by Ober, Josiah
By most indications, the Early Iron Age nadir around 1000 BCE was an exceptionally poor era. That has considerable implications for thinking about the classical efflorescence. On one side, the low EIA bottom left a great deal of room for improvement: Simple reversion to the mean of premodern normal population and consumption, as represented by, say, the Late Bronze Age and Greece in 1830, would represent major progress. On the other hand, the economy of classical Hellas did not have much of a springboard from which to jump off.
The premodern Greek normal represented by the LBA and the early nineteenth century CE was considerably above the EIA nadir—but very far below the classical efflorescence. Demographically, the premodern normal was a population of fewer than a million people for the Greek core of the mainland south of Macedonia and Epirus, along with the Aegean and Ionian Sea islands—the territory defined by the late nineteenth century Greek state. By the last decade of the nineteenth century, the population of core Greece had reached 2.3 million. Yet median non-elite per capita consumption hovered uncomfortably near subsistence—which is, as we will see, the ordinary fate of premodern non-elite populations elsewhere in the world. Greeks responded to Malthusian pressure by emigration.
During the classical efflorescence, Greece was much more densely populated than the premodern norm. In the later fourth century BCE, the population of the part of Hellas occupied by the Greek state in 1890 (ignoring the extended Greek polis world to the west, north, east, and south) was in the range of 2.75 million (my lower bound estimate) to 3.5 million people (Mogens H. Hansen’s upper bound, with discussion below). The population density of modern Greece did not equal that of the extended Greek world in the later fourth century until the 1920s (figure 4.1).
Even the low estimate of classical era population is several times the premodern normal. Moreover, during the same period, there is every reason to believe that, at least in the most highly developed ancient Greek states, many, perhaps most non-elite people were consuming well above the level of subsistence. By the later fourth century BCE, per capita consumption was well above the premodern normal and about twice what it had been 500 years earlier. In sharp contrast to the situation in Greece of the late nineteenth century CE, in the late fourth century BCE there was no evidence of a Malthusian trap: Development was not being constrained by limited resources, and there was no need to export surplus labor through emigration. Indeed, in the era of Plato and Aristotle, Hellas was a net importer of labor (slaves), food, and raw materials—and a net exporter of manufactured goods, cultural products, and expert services. In a word, Hellas, in the era of the classical efflorescence, was wealthy when compared to any point in Greek history before the twentieth century.6
FIGURE 4.1 Comparative Greek demographics, ancient and modern.
NOTES: Modern data adapted from Country National Time Series (CNTS) Greece data (Banks and Wilson 2013). Normalized 0–1 scale. 1 = twice the classical peak.
URBANIZATION: Classical peak for Greek world = 32% of total population in towns of >5,000. CNTS data for percentage of population living in cities of more than 10,000 is available for 1919 to 1978; cities of more than 20,000 from 1850. “Over 5,000” figures estimated for 1919–1978 by adding the difference between the 10,000 and 20,000 figures (median 4.7%) to the 10,000 figure; estimate for 1850 doubles the median difference (9.4%) and adds that to the over-20,000 figure. Density: Classical peak for Greek world = 44 per km2. Modern density = CNTS population divided by area. Population: Classical peak for core Greece = 2.75 million (Ober baseline estimate; Hansen 2006b, 2008 is higher). CNTS population is truncated at 1912 due to discontinuity in 1913. CNTS does not include data from 1914 to 1918.
