INCOME INEQUALITY, POLITICAL EQUALITY, AND TAXATION IN LATE-CLASSICAL ATHENS

Josiah Ober

Stanford University

Introduction

Since the time of August Boeckh ancient Athens has been recognized as a remarkably well-documented example of a pre-modern democratic society.1This is a revised version of a paper originally published as Ober 2017. My thanks to Sven Günther, Dorothea Rohde, Wolfgang Franzen and the other participants in the Boeckh-conference, and three anonymous readers for helpful comments that have led to revisions in the original paper.Athens was not a liberal democracy: there was a large (see below) population of chattel slaves; political participation was limited to adult males; and relatively long-term non-native residents were only rarely granted full citizenship, with its concomitant political participation rights. But Athens was, by pre-modern standards, strikingly politically egalitarian in that virtually all native adult males held full participation rights. There was no property qualification for citizenship. Non-elite citizens participated actively in Athenian legislative institutions, served as jurors on law courts, and held administrative offices. Equal access to institutions for all citizens was ensured by pay for public service and the frequent use of lotteries for service on the council, on juries, and for various offices. Athenian public discourse emphasized equality of voice (isegoria), of vote (isopsephia), and before the law (isonomia). Along with liberty, a commitment to political equality was held to be characteristic of the democratic regime, by ancient democrats and their elite critics alike.2Athenian democracy: Hansen 1999; democracy values and democracy’s critics: Ober 1998. An abbreviated version of some parts of this paper appears in Ober 2016, section 2. My thanks to Walter Scheidel, Claire Taylor, Mark Pyzyk, and Nikolaos Kyriazis for discussion of the issues in this paper.

It is also widely recognized, however, that the Athenians were not practicing egalitarians in respect to the distribution of material goods. Wealth and income were distributed unequally across the resident population; some Athenians were rich, some poor, many others somewhere in the middle. So the general question is this: did Athenian democracy, with its focus on political equality, have any meaningful impact on Athenian (or, more broadly, ancient Greek) economic inequality? Put in other terms: is there reason for contemporary students of economic history to be surprised at the level of Athenian economic inequality, when it is compared to other pre-modern societies? If so, is the surprise occasioned by higher or lower than expected levels of inequality? Was Athens less unequal than other, less democratic, pre-modern states? If so, can democracy plausibly be postulated as a cause of the difference? And if democracy is a cause, what is the mechanism through which the effect is realized? I will suggest that late-classical Athens did in fact experience historically low inequality, caused in part by democracy, through the mechanism of progressive taxation of the wealthy. Progressive taxation lowered the income of the wealthiest. At the other end of the spectrum, tax revenues were (in part) redistributed to poorer Athenians, through pay for military service, pay for civic service, and other welfare distributions.3For recent, contrasting views of and approaches to economic inequality in classical Greece, see Roubineau 2015, arguing that every Greek polis was grounded in deep social status inequalities, resulting in the impoverishment of many; Patriquin 2015, arguing for minimal inequality among citizens at Athens, because democracy could only arise in a society with only a small gap between rich and poor. Neither attempts the sort of modeling offered here.

In other recent work4Ober 2008; 2015a.I have argued that (1) there is reason to believe that late classical Athens experienced relatively low levels of economic inequality along with high levels of economic growth; (2) although Athens was especially prosperous, Athens is not an outlier in the world of the Greek city states: many Greek poleis, and the classical Greek world generally, were remarkably prosperous; (3) prosperity and relatively low levels of economic inequality were related, and both were fostered by historically distinctive Greek citizen-centered political and legal institutions; (4) democracy is an especially strong version of the general Greek phenomenon of the citizen-centered regime, and thus democracy may be expected to have had a particularly strong effect on prosperity and inequality. If these claims are correct, the ancient Athenian (and generally the ancient Greek) case has considerable bearing on questions that lie at the center of debates among social scientists,5E.g. Acemoglu and Robinson 2012; Boix 2015.concerning political economy and development.

