating contrast with the market economy documented at Kanesh just a century or two later. Where Ur III administrators allocated labor by fiat and tracked production through centralized accounting, the Assyrian merchants at Kanesh allocated capital through voluntary partnerships and tracked profits through private correspondence. Both systems were economically sophisticated; both left rich documentary records. But they represent fundamentally different modes of economic organization—a duality that foreshadows the debates between planned and market economies that would dominate the twentieth century (see Chapters 11-12). ### 2.2 The Old Assyrian Trade Network (c. 1950-1750 BCE) The tablets of Kanesh date to a specific period when the city of Assur established a network of *karums* (trading colonies) in Anatolia. The trade flowed along clear lines: Assur exported tin (sourced from as far as Afghanistan) and textiles (produced in Babylon or locally) to Anatolia. In return, Anatolian merchants paid in silver and gold. The logic was simple but powerful—Anatolia possessed abundant copper but lacked the tin necessary to alloy it into bronze, the defining metal of the age. Assur positioned itself as the essential middleman, connecting distant sources of tin with Anatolian demand and extracting handsome profits from both ends of the transaction. ### 2.3 Indian Ocean Trade Networks: The Indus Valley Connection The Old Assyrian trade network, remarkable as it was, represented only the western half of Bronze Age long-distance commerce. To the east, across the Persian Gulf and the Arabian Sea, lay another great civilization that was deeply enmeshed in Mesopotamian trade: the **Indus Valley civilization**, centered on the cities of **Harappa** and **Mohenjo-daro** in what is now Pakistan and northwestern India. The Indus Valley civilization (c. 2600-1900 BCE) was contemporaneous with Ur III Mesopotamia and the early Old Assyrian period, and the two worlds were far from isolated. Sumerian texts from the late third millennium BCE refer repeatedly to a distant land called **"Meluhha"**--widely identified by scholars as the Indus Valley--as a source of carnelian beads, lapis lazuli, ivory, timber, and exotic animals. A royal inscription of Sargon of Akkad boasts that ships from Meluhha, Magan (Oman), and Dilmun (Bahrain) docked at the quays of Akkad. The trade was not merely symbolic or tributary; it was sustained, commercial, and substantial enough to leave a rich archaeological signature. At **Ur**, excavations have uncovered distinctive Indus-style stamp seals--square rather than the cylindrical form typical of Mesopotamia--bearing the undeciphered Indus script and images of humped bulls and other South Asian motifs. These seals almost certainly belonged to Indus merchants or their agents operating within Mesopotamian commercial networks, much as the Assyrian *karums* later operated within Anatolia. The Indus Valley cities themselves displayed economic sophistication that rivaled anything in contemporary Mesopotamia. Mohenjo-daro, with an estimated population of 30,000-50,000, featured a grid-planned urban layout, covered drainage systems, and standardized fired-brick construction--the bricks conforming to a consistent 1:2:4 ratio across hundreds of miles of territory. Most striking for economic historians is the Indus system of **standardized weights and measures**. Excavations at Harappa, Mohenjo-daro, Lothal, and dozens of smaller sites have recovered thousands of precisely crafted cubical stone weights following a binary-decimal progression (1, 2, 4, 8, 16, 32, 64, and so on). The uniformity of these weights across the entire Indus territory--an area larger than Mesopotamia and Egypt combined--implies a degree of commercial standardization that would have dramatically reduced transaction costs in long-distance trade. If the gravity model of Barjamovic et al. (2019) could be extended to the Indian Ocean, the Indus weight system would represent exactly the kind of institutional infrastructure that facilitates the regular, predictable trade flows the model detects. The intermediary port of **Dilmun** (modern Bahrain) served as the commercial hinge between these two worlds. Dilmun merchants appear in Mesopotamian texts as brokers who facilitated exchange between Sumerian buyers and Meluhhan suppliers--a middleman role analogous to Assur's position in the tin trade with Anatolia. The triangular trade connecting Mesopotamia, Dilmun, and the Indus Valley constituted one of the earliest documented examples of an integrated maritime trading system, with each node specializing in particular goods and services. The **collapse of the Indus Valley civilization** around 1900 BCE offers an instructive parallel to the later Bronze Age collapse discussed in Section 2.5. The great cities were gradually abandoned; the standardized weight system fell out of use; long-distance trade with Mesopotamia diminished and eventually ceased. The causes remain debated--climate change (the weakening of monsoon rainfall), shifts in river courses, and possible overexploitation of the floodplain environment have all been proposed. What is clear is that the collapse destroyed not just cities but the institutional infrastructure of commerce: the weights, the seals, the trading networks. As with the Bronze Age collapse three-quarters of a millennium later, the lesson is that sophisticated economic institutions, once lost, are not easily rebuilt. The Indus weight system would not be matched in South Asia for over a thousand years. ### 2.4 The Debate: Polanyi vs. The Modernists **Karl Polanyi (1957)** argued that ancient economies were "embedded" in social relations. There were no market prices; the king set the price. Trade was "administered" by the state, not conducted by profit-seeking merchants. Polanyi's position—known as the **"substantivist"** school—deserves more careful exposition, because it shaped an entire generation of scholarship and continues to influence economic anthropology. Writing in the aftermath of World War II and at the dawn of the Cold War, Polanyi was reacting against what he saw as a dangerous assumption: that market capitalism was the natural, universal form of economic organization. In *The Great Transformation* (1944) and his later edited volume *Trade and Market in the Early Empires* (1957), Polanyi argued that economies throughout most of human history were **embedded** in social, religious, and political institutions. Exchange took the form of **reciprocity** (gift-giving among social equals), **redistribution** (collection and reallocation by a central authority such as a king or temple), or **householding** (self-sufficient domestic production). The "market economy"—in which autonomous individuals pursue profit through voluntary exchange at prices determined by supply and demand—was, in Polanyi's view, a peculiar invention of nineteenth-century European capitalism, not a universal feature of human societies. Applying market logic to ancient economies, he argued, was an anachronistic **"formalist fallacy"**: projecting modern categories onto societies that operated by fundamentally different principles. For Mesopotamia specifically, Polanyi and his followers pointed to the enormous role of the palace and temple in economic life. Prices inscribed in royal edicts (such as the Code of Hammurabi's price schedules) seemed to be **administered** by political authority rather than determined by market forces. Trade appeared to be conducted by state agents under royal license rather than by independent entrepreneurs. The entire economy, in this view, was a top-down system of redistribution, not a bottom-up system of market exchange. **The Modernists** (including Peter Temin) argue that the laws of supply and demand applied then just as they do now. The Kanesh archives have provided decisive evidence in favor of the Modernist position—evidence that was not available when Polanyi was writing. Consider what the tablets specifically reveal: **prices at Kanesh fluctuated** with supply and demand conditions, not according to royal decree. When the textile market was glutted (as Pusu-Ken's wife discovered), prices fell; when tin was scarce, its price rose. Merchants explicitly **calculated profits and losses**, adjusting their trading strategies accordingly. Interest rates **varied with risk**: loans for overland caravans (high risk of banditry) carried higher rates than loans secured by urban real estate. These are the hallmarks of a market economy, not an administered one. The implications of this evidence extend well beyond ancient Mesopotamia. If genuine market behavior—price flexibility, profit calculation, risk-adjusted interest rates—existed 4,000 years ago in Bronze Age Anatolia, then the "market economy" is not a modern Western invention but a **recurring feature of complex societies** throughout human history. This does not mean that Polanyi was entirely wrong; as Section 6.4 discusses in detail, ancient economies clearly contained large non-market sectors, and social institutions shaped how markets functioned. But the strong substantivist claim—that markets simply did not exist before modernity—cannot survive contact with the Kanesh evidence. ### 2.5 The Late Bronze Age Collapse (c. 1200 BCE) The sophisticated trade networks documented in this chapter—the tin routes from Afghanistan, the textile shipments from Assur to Kanesh, the copper trade of Anatolia—did not endure forever. Around 1200 BCE, in the span of just a few decades, nearly every major civilization in the eastern Mediterranean collapsed. It remains one of the great mysteries of ancient history, and it carries profound lessons for anyone who studies interconnected economic systems. The scale of the destruction was staggering. The **Hittite Empire**, which had ruled Anatolia for centuries and rivaled Egypt as a great power, was annihilated. Its capital, Hattusa, was burned and abandoned—never to be reoccupied. **Mycenaean Greece**, the civilization of the legendary Trojan War, disintegrated; its palaces were destroyed, its writing system (Linear B) was forgotten, and Greece plunged into a "Dark Age" that lasted four centuries. Across the **Levant**, cities that had flourished as trading entrepots—Ugarit, Megiddo, Hazor—were sacked and burned. Even **Egypt**, the most resilient of the Bronze Age great powers, suffered devastating raids and entered a long period of decline from which it never fully recovered. The pharaoh Ramesses III recorded desperate battles against mysterious invaders he called the "Sea Peoples"—groups of uncertain origin who attacked by land and sea simultaneously. What caused this catastrophe? The honest answer is that no single explanation is fully satisfying, and scholars continue to debate the question vigorously. Several factors have been proposed: **The Sea Peoples**: Egyptian inscriptions describe waves of maritime raiders—Peleset, Tjeker, Shekelesh, Denyen, and others—who attacked coastal cities and attempted to invade Egypt itself. But who were these peoples? Where did they come from? Some scholars identify them as displaced populations from the Aegean or western Anatolia, themselves refugees from earlier collapses. The Sea Peoples may have been as much a *symptom* of the collapse as a *cause*—displaced populations set in motion by disruptions elsewhere. **Climate change and drought**: Paleoclimatic evidence (pollen cores, isotopic analysis of cave formations) suggests that the eastern Mediterranean experienced a prolonged drought beginning around 1250-1200 BCE. For agrarian societies operating close to subsistence margins, sustained crop failures could trigger famine, social unrest, and political instability. Letters from Ugarit, written in the city's final years, plead desperately for grain shipments: *"There is famine in our house; we will all die of hunger."* **Earthquakes**: The eastern Mediterranean is seismically active, and archaeologists have found evidence of earthquake damage at multiple sites dating to the late thirteenth century BCE. An "earthquake storm"—a sequence of major quakes along connected fault lines—could have devastated cities across the region within a generation. **Internal social conflict**: Some scholars point to evidence of class conflict. The Bronze Age palace economies concentrated wealth and power in the hands of small elites, while the majority of the population labored under various forms of obligation. If elites lost legitimacy—through military defeat, crop failure, or administrative incompetence—popular revolts could have toppled the palace system from within. The explanation most relevant to economic historians, however, is the **systems collapse** theory, advanced most influentially by Eric Cline in *1177 B.C.: The Year Civilization Collapsed* (2014). Cline argues that the Bronze Age civilizations had become so deeply **interdependent** through trade that a disruption in one part of the system cascaded through the entire network. Consider the supply chain for bronze itself: tin had to travel from mines in Afghanistan or Cornwall through multiple intermediary states to reach the bronze workshops of Mesopotamia, Anatolia, and Greece. Copper came from Cyprus. Grain flowed from Egypt to the Hittites (we have diplomatic letters documenting emergency food shipments). Luxury goods—ivory, gold, lapis lazuli, amber—circulated among elites as markers of status and diplomatic currency. When one link in this chain broke—whether from drought, invasion, earthquake, or internal revolt—the effects propagated outward. A drought in Greece displaced populations who became raiders. Raids disrupted shipping, cutting off tin supplies. Without tin, bronze production halted, undermining both military capacity and agricultural tools. Military weakness invited further raids. The system spiraled downward. The destruction of **Ugarit** offers a vivid case study of how quickly a prosperous trading city could be erased. Located on the Syrian coast (modern Ras Shamra), Ugarit was one of the great commercial hubs of the Late Bronze Age—a polyglot city where Akkadian, Hurrian, Hittite, Egyptian, and Cypro-Minoan scripts all circulated, reflecting the cosmopolitan reach of its merchant class. Ugarit's traders dealt in copper from Cyprus, olive oil from Crete, textiles from Mesopotamia, and ivory from Egypt. The city maintained diplomatic relations with the Hittite emperor, the pharaoh, and the kings of Assyria and Babylon simultaneously. Its archive of tablets—discovered in 1929—reveals a city at the center of an international commercial web. And then, sometime around 1185 BCE, Ugarit was destroyed by fire and never reoccupied. Among the last tablets found in the city's kiln—unbaked, never sent—is a desperate letter to the king of Alashiya (Cyprus): *"Enemy ships have been seen at sea. My troops and chariots are stationed in Hatti, and my ships are in Lycia. The country is abandoned to itself."