OTHER STANDARDS OF COMPARISON
The notion that classical Hellas was wealthy comes as a surprise. Among the assumptions about the classical Greek world with which I grew up as a student of ancient history in the 1970s, was that the world of the Greek city-states was poor. This assumption was based in the first instance on what we can call the standard ancient premise. That premise is stated succinctly by the “Father of History”—the fifth century BCE Greek historian Herodotus: “Hellas has always had poverty as its companion.” (7.102.1). In Herodotus’ history, this statement is put into the mouth of Demaratus, a deposed king of Sparta, in the context of a conversation with Xerxes, the Great King of Persia. It has been cited frequently by later historians of the Greek world. Herodotus’ quotable line, along with other passages in Greek authors to a similar effect, contributed to the formation of a standard modern premise. In the pungent early twentieth century prose of the British classical historian and political scientist Alfred Zimmern, “the pioneers who created our European civilization were stricken with poverty all of their days … it was the doom of Athens that Poverty and Impossibility dwelt in her midst from first to last.”7
Along with the familiar claim that it was the Greeks who pioneered European civilization, Zimmern’s comment is notable for its assumption of Athenian exemplarity: For Zimmern, “Athens” stands for “Greece.” How exemplary or exceptional Athens really was, and how much we can extrapolate from Athenian economic performance to the wider Greek economy, remain important questions to which we will return in this and in subsequent Chapters. In the imperial fifth century, but also in the postimperial fourth century, Athens was certainly among the most prosperous of the Greek poleis. For our present purposes, it suffices to say that if classical Athens really was impoverished, as Zimmern claimed, then it is highly likely that the rest of the Greek world fared poorly. If Athens was relatively wealthy, it is reasonable to suppose that residents of the most developed states of the Greek world were quite well off.8
It has long been recognized that the poverty claim in the standard ancient premise must be understood in comparative rather than absolute terms. In Herodotus’ key passage, Demaratus compares the quotidian life of ordinary citizens in Greece (and especially Sparta) with the court of Xerxes, the Great King of Persia. When the lifestyle of the King of Persia is the standard, everyone in classical Greece was certainly comparatively poor. But, equally obviously, when we are attempting to compare the overall performance of ancient economies, “ordinary Greeks vs. court of Xerxes” is the wrong standard of comparison.9
I have suggested above that the right standard of comparison is the premodern Greek normal. But an alternative standard of comparison for the ancient Greek economy might be sought among the most advanced economies of the nineteenth to twenty-first centuries of our era. If Xerxes is self-evidently the wrong comparison for, say, ordinary-citizen Tellus of Athens (later in this chapter), then how about middle-class American John Doe? When we compare median income (in standardized dollars) or consumption (measured, e.g., in per capita energy use) in ancient Greece with the United States or with the most developed countries of western Europe in the nineteenth to twenty-first centuries, the Greek world, once again, looks impoverished.10
As in the case of Demaratus of Sparta’s pointed contrast of ordinary Greeks to ultra-elite Persians, comparing ancient Greek and modern developed-country levels of income or consumption tells us something—but nothing that we do not already know. It is hardly news that median levels of consumption in contemporary developed economies are much higher than median consumption levels in any premodern economy: Tellus of Athens did not shop at Walmart. It is certainly worth investigating just how much higher modern consumption levels actually are. But using contemporary developed-country economies as a standard of comparison is not particularly informative if we are seeking to learn something new about the relative economic performance of premodern societies. Neither the Tellus/Xerxes nor the Tellus/John Doe comparison gives a satisfactory answer to the question, “Are the standard ancient and modern premises right?”
Rather than seeking to determine how much poorer Tellus was than Xerxes or John Doe, when we ask, “Are the standard ancient and modern premises concerning Greek poverty right?” what we want to know is how much better or worse off the ordinary
Greek in the classical period was, when compared to the ordinary individual in the Greek premodern normal situation discussed above, or in other relatively well documented premodern societies. While we cannot answer the comparative welfare question directly, we can provide answers to questions that will in turn give us an indirect way to approach the question of comparative economic performance:
Was the rate of classical Greek economic growth high or low, relative to other premodern economies?
Was the classical Greek world more or less densely populated, and more or less urbanized, than other premodern societies?
Was the distribution of wealth and income across the classical Greek population relatively more or less equitable than that of other premodern populations? What part of the classical Greek population lived at a level high enough above subsistence to qualify as at least minimally decent?