Political development is often associated with changes in the level of economic inequality. Carles Boix (2015) argues, following a broad consensus among historical anthropologists, that very small “stateless” societies of foragers are typically very equal, in both political and material terms. The development of agriculture, radical increases in social scale and complexity, and the beginning of urbanization subsequently led to profound levels of political and economic inequality, which persisted, Boix argues, in most of the world until the Industrial Revolution. Boix’s theory is that economic development (agriculture, much later industrialization) drives political development (centralized monarchies, much later democracy), rather than the other way around. He notes, however, that the culture of the ancient Greek city-states was in some ways exceptional in featuring strong forms of political republicanism and relatively robust growth long before the Industrial Revolution.6Boix 2015, 134-139, 204-205.Boix ultimately dismisses the Greek case as a meaningful alternative path to political and economic development, claiming, “perhaps most decisively, the short duration [italics added] of most republics meant that any economic advantage they might have had could not be consolidated and that it disappeared with their conquest at the hands of a monarchical power - Hellenistic Greece at the hands of Rome, late medieval Italy at the hands of France and Spain.”7Ibid., 88. The notion that the Hellenistic kingdoms were republics and the Roman state a monarchy at the time of the Roman conquest is strange, to say the least.On the other side of the democracy and development debate, Daron Acemoglu and James Robinson (2016) have pointed to the ancient Greek city states as an example of effective “bottom up” political development leading to robust civil society and thus paving the way for economic growth.

Economic growth in the ancient Greek world was strong and sustained, by pre- modern standards.8In addition to Ober 2015b, recent and important books addressing Greek economic growth in the classical era include Acton 2014; Bresson 2015; Harris, Lewis and Woolmer 2015. Taylor 2017 addresses the question of economic inequality in classical Athens.From 800 BC (the end of the “dark” period of the Early Iron Age, thus discounting the “post collapse recovery era” of 1000-800 BC) to 200 BC (the Roman takeover) the Greek world experienced 600 years of impressive economic performance. Within the world of the Greek city-states, there was certainly considerable regional variation in growth rates. Overall growth during the 600-year period undoubtedly masks periods of sharp downturn. Moreover, it is certainly true that ancient Greek growth was much lower than modern growth (by several orders of magnitude) and that it did not last indefinitely. But, on the other hand, it remains the case that the Industrial Age is not much more than 250 years old. So it seems wrong, or at least premature, to say that the Greek economic efflorescence was of “short duration” compared to industrial era modernity.

The evidence for relatively high levels of political equality among native males in many (although not all) Greek city-states is well enough known that I need not rehearse it here. Political equality among native males was especially strongly associated with Greek democracy, and democracy is by far best documented for classical Athens. As I have argued in detail9Ober 2008.Athens was in some ways exceptional as a city-state (very big, very high-performing, with institutions that proved robust to exogenous shock). But in other ways Athens is exemplary of the “citizen centered” and prosperous classical Greek city-states. This is important, insofar as our evidence for the distribution of wealth and income is heavily concentrated in Athens.

A model of Athenian population and income, 330 BC

Economists typically assess material inequality by measuring income. We cannot hope to be precise in our measure of income across the population of even the best documented of the Greek city-states. But I believe that there is currently enough information available to allow for plausible guesses of income distribution for Athens in the late classical era. The postulated date of the model is 330 BC, in the so-called “Lycurgan” era during which Athenian state income and general prosperity was at a high point. This date has the advantage of being after the end of the Athenian Empire (an era in which the Athenian economy was very unusual by Greek standards), after the economy had recovered from the Peloponnesian War, and before the conquests of Alexander fundamentally remade the Greek world.

The model of Athenian population and income is based on 34 groups, each with a hypothetical size and level of income. The model is laid out in detail in the Appendix Table. This model is a variant of a considerably simpler model I proposed in earlier work.10Ober 2015a, Table 4.4.While the general conclusions in respect to levels of income inequality of the two models are similar, the new model assumes a somewhat smaller total population of sub-adult citizens and a substantially larger population of liturgical class (very wealthy) Athenians. The new model attempts to add some more nuance but, like the older one, it necessarily simplifies a much more complex social reality, inter alia by assuming that within-group household incomes are equal.