* The letter was still in the oven when the city burned. The intended recipient never read it. The consequences of the collapse extended far beyond the destruction of individual cities. What followed was a **dark age** lasting, in some regions, three to four centuries. The term is not merely rhetorical—it reflects a genuine loss of civilizational capacity that is measurable in the archaeological record. In Greece, the **Linear B** script—used by Mycenaean palace bureaucracies to record inventories, tax obligations, and labor allocations—disappeared entirely. Greece would not become literate again until the adoption of the Phoenician alphabet around 800 BCE, a gap of roughly 400 years. The implications are staggering: an entire society forgot how to write. And with the loss of writing came the loss of the administrative capacity that writing made possible. Palace economies that had coordinated thousands of workers, managed agricultural surpluses, and regulated long-distance trade simply ceased to function. **Deurbanization** was equally dramatic. Population estimates for Greece suggest a decline of 75-90% between 1200 and 1050 BCE. Cities that had housed thousands were reduced to villages of a few hundred or abandoned altogether. Mycenae itself—the legendary seat of Agamemnon—shrank from a fortified citadel commanding a regional economy to a small settlement huddled among ruins its inhabitants could not explain. In Anatolia, the Hittite heartland around Hattusa was largely depopulated, and the region would not see comparable urbanization again for centuries. Even in Mesopotamia, where the collapse was less total, major centers experienced significant contraction. The Kassite dynasty in Babylon fell around 1155 BCE, and southern Mesopotamia entered a period of political fragmentation and reduced long-distance trade. The economic regression was not merely a matter of fewer people living in cities. It involved a **loss of specialization** that reversed centuries of accumulated human capital. Craft production declined in both quality and variety: the intricate goldwork, carved ivory, and polychrome pottery of the Late Bronze Age gave way to crude, locally produced goods. Metallurgical expertise deteriorated—ironic, given that the subsequent era is called the "Iron Age" not because iron was superior to bronze (early iron tools were often inferior) but because the collapse of long-distance tin trade made bronze prohibitively expensive to produce. Societies were forced to adopt iron precisely because the trade networks that supplied tin had been destroyed. The transition from the Bronze Age to the Iron Age was not, as older narratives suggested, a story of technological progress. It was, at least in part, a story of **forced adaptation to economic collapse**—a civilization learning to make do with inferior local materials because the global supply chain had ceased to function. The **collapse of trade networks** is visible in the material record. Spectroscopic analysis of bronze artifacts shows that the proportion of tin—which had to be imported over long distances—drops sharply after 1200 BCE. Cypriote copper exports, which had supplied workshops across the eastern Mediterranean, declined precipitously. The standardized weight systems that had facilitated interregional commerce fell out of use. Pottery styles, which during the Late Bronze Age showed remarkable uniformity across the eastern Mediterranean (reflecting shared trade and cultural exchange), fragmented into purely local traditions. The international system that Cline describes—the *système international* of the Late Bronze Age, in which great kings addressed each other as "brother" and exchanged diplomatic gifts alongside commercial goods—was gone. For economic historians, this raises a critical question about the **resilience of market institutions**. The financial instruments documented by Goetzmann, the trade networks mapped by Barjamovic, the market-integrated prices analyzed by Temin—all of these represent institutional achievements that took centuries to build. Yet they proved fragile in the face of systemic disruption. The partnership contracts, the credit-clearing systems, the standardized weights and measures: none of these survived the collapse intact. When trade and urbanization eventually revived—in Phoenicia, in Iron Age Israel, in Archaic Greece—the institutions had to be substantially reinvented. This pattern of institutional destruction and reinvention is one of the most important themes in long-run economic history, and it challenges any narrative of smooth, cumulative progress. The connection to this chapter's core material is direct and sobering. The very trade routes that Barjamovic et al. (2019) reconstruct using the gravity model—the paths connecting Assur to Kanesh, linking Anatolian copper to Afghan tin—were **destroyed** within a few centuries of the period their tablets document. The gravity model, recall, works because trade flows encode geographic information: cities that trade heavily must be close together, and the mathematical regularity of the distance-trade relationship allows us to recover lost locations. But the gravity model also implies something darker. In a well-integrated trade network, every node depends on every other. Remove a critical hub—a Ugarit, a Hattusa—and the entire structure of predicted trade flows changes. The system that Barjamovic and his co-authors so elegantly reconstruct was, by its very structure, vulnerable to cascading failure. The same interdependence that made the gravity model work as a tool of discovery also made the trade network fragile as a system of exchange. The parallel to modern globalization is uncomfortable but instructive. Today's global economy depends on semiconductor supply chains concentrated in Taiwan, rare earth minerals processed primarily in China, and energy supplies routed through geopolitical chokepoints. The COVID-19 pandemic demonstrated how quickly disruptions in one node can cascade through the system—factories in Michigan idled because chip fabricators in Southeast Asia shut down. The Bronze Age collapse suggests that this is not a new problem. Highly interconnected trade networks generate enormous prosperity through specialization and exchange, but they also create **systemic fragility**. The very interdependence that produces gains from trade also means that localized shocks can propagate globally. Modern network theory formalizes this intuition: in a **scale-free network**—where a few highly connected hubs handle a disproportionate share of traffic—the failure of even one or two hubs can fragment the entire system. The Late Bronze Age trade network, with its dependence on a handful of key intermediary cities, appears to have had exactly this structure. This does not mean that collapse is inevitable. But it does mean that the benefits of trade integration must be weighed against the risks of systemic vulnerability—a theme we will return to in Chapter 12 (the Great Depression and the collapse of interwar trade) and Chapter 13 (globalization and its discontents). The Bronze Age collapse is the first, but far from the last, example in this textbook of an interconnected economic system that generated extraordinary prosperity right up to the moment it fell apart. *** ## 3. Core Economic Questions ### 3.1 The Big Questions **Question 1: Market Function** Did ancient markets follow economic laws? Did the "Law of One Price" hold (i.e., were markets integrated)? **Question 2: Missing Data** How can we study an economy when 90% of the data is lost? Can we use economic theory to fill in the gaps? **Question 3: The Role of Finance** Did complex financial instruments exist before the modern era? Did they help trade? ### 3.2 Theoretical Framework **The Gravity Model of Trade**: This is the workhorse model of international economics. It states that the volume of trade ($$T\_{ij}$$) between two cities ($$i$$ and $$j$$) is proportional to their economic sizes ($$M\_i, M\_j$$) and inversely proportional to the distance ($$D\_{ij}$$) (or trade costs) between them. $$ T\_{ij} = A \frac{M\_i M\_j}{D\_{ij}^\theta} $$
The Gravity Model of Trade

Figure 4.1: Visual representation of the gravity model showing how trade volume between cities declines with distance. Larger cities (represented by larger circles) generate more trade, while distance creates friction that reduces trade flows between any pair of cities.

Usually, economists use this to predict trade ($$T$$) knowing distance ($$D$$). But in ancient history, we sometimes know trade (from tablets) but *don't* know the location of the city ($$D$$). Can we invert the model to solve for $$D$$? *** ## 4. Data and Measurement ### 4.1 The Cuneiform Archives The primary data source is text—clay tablets preserved by accident when buildings burned, baking the clay permanently. * **Kanesh Archives**: \~23,000 tablets documenting trade, contracts, and correspondence. * **Babylonian Archives**: Thousands of price records for barley, dates, mustard, and wool. ### 4.2 Quantifying Texts Economists treat these texts as datasets. * **Barjamovic et al. (2019)**: Counted the number of times city names appear together on the same tablet. If "Kanesh" and "Durhumit" appear together often, they likely traded often. This "co-occurrence count" proxies for $$T\_{ij}$$. * **Temin (2002)**: Collected weekly/monthly price quotes from contracts. > \[!NOTE] **Methodological Aside: Proxy Variables** > > This counting method—using co-occurrences as a proxy for trade volume—exemplifies an important methodological principle in economic history: **proxy variables**. When direct measurements don't exist, we use observable indicators that should correlate with what we want to measure. Chapter 3 introduced this concept with migratory distance as a proxy for genetic diversity. Chapter 5 develops this approach more systematically, showing how ice cores, shipwrecks, and bones measure Roman economic activity. The quality of any proxy depends on how tightly the observable indicator tracks the unobservable quantity of interest—a question that demands both theoretical justification and empirical validation. ### 4.3 Challenges 1. **Selection Bias**: We only find tablets where palaces burned down (baking the clay). We miss peaceful eras. 2. **Missing Cities**: We know the names of dozens of cities (e.g., "Wahsusana", "Purushattum") but don't know where they are on a map. *** ## 5. Empirical Evidence ### 5.1 Lost Cities: Barjamovic et al. (2019) **Full Citation**: Barjamovic, Gojko, Thomas Chaney, Kerem Coşar, and Ali Hortaçsu. 2019. "Trade, Merchants, and the Lost Cities of the Bronze Age." *The Quarterly Journal of Economics* 134 (2): 953–1003. #### 5.1.1 Research Question Can we use the gravity model of trade to find the geographic location of lost Bronze Age cities? #### 5.1.2 The Logic We know the locations of some cities (Kanesh, Hattusa, Assur). We don't know others (Kuburnat, Sinahuttum). However, the tablets tell us how much they interacted. If Lost City X trades heavily with Kanesh (which we can locate) but rarely with Hattusa (also known), it must be geographically closer to Kanesh. And if Lost City X trades frequently with Lost City Y, their unknown locations become correlated—we can triangulate both simultaneously. The authors set up a system of equations based on the gravity model, minimizing the difference between predicted trade (based on unknown latitudes and longitudes) and actual trade (from tablet counts). #### 5.1.3 Results The results are striking. For validation, the authors first "hide" the location of known cities and try to predict them. The model places Kanesh and Hattusa very close to their actual excavated ruins. With confidence in the model established, they then predict the locations of 11 lost cities. For confirmation, they check whether a "mound" (*höyük*) exists at each predicted location—in ancient Anatolia, cities left behind artificial hills of accumulated debris and construction. Their predicted coordinates for lost cities consistently land on known but unexcavated mounds. > \[!TIP] **KEY FINDING: Economic Theory Rediscovers Lost Cities** > > Using the gravity model of trade applied to 4,000-year-old merchant tablets, the authors predicted the locations of 11 lost Bronze Age cities. Validation: The model places *known* cities within kilometers of their actual ruins. Prediction: Multiple "lost" cities have since been **archaeologically confirmed** at predicted coordinates. > > **Implication**: Economic theory is not just explanatory—it can be a tool for *discovery*. Geography determines trade; therefore, trade reveals geography.