Using these questions as proxies for investigating comparative general welfare, the answer to the question, “Was classical Greece impoverished?” is “obviously not”—Greeks in the age of Plato and Aristotle were not poor when compared to the Greek premodern norm or to people in other ancient or medieval societies. The Greek economy grew briskly in the period 800–300 BCE. By the later classical period, the Greek world was densely populated and highly urbanized. A high percentage of Greeks (or at least of Athenians) lived comfortably above the near-subsistence level of consumption that has been the economic fate of most people for most of human history. By the standards of other premodern economies, Hellas was wealthy. Moreover, Despite the Demaratus–Xerxes exchange and similar passages in the Histories, there is reason to think that Herodotus (and by extrapolation, his original classical Greek readers) knew it.
The “Eastern monarch discusses comparative welfare with a wise Greek” motif of the Demaratus–Xerxes interchange is anticipated in a scene near the beginning of Herodotus’ Histories in which King Croesus of Lydia (in western Anatolia) interviews Solon of Athens. The imagined date of this (perhaps imaginary) interview would have been some time in the sixth century BCE. The subject of their conversation is human happiness, and the context is relative wealth. Croesus expects Solon to acknowledge that Croesus is outstandingly happy on the basis of his superabundant wealth. But to Croesus’ surprise, Solon instead names as the happiest person ever to have lived Tellus of Athens, “who came from a prosperous city … and the circumstances of his life were likewise prosperous, by our standards” (Herodotus, Histories 1.30.4–5).
Tellus is not portrayed by Herodotus’ Solon as a member of a privileged elite pulling down big rents as a result of his violence potential in a natural state. Quite to the contrary, he is depicted as a reasonably but not exceptionally well-to-do, Greek citizen of a reasonably well-to-do Greek polis. Tellus was, in Solon’s pithy account, an ordinary Greek man who had been especially fortunate in his progeny (healthy and excellent children and grandchildren) and in his demise (timely, heroic death in victorious battle in defense of his homeland). Reasonably prosperous material conditions constituted a necessary precondition for Tellus’ exceptional happiness, but the clear implication of the story is that many of Tellus’ fellow-citizens enjoyed a similar prosperity. It was Tellus’ relative advantages in respect to his descendants and death, not his wealth, that made Tellus’ life an especially happy one. It is, however, the background material conditions of that life that concern us here: Tellus and his large and healthy family evidently lived comfortably above the level of bare subsistence, in a community in which living at that level was not regarded as remarkable.
For our purposes, the important point is this: If, as Herodotus’ story implies, the material conditions enjoyed by Tellus and his family were fairly typical, then a decent level of income (i.e., enough to live well above subsistence) must have been fairly common in Athens, and by extrapolation, in other developed Greek states. A society featuring a substantial body of ordinary people, living in decent conditions, consuming well above the level of bare subsistence, would be, by ancient standards, an exceptionally wealthy society. Of course Croesus was vastly richer than Tellus. But, if we think about Lydia as an autocratic state, and the sociopolitical conditions sustaining the great wealth of Croesus and his court, we will have reason to doubt that many non-elite Lydians lived at Tellus’ moderately prosperous level. The figure of “ordinary, relatively prosperous Tellus” was not merely a Herodotean fiction. As we will see, some 40–60% of classical-era Athenian residents (and perhaps of Greeks generally) fit a profile of Tellus-like moderate prosperity. Although a few premodern societies, notably Mesopotamian cities in sixth century BCE Babylonia, paid wages high enough to allow the presumption of quite extensive non-elite prosperity, there are very few documented cases of high-wage societies before the early modern period.11