The population and income model proposed here makes a number of assumptions that are more or less well supported by available empirical evidence. Other historians of classical antiquity, using the same evidence, might make somewhat different assumptions about the population of Athens, its distribution across groups of persons, household size, and the estimated income per household in each group. I have attempted to keep my assumptions about population and income well inside what I take to be the general range within which most estimates by professional Greek historians are likely to fall. In any event, the attempt at modeling allows others interested in the questions raised here to point out where their assumptions align with or differ from mine, and to build their own population and income models accordingly.11Van Wees 2011 discusses a well-known passage in Ath. 6.272c, that records the following figures for Athens’ population, some time after 317 BC: Athenaioi: 21,000; metoikoi: 10,000; oiketai: 400,000. Van Wees’ interpretation of the passage is at odds with all previous scholarship. Van Wees claims that the passage (as it has come down in the manuscript tradition) accurately records a careful census of residents of Athens, putatively conducted by Demetrius of Phaleron. He proposes that the metoikoi include many disenfranchised Athenians (see below). He also proposes that the figure for oiketai includes about 100,000 native women and children, the rest being slaves. None of this seems plausible: inter alia, a total population of 431,000 would yield a per square km density of over 170 persons, which is much too high for Attica, in an era when Athens was no longer an imperial capital or the unique center of Aegean trade. Since Athenian agricultural productivity was limited, in order to feed such a population (ca. 10 slaves to every free household), the non-agricultural sector of the economy of Athens would have to be considerably more productive than any historian antiquity has ever dared suggest (so that food grown outside Attica could be purchased and imported in vast quantities). Extrapolating these figures to other regions of Greece (given that there is no reason to assume that Athens post-317 BC was vastly more densely populated) would yield population figures far higher than the highest estimates proposed by ancient historians. At least for oiketai, the Athenaeus-figure must be regarded as wrong, as most previous scholarship on the passage assumed. The other figures in the passage, even if they were accurate for post-317 BC, need have little bearing on the figures in 330 BC, given the dramatic changes in Athens’ fortunes after 322 BC.

Population

226,703 total in 123,170 households. The population modeled is the total of long-term residents living in Athens’ home territory of Attica in 330 BC. The three primary categories of long-term residents are slaves, free foreigners (metoikoi / metics), and citizens.12These are very general categories; for finer-grained distinctions among Athenian categories of persons (in law and social relations), see Kamen 2013.Some residents are assumed to live in family households, normalized as 4.5 individuals (husband, wife, 2.5 other dependent kin: children under age 18 or other family-members, e.g. unmarried or widowed adult women). Others are assumed to live in one-person households; i.e. not married, no children, no other dependents. “Household” is used here in a highly stylized way, for purposes of demographic counts, rather than as a description of actual living conditions. Thus, all slaves are counted as individual (one-person) households, rather than being included in the households of their masters, although I certainly do not suppose that most slaves lived alone. Some slaves certainly lived in the houses of their masters; other slaves (we do not know how many) did live independently, apart from their masters, and some of these certainly did have families. Yet others may have lived in barracks of some kind, but no archaeological evidence of such arrangements have been identified.13Ancient demography: Scheidel 2007; Holleran and Pudsey 2011. On the invisibility of Athenian slaves in the archaeological record, see Morris 1998b.

Slaves. 80,000 total in 80,000 households: 35% of Athenian population. The category of slaves is the most uncertain in terms of numbers; estimates of Athens’ slave population have ranged from a low of about 20,000 to a high of several hundreds of thousands. Among the problems in estimating slave numbers is the likelihood that no Athenian ever knew how many slaves were resident in Attica. There was no reason for the state to collect that information, and no way for any private individual to know it. The total figure of ca. 80,000 (of all ages and both sexes) has been used for a generation in population estimates, although it rests on no ancient authority. In terms of distribution, I artificially assume that no slaves (even those living apart from their masters) live with their own families. I assume that most slaves are adults (i.e., there are few “house-born” slaves). I do not make any assumption about the distribution by gender. Adding a number of subadult slaves in slave households would increase the total estimated population, but it would not much affect the general model.14Cartledge 1985, 35, seems reasonable: 60,000-100,000 “according to the most cautious estimates.” Kamen in progress, ch. 1, n. 37 surveys a dozen respected classical scholars’ estimates of slave numbers in classical Athens. Estimates are for various dates in the late fifth to late fourth century, and are sometimes stated in terms of maxima or minima. For what it is worth, the mean of the estimates comes to 84,400; the median is 78,000. For the low count: Jones 1964. High count: Taylor 2001; van Wees 2011. For the background conditions that made Greek and Roman slavery profitable, see Scheidel 2005; 2008; Silver 2006.