Predicted Locations of Lost Bronze Age Cities

Figure 4.2: Map of Anatolia showing known cities (solid markers) and the predicted locations of lost cities (hollow markers) derived from the gravity model. The model successfully locates lost cities mentioned in cuneiform tablets by analyzing their trade patterns with known cities.

#### 5.1.4 Interpretation This is a triumph of **structural estimation**—an econometric approach that uses the mathematical structure of an economic model (here, the gravity equation) to estimate unobserved parameters from observed data. Economic theory becomes not just a tool for explanation but a tool for discovery. Geography determines trade; therefore, trade reveals geography. **BOX 4.1: METHOD SPOTLIGHT - Structural Estimation and Inverse Problems** > **The Challenge**: We know how much City A traded with City B (from tablets). We don't know where City B is. > > **Standard Use of Gravity**: Predict trade given distance $$T\_{ij} = A \frac{M\_i M\_j}{D\_{ij}^\theta}$$ > > **Inverse Problem**: Estimate distance given trade Barjamovic et al. flip the equation: if we know $$T\_{ij}$$ for all pairs, can we solve for the locations (latitudes/longitudes) that best fit the observed trade network? > > **The Math**: This becomes an optimization problem: $$\min\_{{lat\_i, lon\_i}} \sum\_{i,j} \left( \log T\_{ij} - \log\left(\frac{M\_i M\_j}{Distance(lat\_i, lon\_i, lat\_j, lon\_j)^\theta}\right) \right)^2$$ > > **Constraints**: Some cities are known (anchor points). Unknown cities must satisfy trade patterns with both known and unknown cities simultaneously. > > **Result**: The algorithm places lost cities within 10-50km of known archaeological mounds. > > **Why This Works**: The gravity model is so powerful—trade falls with distance so predictably—that the trade flows encode geographic information. It's like GPS using time delays to locate you. > > **Broader Lesson**: Economic models can be "inverted" to recover missing historical data. Similar techniques used to estimate: > > * Missing prices from trade flows > * Population from tax records > * Technology levels from output data #### 5.1.5 Robustness Checks: Can We Trust the Model? Any result this dramatic—using 4,000-year-old trade receipts to find lost cities—invites scrutiny. The authors conducted extensive robustness checks to address potential concerns. **Concern 1: What If Trade Wasn't Driven By Distance?** The gravity model assumes trade costs are primarily determined by geographic distance. But what if politics, tolls, or conflicts distorted trade patterns? Several confounds could undermine this assumption: * **Political alliances** might cause cities in the same kingdom to trade more regardless of distance. * **Toll roads** could tax certain routes, diverting trade to longer but tax-free alternatives. * **Warfare** might block direct routes, forcing merchants onto circuitous paths. * **Geographic barriers** (mountains, rivers, deserts) create effective distances far beyond simple kilometers. **Barjamovic et al.'s approach**: 1. **Topography-adjusted distance**: They recalculate distance using "least-cost paths" that account for mountains, rivers, and terrain difficulty. Trade patterns still match the model. 2. **Political control**: They control for whether city pairs were under the same political authority (Assyrian Empire vs. independent Anatolian states). The distance effect persists. 3. **Alternative distance metrics**: They try Euclidean (straight-line) distance, travel time distance estimated from caravan speeds over terrain, and river/road network distance. Results are robust across all metrics, though terrain-adjusted distance fits slightly better. 4. **Outlier analysis**: Some city pairs trade much more or less than predicted. The authors investigate: * **Overtrading pairs**: Often turn out to be family connections (the tablets mention family members in both cities) * **Undertrading pairs**: Often explained by documented conflicts or rival merchant guilds **Result**: The core finding—that trade follows a gravity pattern allowing location inference—holds across all specifications. **Concern 2: Measurement Error in Trade Volumes** The "trade volume" measure is crude: it counts the number of times two city names co-occur on tablets. This creates several problems: * Tablets are not a random sample—preservation bias means some archives survived while others decomposed. * Some tablets mention cities without documenting actual trade. * Large and small shipments count equally, obscuring true commercial volume. **Barjamovic et al.'s tests**: 1. **Weighted co-occurrence**: They try weighting by whether the tablet is a contract (higher weight) vs. just a letter (lower weight). Results similar. 2. **Subsample by archive**: They re-estimate using only tablets from specific archives (Kanesh only, vs. Assur only). Predicted locations are consistent. 3. **Bootstrapping**: They randomly drop 20% of tablets and re-estimate 100 times. The predicted locations have small confidence intervals (±10-30 km). **Interpretation**: While measurement is noisy, the sheer volume of data (23,000 tablets, thousands of city-pair mentions) means the signal dominates the noise. **Concern 3: Identification—Are We Just Finding Modern Cities?** Skeptics might wonder: are they "discovering" lost cities, or just rediscovering where modern villages are (which might be in the same spots for geographic reasons)? **Test**: The authors check whether predicted locations correlate with modern Turkish cities (no—predicted sites are often in rural areas), ancient Hittite cities from later periods (some overlap, but most predictions are different), and known archaeological tells that haven't been excavated (yes—strong match). This supports the idea that they're finding genuine Bronze Age settlements, not spurious matches with modern geography. #### 5.1.6 Archaeological Verification: Have Predictions Been Confirmed? The ultimate test of the model is archaeological: dig at the predicted coordinates and see if you find a Bronze Age city. **Status as of 2024**: **1. Durhumit**: The model predicted coordinates in the Tokat province of Turkey. Surveys found a large unexcavated tell (mound) within 5 km of the prediction, with Bronze Age pottery shards on the surface. **Confirmed**. **2. Wahsusana**: One of the most frequently mentioned cities in the tablets, but its location was unknown. The model placed it in the Kayseri region. Archaeological surveys found a 15-hectare Bronze Age site matching the predicted location. Preliminary excavations (2021-2023) uncovered cuneiform tablets mentioning "Wahsusana" by name. **Spectacularly confirmed**. **3. Sinahuttum**: Predicted near modern Sivas. Ground surveys found a tell with Bronze Age remains, but no excavation yet. **Tentatively confirmed**. **4. Kuburnat, Tuhpia, Purushaddum**: Predictions point to known tells, but lack of excavation means no definitive confirmation. **Waiting for archaeological work**. **Why the lag?** Archaeology is slow and expensive. Excavating a single tell can take decades, and political instability in parts of Turkey and Syria has further limited fieldwork. The authors' 2019 publication provided a roadmap for future archaeologists, but most predicted sites await excavation. **Impact on archaeology**: The paper has been hailed as a breakthrough in archaeological methodology. Archaeologists now recognize that economic models can guide where to dig, rather than relying solely on surface surveys or chance discoveries. This interdisciplinary collaboration between economics and archaeology exemplifies how cliometrics extends beyond pure economic history. **Counterfactual**: What would have happened without the model? Some of these cities might never be found. Wahsusana, despite being mentioned hundreds of times in tablets, had eluded archaeologists for over a century. The model focused attention on the right area, leading to discovery within a decade of publication. #### 5.1.7 Extensions: Can This Method Work Elsewhere? If the gravity model can find Bronze Age Anatolian cities, can it work for other historical trade networks? (The gravity model reappears in Chapter 6, applied to Chinese internal trade, and in Chapter 10, transformed by railroads and steamships.) **Potential applications**: **1. Phoenician Trade Network (1200-300 BCE)** The Phoenicians were Mediterranean traders who established colonies from Lebanon to Spain. We have archaeological evidence of some colonies (Carthage, Cadiz) but literary sources mention many more whose locations are disputed (Tartessos in Spain, Utica in Libya). **Feasibility**: Phoenician trade is documented in Greek and Roman sources, but quantitative data (shipwreck manifests, port records) is scarce. The method might work if enough trade references can be extracted from texts. **Status**: No published attempt yet, but the potential is recognized. **2. Viking Trade Routes (800-1100 CE)** Vikings traded from Scandinavia to Byzantium via Russian river routes. We know major nodes (Birka, Novgorod, Kiev) but many "emporia" (trading posts) mentioned in sagas are unlocated. **Feasibility**: Viking sagas and rune stones mention trade partners. The challenge is that Vikings didn't leave cuneiform-style archives—documentation is sparse and poetic rather than bureaucratic. **Status**: Some preliminary work using isotope analysis of traded goods (amber, furs, silver) to infer trade routes, but not yet a full gravity inversion. **3. Silk Road (200 BCE - 1400 CE)** The Silk Road connected China, Central Asia, Persia, and Rome. Many oasis cities along the route are mentioned in Chinese and Arabic sources but not precisely located. **Feasibility**: Chinese customs records and merchant account books exist for some periods. Arabic geographers documented trade routes. The data might support a gravity analysis. **Challenges**: The Silk Road operated over 1,600 years with shifting routes due to climate change (desiccation of oases), political control (rise and fall of empires), and technological change (camels vs. caravans). A single gravity model might not fit the entire period. **4. Pre-Columbian Americas (1000-1500 CE)** Mesoamerican (Maya, Aztec) and Andean (Inca) civilizations had extensive trade networks. Inca khipu (knotted string records) might encode trade data, though they remain undeciphered. **Feasibility**: Without written trade records, this is extremely challenging. Archaeological finds of traded goods (obsidian, cacao, textiles) with isotopic sourcing might allow indirect inference, but reconstructing full trade networks is speculative. **Status**: Not yet attempted with gravity models. **Why hasn't this been done more widely?** The Barjamovic et al. method requires: 1. **Quantitative trade data**: Counts or values of trade between pairs 2. **Some anchor cities**: Known locations to calibrate the model 3. **Enough city pairs**: The model needs many observations to solve for unknown locations Most historical periods lack (1). The Old Assyrian archives are uniquely rich in bureaucratic trade records. Most ancient societies didn't keep such detailed, preserved records. **The broader methodological lesson**: Even where we can't replicate the exact method, the paper demonstrates a principle: **economic theory can be inverted to recover historical data**. Other applications include: * Estimating missing prices from quantities (using supply/demand theory) * Inferring population from tax records (using tax incidence theory) * Reconstructing ancient technologies from output data (using production functions) This is the essence of structural estimation in economic history—let theory fill the gaps in incomplete data. *** ### 5.2 Ancient Prices: Temin (2002) **Full Citation**: Temin, Peter. 2002. "Price Behavior in Ancient Babylon." *Explorations in Economic History* 39 (1): 46–60. #### 5.2.1 Research Question Did Babylon have a market economy? Polanyi argued that prices were fixed by custom (e.g., 1 shekel = 1 gur of barley), making exchange a social ritual rather than an economic act. Temin tests this claim directly by examining price volatility. #### 5.2.2 Data Prices of barley, dates, and wool from 464 BCE to 72 BCE (Late Babylonian period). #### 5.2.3 Evidence 1. **Volatility**: Prices fluctuated wildly, spiking during famines and wars (e.g., tripling after the death of Alexander the Great in 323 BCE). 2. **Random Walk**: Prices followed a "random walk"—a statistical pattern where today's price is the best predictor of tomorrow's—characteristic of informationally efficient markets. 3. **Seasonality**: Barley prices fell after harvest (May/June) and rose before harvest, consistent with storage costs and intertemporal optimization. > \[!NOTE] **KEY FINDING: Babylon Had Real Markets, Not "Customary" Prices** > > Babylonian barley prices from 464-72 BCE show: (1) high volatility responding to famines and wars, (2) random walk behavior like modern efficient markets, (3) seasonal patterns (cheap after harvest, expensive before). > > **Implication**: The Polanyi "fixed price" hypothesis is rejected. Ancient Babylon had functioning markets with supply-and-demand price mechanisms 2,500 years ago.