We still lack the detailed studies of the economies of ancient Middle Eastern societies that would allow meaningful comparisons to the economy of Hellas. But it seems very unlikely, on the face of it, that the Lydian empire, or the Persian Empire, or indeed any other ancient empire substantially outperformed the Roman Empire of the first and second centuries CE in terms of per capita consumption or urbanization. We now have reasonable estimates of Roman imperial economic performance. Based on those estimates, Rome appears, by ancient standards, to have been an exceptionally prosperous and urbanized empire. The Roman economy was much bigger, and much more diverse, than the Greek economy. Comparisons can be misleading, but, by certain measures (aggregate and per capita economic growth, urbanization, and income distribution), the overall Greek economy of ca. 500–300 BCE appears to have outperformed the overall Roman economy of ca. 100 BCE–200 CE.12
If it is true that Rome economically outperformed other ancient empires and that Hellas performed well relative to Rome, then it is fair to say that the society of Hellas was, in fact, relatively wealthy by ancient standards. Moreover, certain features of the Greek economy (or at least the Athenian economy) of the fifth and fourth centuries BCE compare favorably with the most advanced premodern European economies—Holland and England in the fifteenth to eighteenth centuries CE.13
The explanation I advance in chapter 5 for the comparative wealth of Hellas in the classical era is, in the first instance, political and institutional. The point of the comparisons is to falsify the standard ancient and modern premises about Hellenic poverty, to highlight the role of politics in Greek social and economic development, and thus to begin to solve the puzzle of the remarkable efflorescence of classical Hellas.
Here are three premises that, if correct, give us good reasons to believe that classical Hellas was indeed comparatively wealthy:
Premise 1
The Greek economy grew steeply and steadily from 1000 to 300 BCE, both in its aggregate size and in per capita consumption.
Premise 2
By the fourth century BCE, Greece was densely populated and remarkably urbanized, yet living standards remained high.
Premise 3
Wealth was distributed relatively equitably across Greek populations; there was a substantial middle class of people who lived well above bare subsistence yet below the level of elite consumption.
The three premises, in the form of descriptive statements about the Greek economy, are based on economic models that are in turn based on extensive collections of empirical evidence for ancient Mediterranean and early modern European economies.
The rest of this chapter offers evidence that tests, and ultimately supports, the validity of each of the three premises. In the next chapter (ch. 5), drawing on the theory of decentralized cooperation developed in chapter 3, I offer two hypotheses to explain how and why the Greek economy performed comparatively well. The historical narrative that follows (chs. 6–11) tests those hypotheses.
HIGH AGGREGATE AND PER CAPITA GROWTH
My Stanford colleague, the archaeologist and historian Ian Morris, has assembled an impressive array of data for measuring Greek economic growth in the period 800–300 BCE.14 The first factor to consider i
n measuring Greek economic growth is demographic change. It is uncontroversial to state that the population of Hellas grew substantially in the half-millennium 800–300 BCE. On the basis of J. K. Beloch’s late nineteenth century surveys of literary evidence, supplemented by recent work in survey and excavation archaeology, Morris posits that the population of “the Aegean and the colonies in southern Italy and Sicily” rose from under 500,000 people in the ninth century to perhaps 4 million people in the fourth century. If this is correct, the Greek population of this part of the world increased about tenfold and the per annum demographic growth rate was over 0.4%. As Morris points out, this is a comparatively high rate of sustained demographic growth in a premodern society.15
Morris’ figures are only estimates, but in order for Morris’ posited demographic growth rate to be much too low, we would have to assume that the population of the Aegean/Italian–Sicilian Greek world in 800 BCE was much larger than 500,000, or that in 300 BCE the relevant parts of the Greek world had a population much less than 4 million. Neither counterfactual is plausible: For the early period, archaeologists have expended a great deal of time and effort searching for and analyzing sites from the Greek Early Iron Age (also known as the Dark Age), and they have done their best to show that the Dark Age was not so dark as all that (ch. 6). Despite their best efforts, known Early Iron Age Greek occupation sites remain comparatively sparse and small; as noted above, the nadir of Greek population, around 1000 BCE, was probably about 330,000.