Metics. 31,633 total in 12,835 households: 14% of Athenian population. Metics included both migrants (Greek and non-Greek) who chose to live in Attica, presumably often for economic reasons, their descendants, and manumitted slaves. Unlike slaves, the Athenian state did keep a record of long-term resident free non-citizens, each of whom (male and female adults) was required to pay an annual poll tax to the state. One ancient passage (Ath. 6.272c: see above n. 6) claims that there were 10,000 (presumably adult) metics in Athens in 317 BC, during the oligarchy established by Demetrius of Phaleron under Macedonian auspices. I have assumed that in 330 BC there were more adult metics than that. The Athenian economy in 330 BC was stronger than it was later, and the democratic regime more welcoming to foreign residents. Moreover, I assume that in addition to the “official metics” who paid that poll tax (and thus were available to be counted in an official list) there were some “undocumented” metics who did not pay the tax and so could not be officially counted. I assume that some metics lived alone (i.e. in households of 1, slaves aside). Others will have lived in family households with children and perhaps some adult family-member dependents.15Metics in Athens with discussion of possible numbers: Whitehead 1977; Kamen 2013; Akrigg 2015. Van Wees 2011, 104, argues that even 10,000 metics is “implausibly high,” based on an overreading of Thuc. 2.31.2. Van Wees 2011 uses the Thucydides-passage to produce a citizen-to-metic ratio of 5:1. That ratio is in turn imagined to be constant from the fifth through the late fourth century BC. The argument for the ratio is strained, and its assumed invariance over time, is implausible. Metoikia as an institution outside Athens: Whitehead 1984.

Citizens. 115,070 total in 30,337 households: 51% of Athenian population. The citizen population has received the most attention from ancient historians. Mogens Herman Hansen (1986; 1988; 2006) has done especially impressive work on the number of Athenian citizens in the fourth century; my assumptions make only minor alterations in his model, which is predicated on the assumption that there were at least 30,000 adult male citizens and on standard life tables developed by demographers for pre-modern societies. Like Hansen, I assume a roughly equal number of citizen males and females. The question of the distribution of individuals across age categories is complex. Hansen’s “mortality level 4 growth rate 0.5%” model was based on a comparison with demographic estimates of the population structure of the high Roman Empire; the basic work was done a generation ago by the Roman historian, Keith Hopkins, and similar assumptions are still being made by subsequent Roman demographers.16See Akrigg 2011.Roman demographers in turn borrowed from assumptions about European populations before 1800. Hansen’s choice of demographic model seems certainly to be within the right general range, but the Athenian population may have varied from the model in certain ways.

A key question is the ratio of adults (age 18 and over) to children, which in turn depends of life expectancy and average number of live births per woman. The rate of infant/young child mortality has an especially strong effect on the structure of the overall population. Hansen’s model results in males aged 18-80 being 57.4% of the total male citizen population. My earlier model left adult males as only 44.4% of the total male population, implying that there would have been many more children in citizen households.17See Ober 2015.That figure resulted from what I now regard as an overly-strong simplifying assumption, i.e., that each adult citizen headed of a family of 4.5 persons. That obviously misrepresents the reality, in that a male citizen of, say, 18-20, was unlikely to be married with 2.5 children or other family-member dependents. Likewise, some older male citizens would no longer be heads of households supporting children under 18. In the new model I have attempted to correct for this by postulating that there were 6,125 unmarried citizen men over age 18 (group 11: “citizen 2”).

For citizens with families, I think we may allow marginally lower childhood mortality than in Hansen’s model, on the assumption that the life tables he used average across subpopulations, at least some of which may be assumed to outperform and others to underperform the median. So, insofar as the model is based on the demography of the Roman Empire, we may reasonably assume that some parts of the empire (i.e. certain ancient populations) diverged somewhat from the norm. My model assumes a date in the third quarter of the fourth century, which appears to be near the peak of the classical Greek eラorescence. As we will see, Anna Lagia (2015, below) has offered new evidence that suggests that classical Greek (and specifically Athenian) nutrition was overall better than it was during the Roman era. In sum, I am assuming economic conditions that appear to be more favorable than at any point in Greek history before the 20th century. The upshot is an assumption that there were just over 30,000 adult male citizens, and that this represented 52.7% of all citizen males (compared with Hansen’s 57.4%). The total number of citizens of both genders and all ages comes to just over 115,000.