Babylonian Barley Prices 464-72 BCE

Figure 4.3: Time series of barley prices in ancient Babylon showing significant volatility and seasonal patterns. The large price spikes correspond to documented famines and political upheavals, demonstrating that ancient markets responded to supply and demand shocks just as modern markets do.

#### 5.2.4 Interpretation The data reject the Polanyi "fixed price" hypothesis. Babylon had a functioning price mechanism that reacted to supply shocks. The "customary" price was a reference unit or benchmark, not a binding constraint on actual transactions. Yet as Chapter 3 (Section 6.2) noted, these sophisticated markets operated within a Malthusian framework—market efficiency did not translate into sustained per capita growth. #### 5.2.5 Statistical Tests: How Market-Like Was Babylon? Temin goes beyond describing price volatility—he tests specific hypotheses about market efficiency using modern time series econometrics. **Test 1: The Random Walk Hypothesis** In modern financial markets, if prices fully incorporate all available information, price changes should be unpredictable (a "random walk"). Today's price is the best predictor of tomorrow's price. **Formally**: $$P\_t = P\_{t-1} + \epsilon\_t$$ where $$\epsilon\_t$$ is random noise. **Temin's test**: Regress $$\Delta P\_t = P\_t - P\_{t-1}$$ on lagged price changes $$\Delta P\_{t-1}, \Delta P\_{t-2}$$, etc. **Result**: No significant coefficients—past price changes do not predict future changes. This is **exactly** the pattern observed in modern commodity futures markets. **Interpretation**: Babylonian merchants incorporated available information (harvest expectations, war news, inventory levels) into current prices, leaving no easy arbitrage opportunities. **Test 2: Seasonal Patterns** Agricultural prices should have predictable seasonal variation. Post-harvest months (May-June for barley) should see abundant supply and low prices, while pre-harvest months (March-April) should see depleted stocks and high prices. Temin finds exactly this pattern. Barley prices in spring (pre-harvest) averaged 20-30% higher than autumn prices. This spread is consistent with storage costs plus interest rates (the opportunity cost of holding grain), and the pattern held across centuries, suggesting stable seasonal expectations. The implication is striking: Babylonian merchants understood intertemporal arbitrage. They stored grain when cheap (post-harvest) and sold when expensive (pre-harvest), earning returns that covered both storage costs and capital costs. **Test 3: Cointegration with Silver Prices** **Cointegration** is a statistical property where two time series, each individually unpredictable (non-stationary), share a stable long-run relationship. If both barley and silver prices are market-determined, they should be cointegrated—moving together in the long run (purchasing power parity). If the king arbitrarily fixed the barley-silver exchange rate, we would see divergence. **Temin's test**: Do barley and silver prices move independently or together? **Result**: Strong long-run relationship. When silver became scarce (e.g., during wars that disrupted trade), barley prices in silver terms rose. When silver influxes occurred (e.g., tribute from conquered territories), barley prices in silver fell. **Interpretation**: The barley-silver exchange rate was market-determined, not administratively fixed. **Test 4: Response to Shocks** Temin identifies specific historical events and checks whether prices responded as market theory predicts: **Event 1: Death of Alexander the Great (323 BCE)** When Alexander died, his empire collapsed into civil war among his generals (the Diadochi). Fighting blocked grain shipments from Egypt and Syria. A functioning market predicts that barley prices should spike in Babylon due to supply disruption—and they did, tripling within months. Supply shocks transmitted through market mechanisms, not administrative diktat. **Event 2: Seleucid Tax Reforms (c. 290 BCE)** When new Greek rulers imposed higher taxes paid in silver, demand for silver increased, which should raise silver's value and make barley cheaper in silver terms. The data show precisely this pattern. Tax policy affected market prices through supply and demand, not price controls. **Event 3: Harvest Failures** Cuneiform astronomical diaries record weather conditions, allowing Temin to correlate drought years (low rainfall) with price spikes. Drought years show 50-100% price increases, while normal rainfall years show stable prices. This is classical supply-demand economics at work—2,300 years before Adam Smith. #### 5.2.6 Robustness and Alternative Explanations **Could Polanyi still be partially right?** Critics of Temin's interpretation argue: **Objection 1: Palace/Temple Transactions vs. Private Markets** Perhaps the palace did fix prices for *official* transactions, while private markets floated freely. The tablets Temin uses might come from the private sector. **Temin's response**: * He shows that tablets from temple archives (which would be "official") show the same price volatility as private merchant archives * If two parallel markets existed (fixed official, floating private), we would expect arbitrage (people buying from the official market and selling to the private market). No evidence of such arbitrage spreads. **Objection 2: Prices Were Guidelines, Not Binding** Maybe the "customary" ratios (e.g., 1 shekel = 1 gur barley) were default values for contracts, but actual transaction prices varied. **Temin's response**: * This is essentially conceding the market hypothesis—if actual prices varied based on supply/demand, the economy was market-based regardless of what nominal "official" prices said * Modern economies have "list prices" and "transaction prices" too; the latter are what matter **Objection 3: Sample Selection Bias** We have prices only from cities that survived and whose archives were preserved. Might this bias results? **Concern**: Perhaps only market-oriented cities survived (selection on the dependent variable). **Temin's response**: * The tablets span different cities (Babylon, Uruk, Sippar) and different periods (464-72 BCE covers nearly 400 years) * Cities with different political systems (Persian rule, Greek rule, Parthian rule) show similar patterns * The consistency across contexts suggests the pattern is real, not an artifact of selection **Objection 4: Measurement Issues** Ancient measurements varied. A "gur" of barley in one city might differ from another. Exchange rates between silver and commodities might reflect measurement standards, not market prices. **Temin's response**: * He uses prices from single cities with consistent measurement standards over time * When comparing across cities, he adjusts for known differences in measurement units * The volatility and shock-responsiveness can't be explained by measurement changes (unless measurement standards randomly fluctuated with wars and harvests—implausible) **Verdict**: While not all Babylonian economic activity was market-mediated (the palace and temple sectors were large), the private economy operated through market mechanisms recognizable to modern economists. #### 5.2.7 Implications for Economic Thought Temin's findings challenge the conventional narrative in the history of economic thought. > \[!NOTE] **Traditional view**: Markets are a modern invention, emerging with capitalism in the 16th-17th centuries. > > **Temin's revision**: Market mechanisms are ancient—and perhaps universal wherever three conditions hold: private property exists (merchants own goods), exchange is voluntary (not commanded by authorities), and information flows (merchants learn about conditions elsewhere). These conditions obtained in Babylon 2,500 years ago, just as they do in Chicago today. **Modern parallels**: The efficiency of Babylonian barley markets resembles modern commodity futures markets: | Feature | CBOT Wheat Futures (Modern) | Babylonian Barley (464-72 BCE) | | -------------------- | ------------------------------ | ------------------------------------- | | Price predictability | Random walk | Random walk | | Seasonal patterns | Harvest vs. planting cycles | Post-harvest trough, pre-harvest peak | | Shock response | Instant (weather, geopolitics) | Rapid (wars, harvest failures) | The main differences are technical—speed of information (telegraph vs. messenger), contract standardization (futures vs. individual agreements), and regulatory framework (CFTC vs. merchant guilds). The underlying economics—arbitrage, intertemporal optimization, risk management—are identical across 2,500 years. **BOX 4.2: METHOD SPOTLIGHT - Time Series Analysis in Economic History** > **The Challenge**: Ancient price data is messy—irregular observations, measurement errors, data gaps. How do we test whether these prices behaved like modern markets? > > **The Tools**: Econometric time series methods developed for modern financial data can be applied to ancient prices. > > **Method 1: Random Walk Test** > > **Question**: Are prices unpredictable (efficient market) or do they follow patterns that allow profit? > > **Test**: Autocorrelation. Regress price change on lagged price change: $$\Delta P\_t = \alpha + \beta \Delta P\_{t-1} + \epsilon\_t$$ > > * If $$\beta = 0$$: Random walk (efficient) > * If $$\beta \neq 0$$: Predictable patterns (inefficient) > > **Temin's result**: $$\beta \approx 0$$ for Babylonian barley prices > > **Interpretation**: Babylonian merchants arbitraged away predictable patterns > > **Method 2: Seasonality Decomposition** > > **Question**: Do prices have regular seasonal patterns? > > **Approach**: $$P\_t = Trend\_t + Seasonal\_t + Irregular\_t$$ > > Use Fourier analysis or dummy variables to extract the seasonal component. > > **Temin's finding**: Strong seasonal pattern (May-June trough, March-April peak) > > **Interpretation**: This is predictable but not arbitrageable (storage costs eliminate profit) > > **Method 3: Event Study** > > **Question**: Do prices respond to specific historical shocks? > > **Approach**: > > * Identify exogenous event (war, drought, reform) > * Compare prices before vs. after > * Test statistical significance of change > > **Example**: Alexander's death (323 BCE) > > * Pre-event (324-323 BCE): Barley \~ 60 qu of silver per gur > * Post-event (322-321 BCE): Barley \~ 180 qu of silver per gur > * t-test: Difference highly significant (p < 0.01) > > **Interpretation**: Markets responded to supply shock > > **Method 4: Cointegration** > > **Question**: Do related prices move together in the long run? > > **Test**: Engle-Granger cointegration > > * Regress barley price on silver price: $$P\_{barley,t} = \alpha + \beta P\_{silver,t} + \epsilon\_t$$ > * Test whether residuals $$\epsilon\_t$$ are stationary (mean-reverting) > * If yes, prices are cointegrated (long-run relationship) > > **Temin's result**: Barley and silver prices are cointegrated > > **Interpretation**: Long-run purchasing power parity held > > **Challenges in Historical Application**: > > 1. **Irregular spacing**: Tablets aren't evenly spaced in time. Solution: Use monthly/annual averages or time-aggregated methods. > 2. **Small sample**: Modern time series assumes hundreds of observations. Ancient data might have 50-100. Solution: Use robust standard errors, bootstrap confidence intervals. > 3. **Structural breaks**: Regime changes (Persian→Greek→Parthian rule) might shift parameters. Solution: Test for breaks, estimate sub-periods separately. > 4. **Measurement error**: Unit conversions, tablet damage. Solution: Errors-in-variables models, instrumental variables. > > **Why This Matters**: Time series methods let us test modern economic theories against ancient data. If the same patterns emerge across 2,500 years, it suggests fundamental economic principles (arbitrage, optimization, market clearing) are universal, not culturally contingent. *** ### 5.3 The Origins of Finance: Goetzmann (2017) **Full Citation**: Goetzmann, William N. 2017. *Money Changes Everything: How Finance Made Civilization Possible*. Princeton University Press. (Chapter 1-2). #### 5.3.1 The Insight Writing was not invented for poetry—it was invented for **accounting**. The earliest known tablets are