Income

7,179.6 Talents total per annum

The basic units of income are the drachma (dr: a silver coin with standard weight of 4.3 grams) and the talent (T: 6000 drachmas). All income figures are pretax. Per above, metics paid a poll tax (at the rate of 12 dr per annum for a male, 6 dr per annum for a female). There was no standard tax on income, but elite Athenians paid periodic property taxes (eisphora) and other taxes in the form of mandatory liturgies and donations (epidoseis) to state projects.18On Athenian taxation, see Lyttkens 2012; Ober 2015b; Fawcett 2016.

Because we know enough about the cost of grain, income in the form of drachmas per day can be converted to standard per diem “wheat wages,” and then into multiples of bare subsistence.19See Scheidel 2010.Estimates of income are based on the following basic assumptions:20Discussed in more detail in Ober 2015, 91-97.to survive each individual must consume at or above the level of bare subsistence (S). When income is in the form of daily wages, I assume 300 working days per year. Adult male “head of household” is the primary source of income for family households, but adult women and other family members are assumed to bring in some income.

Wage earners (including slaves). Groups 1-14. I assume that the great majority of Athenians made their living from income that was in the form of, or can be imagined as, wages. Of course not all Athenians actually received wages: many received some or all of their income from working privately-held or leased agricultural land. Where we have the evidence of actual wages in later fourth century Athens, these were relatively high when compared to the normal wage paid to unskilled laborers in other pre-modern societies. Documentary evidence suggests that, in the later fourth century, unskilled labor by adult males was (at least sometimes: the evidence remains anecdotal) paid at the rate of 1.5 dr per day; social status (slave, metic, citizen) does not appear to affect pay rates. Skilled laborers were (at least sometimes) paid at higher levels, up to 2.5 dr per day. When converted into wheat wage equivalents, and combined with the assumption of at least minimal income by other family members this yields a high family income: 3.7 to 4.6 times above the level of bare subsistence (i.e. 3.7-4.6 S). These Athenian pay rates are comparable to pay rates in early modern Holland and much higher than other known ancient and medieval societies, which average around 1.6 S.21See further Scheidel 2010; Allen 2009.The income model assumes that some adult male residents were paid below an unskilled labor rate of 1.5 dr noted above, and that others, assumed to be skilled laborers, were paid above those rates.

The income of slaves is modeled in the form of wages. The wages of slaves in the model are assumed to be the portion of pay that is left to the slave, to support his/her consumption and savings, after the master has extracted the bulk of the slave’s wages. I assume that some slaves were left with very little beyond bare subsistence (group 1), that others were left with somewhat more (groups 2 and 3), and that some skilled slaves were left with “take home” income comparable to that of an unskilled laborer (group 8). We do not actually know what part of a slave’s income was typically retained by the slave, but we do know that Athenian slaves were frequently manumitted and that manumission at least sometimes involved a financial transfer of what can be imagined as slave savings to the master in a sort of “balloon payment,” whereby the slave in effect purchased himself. Slave to master manumission-time wealth transfer is at least is one reason that Athenian income inequality is likely to be substantially lower than wealth inequality.22Manumission: Zelnick-Abramovitz 2005; 2009; Akrigg 2015.