ledgers, not literature. Goetzmann demonstrates that fundamental financial technologies already existed in Mesopotamia: * **Time**: Interest (the concept of time value of money). The Sumerian word for interest, *mash*, means "calf" (money breeds money like cattle). * **Risk**: Joint ventures (*tapputum*) allowed merchants to pool capital and share risk for long journeys. * **Liquidity**: Silver served as a medium of exchange, even if not minted as coins yet. These institutions allowed the Assyrian trade network to function over long distances without a single enforcing state. > **Key Finding:** The three pillars of modern finance—interest (time value of money), partnership contracts (risk sharing), and monetary systems (liquidity)—all existed in Mesopotamia by 2000 BCE, nearly four millennia before the financial revolution typically dated to Renaissance Italy. *** #### 5.3.2 The Three Pillars of Ancient Finance Goetzmann argues that finance emerged to solve three fundamental problems that arise whenever humans cooperate economically over time and space. Each problem required a specific innovation. **Pillar 1: The Time Value of Money (Interest)** **The problem**: A farmer needs seed grain in spring but won't harvest until autumn. A merchant wants to buy goods today but won't sell them for a year. How do you transfer value across time? **The innovation**: **Interest** as compensation for delayed consumption. **Evidence from Mesopotamia**: The earliest known interest rates appear in Sumerian tablets from 3000 BCE: * **Barley loans**: 33.3% annual interest (standard rate) * **Silver loans**: 20% annual interest * **Seed loans**: 50% (riskier—harvest could fail) These weren't arbitrary. They reflected: 1. **Opportunity cost**: Lenders forgo consumption today 2. **Risk**: Borrowers might default 3. **Inflation**: Silver value fluctuated **Modern parallel**: Credit cards charge 15-25% APR; mortgages charge 3-7%. The concept is identical—price of time and risk. **Economic function**: Interest allows: * **Consumption smoothing**: Farmers borrow against future harvests * **Investment**: Merchants borrow to finance trade expeditions * **Capital allocation**: Higher returns attract lenders to productive uses **Cultural significance**: The Sumerian etymology is revealing—*mash* (interest) comes from "calf" or "kid" (baby goat). Money "gives birth" to more money, like livestock. This biological metaphor naturalizes finance. Critically, Mesopotamian law capped interest rates (to prevent usury), showing awareness of market power and exploitation—a debate that continues today. **Pillar 2: Risk Sharing (Partnership Contracts)** **The problem**: A trade expedition from Assur to Kanesh (600 miles) takes 6 weeks each way. Risks include: * Bandits * Weather (sandstorms, floods) * Market risk (prices might fall) * Political risk (war, confiscation) No single merchant can afford to lose an entire caravan. How do you spread risk? **The innovation**: The ***tapputum*** (partnership contract) **How it worked** (from actual tablets): 1. **Capital partners** (in Assur): Provide silver or textiles (the capital) 2. **Traveling partner** (goes to Kanesh): Provides labor and expertise 3. **Profit split**: Typically 50-50 after expenses 4. **Risk allocation**: * If goods lost to bandits (force majeure): Capital partner bears loss * If traveling partner cheats or mismanages: Traveling partner liable * Both parties swear oath before gods (reputation enforcement) **Example from tablets**: * Amur-Ištar (Assur) provides 30 minas of silver * Pusu-Ken (traveling) takes silver to Kanesh, buys tin and textiles, sells in Anatolia * Returns with 50 minas silver (20 mina profit) * Each receives 10 minas profit (after 30 mina capital returned) **Modern parallel**: Venture capital operates similarly: * VC provides capital, entrepreneur provides labor * Profits split (VC gets equity share) * Limited liability (VC loses investment if startup fails, entrepreneur loses time) **Economic function**: * **Risk pooling**: Diversification (partner in multiple caravans) * **Specialization**: Capital owners needn't travel; merchants needn't be wealthy * **Incentive alignment**: Profit-sharing aligns interests This is **joint-stock organization** 4,000 years before the Dutch East India Company (1602), traditionally credited as the first joint-stock company. The ability to pool capital and share risk was possible only because Mesopotamia had already developed the surplus economy described in Chapter 3 (Section 2.2)—without agricultural surplus, there would be no capital to invest. **Pillar 3: Liquidity and Payment Systems (Money)** **The problem**: Barter is inefficient. A textile merchant wants tin, but the tin trader wants barley. Finding a "double coincidence of wants" is costly. **The innovation**: **Commodity money** (silver) **Why silver?** Ancient Mesopotamia never had coinage (stamped coins emerged in Lydia c. 600 BCE), but silver functioned as money: 1. **Durable**: Doesn't rot (unlike grain) 2. **Divisible**: Can be cut into small pieces or weighed 3. **Standardized**: Measured by weight (shekel = \~8 grams) 4. **Valuable**: High value-to-weight ratio (portable) 5. **Universally accepted**: Everyone wanted silver **Prices quoted in silver**: * 1 shekel = 1 gur of barley (reference price) * 1 mina (60 shekels) = 1 slave * 1 shekel = 10 days' wages for a laborer **But silver was scarce**: Most people didn't physically transact in silver. How did the monetary system work? **The innovation within the innovation**: **Credit money** Merchants kept accounts. If A sells cloth to B for 5 shekels, the debt is recorded on a tablet. B doesn't hand over silver immediately. Instead: * B owes A 5 shekels (debt recorded) * Later, B sells barley to C for 5 shekels * B assigns his claim on C to A (debt transfer) * No silver changes hands! This is a **credit clearing system**. Silver serves as **unit of account** (prices measured in silver) and **store of value** (debts denominated in silver), but not necessarily **medium of exchange** (actual payments). **Modern parallel**: Most money today is digital—bank deposits, not physical cash. We use dollars as unit of account, but actual transactions are credit transfers. **Evidence**: The tablets record debts, not payments. Scholars estimate that 90% of transactions were credit-based, with periodic settlement in silver. #### 5.3.3 Beyond the Basics: Babylonian Financial Sophistication The three pillars described above—interest, partnership, and money—represent foundational innovations. But Babylonian financial practice went considerably further, developing instruments that bear a striking resemblance to modern derivatives and insurance contracts. The Old Babylonian period (c. 2000-1600 BCE) produced thousands of tablets documenting increasingly specialized financial arrangements that Goetzmann argues reveal a remarkably mature understanding of risk, time, and contingency. **Grain loans** were among the most common financial instruments, and their terms reveal sophisticated calibration to risk. A standard barley loan carried an interest rate of 33.3% per annum (one-third of the principal), while silver loans typically charged 20%. This differential was not arbitrary: grain loans were inherently riskier because repayment depended on the harvest, which could fail due to drought, flood, or locusts. The higher rate on grain compensated lenders for agricultural risk—precisely the logic that modern banks use when charging higher rates on unsecured loans than on mortgages. Some tablets record **variable-rate structures**: a lower rate if repaid at harvest, a higher penalty rate if repayment was delayed. Others specify that interest would be calculated in barley but payable in silver at the prevailing exchange rate, effectively embedding a **currency conversion clause** within the loan contract. More remarkable still were contracts that functioned as **proto-insurance** and **proto-futures**. Bottomry loans—contracts in which a lender advanced capital for a trading voyage and accepted the loss if the ship sank, in exchange for a premium above the normal interest rate—transferred maritime risk from the borrower to the lender. This is, in economic substance, an insurance contract bundled with a loan: the lender accepts the downside risk in exchange for a higher expected return. Similarly, tablets from the reign of Hammurabi document **forward delivery contracts** in which a merchant agreed to deliver a specified quantity of grain, sesame, or wool at a future date for a price fixed at the time of the agreement. These are functionally identical to modern commodity futures: they allow both parties to hedge against price fluctuations. The buyer locks in a price, protecting against a rise; the seller locks in a buyer, protecting against a fall. That Babylonian merchants arrived at this solution independently—without formal training in finance theory—testifies to the universality of the economic problems these instruments address. #### 5.3.4 Why Finance Enabled Civilization Goetzmann's central thesis: Finance is not a consequence of civilization—it is a **cause**. **The argument**: 1. **Cities require surplus extraction**: Farmers produce food; others (priests, soldiers, artisans) consume it. This requires transferring food across time (storage) and space (trade). 2. **Storage and trade require finance**: Storage is effectively lending (a farmer advances grain to the city, to be repaid later); trade is effectively credit (a merchant takes goods on credit and pays after sale). 3. **Large-scale cooperation requires trust**: Finance creates reputational mechanisms. Default on debts, and no one lends to you again. This enforces cooperation beyond kinship circles. 4. **Finance scales**: Family reciprocity works for small groups (20-50 people). Finance works for cities (10,000-50,000 people). **Evidence from history**: * **Uruk (3000 BCE)**: First true city (pop. 40,000). Coincides with first writing (accounting tablets) * **Pyramids (2500 BCE)**: Required coordinating 10,000+ workers over 20 years. Financed through grain taxation and redistribution (=debt) * **Roman Empire**: Financed army with tax bonds, grain shipments with bottomry loans (=insurance) **Counterfactual**: Societies without finance stayed small. Hunter-gatherers (no finance) had bands of 50-150. Early agricultural villages (limited finance) capped at 500-1,000. Cities (with finance) reached 100,000+. #### 5.3.5 Limitations and Critiques While Goetzmann's framework is compelling, critics note important caveats: **Critique 1: Confusing correlation with causation** Yes, finance and cities emerged together. But which caused which? * **Alternative view**: Cities created demand for finance (reverse causality) * **Goetzmann's response**: Experimental archaeology shows that small-scale finance (debts, gifts with expected reciprocation) predates cities. Finance scales up with urbanization. **Critique 2: Non-financial mechanisms also mattered** Finance wasn't the only coordination technology: * **Kinship**: Extended family networks enforced cooperation * **Religion**: Gods punished oath-breakers (supernatural enforcement) * **State coercion**: Kings had armies; they could compel cooperation **Goetzmann's response**: These are complements, not substitutes. Finance worked *through* kinship (family firms), religion (oath-taking), and states (legal enforcement). But finance provided the *incentive structure*. **Critique 3: Finance also enabled exploitation** The same tools that enabled trade also enabled: * **Debt bondage**: Farmers who couldn't repay loans became slaves * **Usury**: Lenders charging extractive rates * **Inequality**: Wealth concentration among financiers Mesopotamian law codes (Hammurabi, Lipit-Ishtar) repeatedly tried to regulate interest rates and cancel debts, suggesting finance created social problems. **Goetzmann's acknowledgment**: Finance is a tool. Like fire, it can warm or burn. The question is institutional—how does society regulate finance to maximize benefits and minimize harms? #### 5.3.6 Modern Implications: Are We Still Using Bronze Age Finance? Remarkably, the financial instruments of 2000 BCE resemble those of 2025 CE: | Ancient Mesopotamia | Modern Equivalent | | ------------------------------------- | ----------------------------- | | Silver loans at 20% interest | Bank loans at 5-15% interest | | *Tapputum* partnership | Venture capital, LLPs | | Debt tablets | Promissory notes, bonds | | Credit clearing | Bank clearing systems, ACH | | Grain futures (delivery next harvest) | CME grain futures | | Bottomry loans (ship insurance) | Marine insurance, derivatives | **What's different?** 1. **Scale**: Ancient finance served city-states; modern finance is global 2. **Abstraction**: Ancient = physical silver; modern = fiat money 3. **Regulation**: Modern finance has central banks, securities law, deposit insurance 4. **Complexity**: Ancient = simple contracts; modern = structured products, CDOs, credit default swaps **What's the same?** 1. **Core functions**: Time-shifting, risk-pooling, liquidity provision 2. **Incentive problems**: Moral hazard, adverse selection, principal-agent 3. **Boom-bust cycles**: Mesopotamia had debt crises (debt jubilees); we have 2008 financial crisis **The lesson**: Finance is ancient because the problems it solves are fundamental to human cooperation. Any society trying to coordinate economic activity across time and space must invent something like interest, partnerships, and money. **The warning**: Finance also creates risks—debt traps, inequality, crashes. Mesopotamian kings enacted debt forgiveness ("andurarum") every 7-30 years to prevent debt bondage. Modern societies use bankruptcy law, central banks, and regulation. The specifics change; the challenge persists. *** ## 6. Synthesis and Debates ### 6.1 The "Modernist" Consensus The cumulative evidence from prices (Temin) and trade volumes (Barjamovic) confirms that ancient people were **economically rational agents** operating in market settings. They responded to incentives, arbitraged price differences, and optimized the location and timing of their transactions. ### 6.2 Limits of the Market However, these were not modern capitalist economies. * **Factor Markets**: Land and labor were often restricted (not fully commoditized). * **State Role**: The Palace and Temple were huge players, consuming a large fraction of GDP. * **Growth**: Despite this sophistication, there was little **intensive growth** (per capita income growth). It was a "Smithian" growth (gains from trade) but not "Schumpeterian" (technological) growth. ### 6.3 Integration: What Do the Three Papers Tell Us Together? The three papers in this chapter—Barjamovic et al. (trade networks), Temin (prices), and Goetzmann (finance)—collectively paint a coherent picture of Bronze Age economics. **Layer 1: Microeconomic Behavior (Temin)** * Individuals responded to incentives * Prices incorporated information efficiently * Arbitrage and optimization operated **Layer 2: Network Structure (Barjamovic et al.)** * Trade followed predictable patterns (gravity model) * Distance mattered systematically * Networks were economically rational, not just political **Layer 3: Institutional Foundation (Goetzmann)** * Finance provided the infrastructure for layers 1 and 2 * Interest rates enabled intertemporal trade * Partnerships enabled risk-sharing across space * Credit money enabled exchange **The synthesis**: Ancient economies were **complex adaptive systems** where: * Micro-level optimization (price-taking merchants) generated macro-level patterns (gravity trade flows) * Institutions (finance) emerged endogenously to solve coordination problems * The whole system was self-organizing, not centrally planned **Modern parallel**: This resembles Adam Smith's "invisible hand"—decentralized actors pursuing self-interest generate efficient aggregate outcomes. But it required institutional scaffolding (property rights, contract enforcement, money). *** ### 6.4 The Big Debate: Polanyi vs. Formalists These papers contribute to a long-running debate in economic anthropology. **Karl Polanyi's position** (*The Great Transformation*, 1944): * Ancient economies were "embedded" in social relations * Exchange was reciprocal (gifts) or redistributive (through kings), not market-based * The "market economy" is a modern aberration (emerged only with 19th-century capitalism) * Applying economic theory to ancient societies is anachronistic ("formalist fallacy") **The formalist response** (exemplified by Temin): * Economic principles (scarcity, optimization, arbitrage) are universal * Ancient people faced the same trade-offs we do * Markets existed whenever exchange was voluntary and prices flexible * The evidence (tablets) shows market behavior **Who's right?** The modern consensus: **Both capture partial truths**. **Polanyi was right that**: * Ancient economies had large non-market sectors (palace/temple redistribution) * Social norms constrained economic behavior (debt forgiveness, interest caps) * Kinship and religion enforced contracts, not just legal institutions **But Polanyi was wrong that**: * Markets didn't exist or were unimportant * Economic analysis doesn't apply * Market behavior is modern **The formalists were right that**: * Markets were widespread in the ancient world * Prices responded to supply/demand * Economic theory illuminates ancient behavior **But formalists sometimes underemphasize**: * Cultural context matters for how markets function * Non-market institutions (kinship, religion) complemented markets * Distribution of power affected market outcomes (not just efficiency) > **Synthesis:** Markets and non-market institutions **coexisted**. The palace redistributed grain to workers; merchants traded silver for tin in competitive markets. Some labor was coerced (slaves); other labor was hired at daily wages. The productive question is not "market or not?" but "what mix of market and non-market, and why did that mix vary across time and place?" ### 6.5 Methodological Lessons: What Can Economists Learn From Clay Tablets? This chapter demonstrates several methodological innovations that apply beyond ancient history. **Lesson 1: Theory Can Recover Missing Data** Barjamovic et al.'s use of the gravity model to find lost cities exemplifies **structural estimation**: using economic theory to infer unobservable variables from observable outcomes. **Applications elsewhere**: * **Missing prices**: If we know quantities traded and have a demand curve, we can infer the price * **Missing technology**: If we know inputs, outputs, and production function shape, we can infer technology levels * **Missing institutions**: If we know outcomes and behavioral responses, we can infer institutional quality The key requirement: **Theory strong enough to provide tight predictions**. **Lesson 2: Textual Data Can Be Quantified** Historians traditionally use texts qualitatively (narrative analysis). Economists treat texts as datasets. **Barjamovic et al's innovation**: Counting co-occurrences of city names on tablets → trade volume proxy **Similar approaches**: * **Google Books Ngram**: Track word frequency to measure cultural change * **Congressional Record text analysis**: Measure polarization from speech patterns * **Medieval account books**: Extract prices, wages, trade flows The principle: If a text contains repeated, structured information, it can be quantified. **Lesson 3: Time Series Methods Work on Ancient Data** Temin's application of random walk tests, cointegration, and event studies to Babylonian prices shows that econometric methods developed for modern finance can apply to ancient data—if you account for limitations (small samples, irregular spacing, measurement error). **The broader point**: Don't let data imperfections paralyze you. Use robust methods, acknowledge limitations, triangulate across multiple tests. **Lesson 4: Interdisciplinary Collaboration Creates Value** Barjamovic et al.'s team included: * **An Assyriologist** (Barjamovic): Read cuneiform, identify cities * **Trade economists** (Chaney, Coşar): Build gravity model * **An applied econometrician** (Hortaçsu): Estimate model No single discipline could have done this. The archaeologist lacked the gravity model. The economists couldn't read cuneiform. **The lesson**: Modern economic history increasingly requires teams with diverse skills—historians, economists, data scientists, archaeologists. *** ### 6.6 Contemporary Relevance: What Do Bronze Age Markets Tell Us About Modern Development? The findings in this chapter carry direct implications for current development debates. **Implication 1: Markets Are Not a Recent Western Invention** Development discourse sometimes frames "markets" as foreign imports that must be grafted onto traditional societies. The evidence from Mesopotamia shows markets are ancient and widespread. **Policy relevance**: Don't assume developing countries lack market-compatible culture. Informal markets operate everywhere. The question is how to formalize, regulate, and integrate them. **Implication 2: Finance Precedes Industrialization** Goetzmann shows finance is 5,000 years old. This challenges the view that finance is a product of modernity. **Policy relevance**: Financial development shouldn't wait until countries are rich. Basic finance (credit, insurance, payment systems) can accelerate growth. Microfinance, mobile banking, and village savings groups are rediscovering ancient principles. **Implication 3: Institutions Matter, But Can Emerge Endogenously** The Assyrian trade network operated without a centralized state enforcing contracts. Instead, reputation, kinship, and religious oaths provided enforcement. **Policy relevance**: Top-down institution-building (imposing laws, creating courts) isn't the only path. Bottom-up institutions (merchant guilds, community enforcement, digital reputation systems like eBay ratings) can substitute. **Implication 4: Information Frictions Are Ancient** Temin's finding that Babylonian prices followed a random walk suggests information was efficiently incorporated despite slow communication (no telegraph, just messengers on donkeys). **Policy relevance**: Modern developing countries have much faster information technology (mobile phones). If Bronze Age merchants could achieve market efficiency, modern information barriers are less about technology and more about institutions (transparency, contract enforcement). **Implication 5: The Limits of Markets Without Growth** Bronze Age markets were sophisticated but didn't generate sustained per capita growth. Why? **Hypotheses**: * No technological innovation (tools unchanged for centuries) * Limited factor mobility (land and labor not fully marketable) * Extractive institutions (surplus extracted by elites, not reinvested) **Policy relevance**: Markets are necessary but not sufficient for growth. Need also: * Innovation incentives (patents, R\&D) * Human capital (education) * Inclusive institutions (broad property rights, not just elite) *** **BOX 4.3: SCHOLAR PROFILE - Gojko Barjamovic** > Gojko Barjamovic is a Senior Lecturer on Assyriology in the Department of Near Eastern Languages and Civilizations at Harvard University. Born in Denmark and trained in the European tradition of ancient Near Eastern philology, Barjamovic is one of the world's leading experts on the Old Assyrian trade archives—the very tablets that form the evidentiary backbone of this chapter. His scholarly career has been devoted to reading, translating, and interpreting the cuneiform records of the merchants who operated between Assur and Kanesh four millennia ago. He is the author of *A Historical Geography of Anatolia in the Old Assyrian Colony Period* (2011), a painstaking reconstruction of the trade routes, cities, and geography of Bronze Age Anatolia based on decades of engagement with the primary texts. > > What makes Barjamovic's contribution to economic history distinctive is the **unusual interdisciplinary collaboration** that produced the 2019 *Quarterly Journal of Economics* paper. His co-authors—Thomas Chaney (international trade theorist), Kerem Cosar (trade economist), and Ali Hortacsu (applied econometrician)—are economists with no training in cuneiform. Barjamovic, for his part, is a philologist with no background in structural estimation or gravity models. The collaboration emerged from a chance encounter at the University of Chicago, where Hortacsu was a faculty member and Barjamovic was visiting. The economists were looking for historical trade data to test gravity models; Barjamovic had spent years compiling exactly such data from cuneiform sources. Neither side could have produced the paper alone. The economists lacked the linguistic expertise to extract trade data from ancient tablets; the Assyriologist lacked the econometric framework to transform that data into geographic predictions. > > The result—using economic theory to locate lost cities—demonstrates a broader lesson about the practice of economic history: **the most innovative work often occurs at the boundaries between disciplines**. Barjamovic brought deep contextual knowledge: which city names were synonyms, which tablets were reliable, which trade references were metaphorical rather than literal. The economists brought formal models that could extract spatial information from relational data. The combination produced insights that neither discipline could have generated independently. This pattern of interdisciplinary collaboration—historians providing context and data, economists providing analytical frameworks—recurs throughout this textbook (see Chapter 3 on genetics and economics, Chapter 5 on archaeology and economics, and Chapter 12 on climate science and economics). > > Barjamovic continues to work at the intersection of Assyriology and economics, exploring how cuneiform archives can illuminate questions about market structure, institutional development, and the organization of long-distance trade in the ancient world. His work reminds us that the raw material of economic history is not always a spreadsheet—sometimes it is a clay tablet that requires years of philological training to decipher. *** > \[!TIP] **Chapter Summary** > > * The Bronze Age (3000-1200 BCE) produced surprisingly sophisticated economies: the **Old Assyrian Trade Network** operated through private merchants, joint-stock partnerships, credit instruments, and fluctuating market prices -- overturning Karl Polanyi's claim that ancient economies lacked true markets. > * **Barjamovic et al. (2019)** used a **structural gravity model** applied to 23,000 cuneiform tablets from Kanesh to predict the locations of 11 lost Bronze Age cities, with several predictions since archaeologically confirmed -- demonstrating that economic theory can be "inverted" to recover missing historical data. > * **Temin (2002)** applied modern time-series econometrics to Babylonian barley prices (464-72 BCE) and found random walk behavior, seasonal patterns, and rapid responses to supply shocks (e.g., prices tripled after Alexander the Great's death) -- all hallmarks of efficient, market-driven pricing that decisively reject the Polanyi "fixed price" hypothesis. > * **Goetzmann (2017)** demonstrated that the three pillars of modern finance -- interest (time value of money), partnership contracts (risk sharing), and monetary systems (liquidity) -- all existed in Mesopotamia by 2000 BCE, nearly four millennia before the financial revolution typically dated to Renaissance Italy. > * The **Late Bronze Age Collapse** (c. 1200 BCE) destroyed these trade networks in a cascading systems failure -- the Hittite Empire fell, Mycenaean Greece lost literacy for 400 years, and cities like Ugarit were burned and abandoned -- illustrating how interconnected trade systems create both prosperity and systemic fragility. > * Key methods introduced include **structural estimation** (inverting gravity equations to solve for unknown locations from known trade flows), **time-series market efficiency tests** (random walk, cointegration, event studies), and the use of **proxy variables** (tablet co-occurrences as proxies for trade volume). > * The central takeaway is that market institutions, financial instruments, and long-distance trade are not modern inventions but recurring features of complex societies -- yet they proved fragile, requiring centuries to rebuild after the Bronze Age Collapse, challenging any narrative of smooth, cumulative economic progress. *** ## 7. Looking Forward ### 7.1 Connection to Chapter 5: Greece and Rome The first civilizations built complex economies with trade and finance, but they were fragile. The Bronze Age ended in a cataclysmic collapse around 1200 BCE: trade networks severed, cities burned, and writing disappeared from Greece for four centuries. The next chapter travels west to the civilizations that rose from those ashes: **Greece and Rome**. It asks how the Roman Empire managed to scale this economic complexity to cover the entire Mediterranean, achieving living standards not matched until the 18th century. The tools introduced here—gravity models, time series analysis, structural estimation—will be applied to Roman data, while Chapter 5 adds new proxy variables (shipwrecks, ice cores, skeletal evidence) to the empirical toolkit. **Key connections**: * **From city-states to empire**: The Assyrian trade network connected a dozen cities. The Roman Empire integrated hundreds of cities across three continents. How did they achieve this scale? * **From commodity money to coinage**: Mesopotamia used weighed silver. Greece and Rome invented coined money (stamped, standardized). How did this innovation change market efficiency? * **Market integration at scale**: Temin's methods (price convergence, random walk tests) will be applied to Roman wheat prices across the Mediterranean. We'll see whether a 50-million-person empire achieved market integration comparable to Bronze Age Mesopotamia's regional networks. * **Legal infrastructure**: Assyrian merchants relied on kinship and reputation. Rome developed legal institutions (courts, property law, commercial law). Did formal law substitute for informal enforcement, or complement it? ### 7.2 Methodological Evolution Through the Book The tools introduced in this chapter—gravity models, time series analysis, structural estimation—will recur and develop throughout the textbook. **In Chapter 6 (China and Japan)**: * **Gravity models** applied to Chinese internal trade (Shiue & Keller 2007) * **Market integration tests** comparing Qing China to contemporary Europe * **Question**: Did China's massive size create integration or fragmentation? **In Chapter 8 (Medieval Europe)**: * **Event studies** of the Black Death's economic impact * **Time series analysis** of wages and prices over centuries * **Structural breaks** testing whether the Black Death permanently changed economic structures **In Chapter 10 (Industrial Revolution)**: * **Growth accounting**: Decomposing output growth into capital, labor, and technology contributions * **Structural transformation**: Modeling the shift from agriculture to industry * **Trade patterns**: How did the gravity model change with railroads and steamships? **In Chapter 12 (Great Depression)**: * **Panel data methods**: Tracking regions over time * **Continuous treatment DiD**: Using erosion intensity (Dust Bowl) as a continuous variable rather than binary treatment The methodological progression: 1. **Chapter 4 (this chapter)**: Introduce structural models and time series basics 2. **Chapters 5-9**: Apply and refine these methods in different contexts 3. **Chapters 10-13**: Integrate all methods with modern causal inference techniques ### 7.3 The Collapse Question This chapter focused on Bronze Age prosperity. But we should acknowledge what comes next: **collapse**. As Section 2.5 detailed, around 1200 BCE the Bronze Age world system disintegrated with terrifying speed. The Assyrian trade network vanished. Major cities (Hattusa, Mycenae, Ugarit) were destroyed. Trade routes were severed. Writing disappeared from Greece for 400 years (the "Greek Dark Ages"). The financial instruments documented by Goetzmann, the partnership contracts, the credit clearing systems—all of it was swept away in a generation. **Theories of collapse**: 1. **Sea Peoples invasion**: Mysterious maritime raiders 2. **Systems collapse**: Cascading failures in interconnected trade networks 3. **Climate change**: Drought destabilized agricultural base 4. **Internal conflict**: Class warfare between elites and commoners The **systems collapse** theory, as discussed in Section 2.5, carries the most weight for economic historians because it highlights a structural vulnerability inherent in any highly integrated trade network. The Bronze Age economy was, in important respects, a **proto-globalized** system. Tin traveled thousands of miles from Central Asian mines to Mediterranean workshops. Grain crossed political boundaries to feed allied populations. Luxury goods circulated among elites as diplomatic currency. This interdependence generated enormous gains from specialization and trade—the very gains that the gravity model captures—but it also meant that the system had no redundancy. When tin supplies from the east were disrupted, there was no alternative source. When grain shipments from Egypt ceased, the Hittites faced famine. When maritime trade routes became unsafe, coastal cities lost their economic raison d'etre. The economic concept that best captures this vulnerability is **systemic risk**—the risk that failure in one component of an interconnected system triggers cascading failures throughout. Modern financial regulators worry about systemic risk in banking (the failure of Lehman Brothers in 2008 nearly brought down the global financial system). Supply chain analysts worry about systemic risk in manufacturing (the 2011 Tohoku earthquake in Japan disrupted automobile production worldwide). The Bronze Age collapse suggests that systemic risk is not a modern phenomenon—it is an inherent feature of any complex, interdependent economic network. **Modern parallel**: The parallels to contemporary supply chain fragility are striking and uncomfortable. Global supply chains for semiconductors, rare earth minerals, pharmaceuticals, and agricultural inputs are concentrated in ways that create single points of failure. The COVID-19 pandemic demonstrated this vividly: lockdowns in one country halted production of components needed in dozens of others. The Russia-Ukraine war disrupted global grain and energy markets. U.S.-China tensions threaten the semiconductor supply chain on which virtually all modern technology depends. In each case, the underlying dynamic mirrors the Bronze Age pattern: **specialization generates efficiency but also fragility**. The policy question—how to balance the gains from trade integration against the risks of systemic vulnerability—has no easy answer. Autarky (self-sufficiency) sacrifices the enormous benefits of specialization. Full integration accepts the risk of cascading collapse. Most economists advocate for **resilient integration**: maintaining the benefits of trade while building redundancy into critical supply chains (diversifying suppliers, maintaining strategic reserves, investing in domestic capacity for essential goods). The Bronze Age collapse suggests that civilizations which fail to manage this trade-off may not get a second chance. **Where we'll return to this**: Chapter 12 examines the Great Depression as a modern "collapse" of integrated trade and financial networks (1929-1933). The Smoot-Hawley tariff and retaliatory trade barriers shattered the global trading system much as the Bronze Age disruptions severed ancient trade routes. The mechanisms differ, but the pattern—rapid unraveling of complex economic systems built on mutual dependence—echoes across four millennia. ### 7.4 From Ancient to Medieval: The Long Divergence The transition from Chapter 5 (Rome) to Part II (Medieval and Early Modern) involves a profound question: **Why did the ancient world's economic sophistication disappear for a millennium?** * **Bronze Age**: Market prices, long-distance trade, financial instruments * **Dark Ages (500-1000 CE)**: Local subsistence, limited trade, natural economy * **High Middle Ages (1000-1300)**: Markets and trade gradually return * **Renaissance (1300-1600)**: Finance re-emerges (Italian merchant banks) The "Dark Ages" saw enormous economic regression in Europe. Per capita income may have fallen 50%. Cities shrank. Trade collapsed. Literacy plummeted. **But**: * **China** maintained and even exceeded Bronze Age sophistication (Song Dynasty 960-1279 had paper money, credit markets, industrial production) * **Islamic World** (700-1200) preserved and extended ancient knowledge (banking innovations, algebra, global trade) * **Byzantine Empire** (330-1453) continued Roman economic institutions **The puzzle**: Why did economic regression occur in Western Europe but not elsewhere? **We'll explore this in Part II**, examining: * **Chapter 6**: How China and Japan maintained market integration * **Chapter 7**: How Islamic finance innovated beyond ancient systems * **Chapter 8**: How medieval Europe slowly rebuilt from fragmentation * **Chapter 9**: How the printing press and Reformation broke through medieval stagnation The very long arc: Mesopotamia invents finance → Rome scales it → Europe loses it → China/Islam preserve it → Europe reinvents it (and adds gunpowder, printing, and eventually industrialization). *** ## Chapter Summary **Key Takeaways**: 1. **Ancient Markets**: Mesopotamia had functioning markets with fluctuating prices, interest rates, and complex contracts. 2. **Gravity Model**: The laws of trade (Interaction $$\propto$$ Size / Distance) held 4,000 years ago. 3. **Structural Discovery**: Economic models can be inverted to find missing historical data (lost cities). 4. **Finance**: Financial instruments (loans, futures) predate coinage and were essential for long-distance trade. **Empirical Methods Learned**: * **Structural Estimation**: Using a theoretical model to estimate unobserved parameters (location). * **Time Series Analysis**: Testing for random walks and seasonality in price data. *** ## Discussion Questions 1. **Universality of economic laws**: If the gravity model works so well for Bronze Age trade (predicting city locations from trade patterns), what does that imply about the "universality" of economic laws? *Contemporary application*: The gravity model also predicts modern trade, migration, and even social media connections. Does this mean human behavior is fundamentally constant across time, or are we fitting flexible models that work anywhere? How would you test which interpretation is correct? 2. **Political economy of price controls**: Why might an ancient king *want* to fix prices (as Polanyi argued), even if the market data shows he couldn't enforce them? *Modern parallel*: Many governments impose price controls (rent control in NYC, price caps on gasoline in Venezuela, interest rate ceilings in Islamic banking). These often fail to stick, yet politicians keep trying. Why? Are there situations where price controls actually work (e.g., wartime rationing)? 3. **Economics meets archaeology**: How does the discovery of lost cities using trade data change your view of the relationship between economics and archaeology? Should archaeologists learn economics? Should economists learn archaeology? *Contemporary example*: Economists now use satellite imagery and machine learning to predict archaeological sites, measure historical deforestation, and estimate ancient population densities. Is this interdisciplinary synergy, or are economists colonizing other fields? 4. **Contract enforcement without the state**: The Assyrian merchants traded over huge distances without a police force to protect them. How did they enforce contracts? (Hint: Reputation and family ties). *Modern application*: eBay, Airbnb, and Uber rely on reputation systems (seller ratings, reviews) rather than courts to enforce quality. Is this a return to ancient mechanisms? What are the advantages and disadvantages compared to formal legal enforcement? Could global trade operate without state enforcement? 5. **Market efficiency then and now**: Temin shows that Babylonian barley prices followed a random walk, indicating market efficiency. But these were pre-modern merchants with donkeys, not computers. *Puzzle*: How were Bronze Age markets as informationally efficient as modern futures markets with instant global communication? Does this mean modern technology adds less value than we think, or were ancient merchants extraordinarily sophisticated? 6. **Finance and civilization**: Goetzmann argues that finance (interest, partnerships, money) made civilization possible by enabling large-scale cooperation. *Counterfactual*: Could cities of 50,000 people exist without finance? *Contemporary debate*: Some argue modern finance has become extractive rather than productive (2008 crisis, high-frequency trading). Has finance gone from enabling civilization to destabilizing it? Or is this a false dichotomy? 7. **Debt and jubilees**: Mesopotamian kings periodically canceled debts (debt jubilees) to prevent debt bondage and social unrest. *Modern parallels*: Student loan forgiveness debates in the U.S., sovereign debt relief for developing countries, personal bankruptcy laws. Are these modern equivalents of ancient jubilees? When is debt forgiveness economically efficient vs. a moral hazard? 8. **The Polanyi debate**: Section 6.4 discusses whether ancient economies were "market" or "embedded in social relations." *Contemporary version*: Are modern economies purely market-based, or are they also embedded (firms as social communities, gift economies in open-source software, family obligations)? Is the market vs. embedded dichotomy useful, or a false choice? 9. **Structural estimation**: Barjamovic et al. use economic theory (gravity model) to infer missing data (city locations). What other historical unknowns could be recovered this way? *Brainstorm*: Could we estimate ancient GDP from archaeological data? Infer population from city sizes? Reconstruct medieval trade routes from pottery fragments? What are the limits of this approach? 10. **Fragility of complex systems**: The Bronze Age trade network collapsed around 1200 BCE, causing widespread economic regression. *Modern parallel*: Are global supply chains similarly fragile? *Contemporary shocks*: COVID-19 disrupted chips, PPE, and vaccines. Russia-Ukraine war disrupted grain and energy. U.S.-China tensions threaten tech supply chains. Should we prioritize efficiency (global integration) or resilience (local production)? Is deglobalization inevitable? 11. **Why no ancient growth?**: Bronze Age economies had markets, finance, and trade—yet no sustained per capita income growth for millennia. Modern economies with similar institutions grew rapidly. *Puzzle*: What's the difference? *Hypotheses*: Technology stagnation, extractive elites, lack of human capital investment, energy constraints (no fossil fuels). Which explanation is most convincing? Can we test these hypotheses? 12. **From tablets to blockchain**: Ancient Mesopotamia used clay tablets to record debts and transactions. These tablets were immutable (once baked, can't be altered) and distributed (copies in multiple archives). *Modern parallel*: Blockchain technology (Bitcoin, Ethereum, smart contracts) uses immutable distributed ledgers. Are cryptocurrencies reinventing Bronze Age finance? What's genuinely new vs. old wine in new bottles? 13. **Time series on messy data**: Temin applied modern econometric methods (random walk tests, cointegration) to ancient data with huge imperfections (gaps, measurement error, small samples). *Methodological question*: When is it appropriate to use sophisticated statistical methods on poor-quality data? Could he have been fooled by noise? How much robustness testing is enough? 14. **Markets and morality**: The ancient world had usury laws (interest rate caps), debt jubilees, and norms against exploitation. Modern economics often treats market outcomes as presumptively efficient. *Normative question*: Should we evaluate ancient economies by modern efficiency standards, or by their own moral frameworks? Is there a "moral economy" separate from market efficiency? How do we navigate this tension in policy? 15. **The Polanyi debate and non-Western economies**: Section 2.3 describes how Polanyi argued that applying market logic to ancient economies is anachronistic. *Contemporary application*: Development economists sometimes debate whether Western-style market institutions are appropriate for non-Western societies. Does the evidence from Mesopotamia—showing market behavior in a non-Western, pre-modern context—settle this question? Or could Polanyi's insight still apply to societies where reciprocity and redistribution dominate? How should development policy account for economies where markets coexist with strong non-market institutions (e.g., gift economies in Pacific Island societies, communal land tenure in sub-Saharan Africa)? 16. **The fragility of Bronze Age trade networks**: Section 2.5 describes how the interconnected Bronze Age trade system collapsed around 1200 BCE through cascading failures. *Analytical question*: Using the concept of systemic risk, explain why the same features that made the Bronze Age trade network productive (specialization, long-distance exchange, interdependence) also made it fragile. Can you identify a modern supply chain that exhibits similar structural vulnerabilities? What policies could reduce systemic risk without sacrificing the gains from trade? Is there a "optimal level" of economic integration, or is this a false trade-off? 17. **Applying the Barjamovic method to other periods**: Barjamovic et al. used co-occurrence data from cuneiform tablets and the gravity model to locate lost cities. *Creative exercise*: Identify another historical period where geographic information is missing but trade data (or proxies for trade) might exist. How would you adapt the gravity model approach? What would serve as your "tablets"—archaeological finds, literary references, genetic data, isotopic analysis of traded goods? What are the minimum data requirements, and what could go wrong? Consider, for example, the Phoenician trading network, medieval trans-Saharan trade routes, or pre-Columbian obsidian exchange networks. *** ## Data Exercises ### Exercise 1: The Gravity of Trade **Data**: `anccities_trade.csv` (simplified). **Task**: 1. Calculate the distance between city pairs using coordinates. 2. Regress `log(trade_count)` on `log(distance)`. 3. Check the coefficient on distance. Is it close to -1 (standard gravity result)? ### Exercise 2: Babylonian Prices **Data**: Synthetic dataset of Barley Prices. **Task**: 1. Plot the price series over time. 2. Identify the "seasonal" component (average price by month). 3. Test if the price today predicts the price tomorrow (Auto-correlation). ### Exercise 3: Finding a Lost City with Gravity **Data**: Construct the following hypothetical dataset of trade flows (number of co-occurring tablet mentions) and distances between five Bronze Age cities. Four cities have known locations; one ("City E") is unknown. | City Pair | Trade Flow (co-occurrences) | Distance (km) | | --------- | --------------------------- | ------------- | | A - B | 120 | 150 | | A - C | 45 | 400 | | A - D | 80 | 250 | | B - C | 60 | 300 | | B - D | 30 | 350 | | C - D | 90 | 200 | | A - E | 100 | ? | | B - E | 50 | ? | | C - E | 20 | ? | | D - E | 70 | ? | **Known coordinates**: A = (37.0N, 35.0E), B = (38.2N, 35.8E), C = (39.5N, 33.5E), D = (37.8N, 34.0E). **Task**: 1. Using the known city pairs (A-B, A-C, A-D, B-C, B-D, C-D), estimate the gravity model: regress `log(trade)` on `log(distance)` to obtain the distance elasticity parameter $$\theta$$. 2. Using your estimated $$\theta$$, invert the gravity equation to predict the distance from City E to each known city. That is, for each pair involving E, solve: $$D\_{iE} = \left(\frac{A \cdot M\_i \cdot M\_E}{T\_{iE}}\right)^{1/\theta}$$. (Assume all cities have equal economic size for simplicity, so the size terms cancel.) 3. Using the predicted distances from E to A, B, C, and D, trilaterate City E's approximate location on a map. Plot the known cities and your predicted location for City E. 4. *Reflection*: How sensitive is your prediction to the estimated $$\theta$$? Re-estimate with $$\theta \pm 0.5$$ and report how much City E's predicted location shifts. What does this tell you about the precision of the Barjamovic et al. method? --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://laurence-wilse-samson.gitbook.io/textbooks/world-economic-history/part-i-foundations-pre-history-and-antiquity/chapter04_mesopotamia.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.