Elites. Groups 15-34. The high end of the income distribution model consists of 5,265 people in 1,170 households; 835 of these households are citizens, 335 are metics. These high-income groups constitute 2.32% of Athenian population. I assume that household income in these high-end groups includes various nonwage sources, such that an estimate of daily wages is no longer the relevant unit of income; rather income ought to be thought of in terms of a return on invested wealth. We know that in Athens households with substantial wealth, generally estimated in the range of 3-4 T, were subject to paying direct taxes on wealth (eisphora and occasional liturgies, trierarchic and festival). In this model I have accepted the argument that the liturgical class included approximately 1,200 households (metic and citizen), rather than the estimate of 400 households that I used in my earlier model.23The lower figure of ca. 400 liturgical estates: Davies 1971; the higher figure of ca. 1,200 liturgical estates: Hansen 1999, 110-115. Due to the simplifying assumption of equal within-group income, the cut off between groups 14 (citizen 5: annual household income 720 dr) and 25 (citizen 6: annual household income 1000 dr) is artificial. Ancient sources referring to 1,200 citizens in the liturgical class (on which see Hansen, above) can be accommodated by assuming that a few hundred of the wealthiest group 14 households were subject to paying liturgical taxes.

Assuming a return on wealth (whether in the form of real estate or other assets) in the range of 5-8% per annum, a household in the bottom rank of the highincome tier (groups 15, 25), with an estimated income of 1000 dr per annum, might possess total wealth in the range of 2-3.5 T. Households in higher-income groups (16-24, 26-34) would have comparably greater wealth. I assume that the lower of the high-income groups contained relatively more households. There are only a handful of metic and citizen households in the very wealthiest groups (24, 34), with estimated annual incomes of 40,000 dr. Each of these ultra-high-income households might have possessed total wealth in the range of 100 T. Estimates of the income of the highest-income groups can, however, be little more than guesses, in part because of the issue of hidden wealth. I assume that many elite families sought to hide (i.e., render “invisible” and thus untaxable) a substantial part of their wealth and that some were successful in doing so.

Discussion

Based on the hypothetical model of population and income sketched above, and detailed in the Appendix Table, we can calculate various figures relevant to the question of the extent of income inequality in late classical Athens. These figures are, of course, hypothetical, in that they are based on the assumptions built into the model. Once again, it is important to remember that other historians could make different assumptions, and thus end up with different figures.

Gini Coefficient of Inequality and Inequality Extraction Ratio. The (pretax) income Gini coefficient for the whole of Athenian society (including slaves and resident foreigners) comes to 38.24The Gini coefficient is a standard measure of inequality, widely used in the social sciences; the higher the number (the closer to 100), the greater the inequality in the population in question. Here what is measured (via the hypothetical model) is income; in a perfectly unequal society (Gini: 100), all of the income would be earned by one person (or income group); this is obviously impossible in practice, so the question is how far the population in question diverges from that level. The derived Gini is “Gini1,” which estimates between-group inequality, assuming within-group inequality to be zero. See Milanovic, Lindert and Williamson 2011, 260 for the distinction between Gini1 and Gini2; as they note, the difference is very small for most of the historical societies in their study group.This is marginally higher than the “Roman

The “optimistic” model in Ober 2015b: Table 4.4 yielded an income Gini of 40. Empire in AD 14” Gini of 36.4,25As estimated by Milanovic, Lindert, and Williams 2011; see Table 1, below.but it is lower than the Roman Empire Gini of 42-44, as estimated by the ancient economic historians, Walter Scheidel and Steven Friesen.26See Scheidel and Friesen 2009.The Lorenz curve for Athens is illustrated in Figure 1. It is quite different in shape from the curve illustrated by Scheidel and Friesen for Rome. The difference arises from the substantially larger Athenian population of persons who fall in the middle range, between the richest and poorest. The estimated Athenian income Gini is substantially lower than that estimated by Branko Milianovic, Peter Lindert and Jeffrey Williamson for relatively high performing early modern European economies.27Scheidel and Friesen 2009, 84-85. By way of early modern comparisons, Milanovic, Lindert and Williamson 2011 (Table 2) report the income Gini for Tuscany in 1427 = 46; Holland in 1561 = 56; England and Wales in 1688 = 45; France in 1788 = 56.

Mean Athenian income per capita (roughly GPD per capita) comes to 0.032 T (192 dr). Moving from this per capita mean to the mean income of a family household allows us to calculate according to multiples of the subsistence minimum. Mean household income comes to approximately 4.5 S. That estimate in turn allows us to measure inequality in Athens according to the Inequality Extraction Ratio, a method that is in some respects more useful for assessing inequality in pre-industrial societies than is the simple Gini coefficient because it factors in consumption as a multiple of subsistence.

Milanovic, Lindert and Williamson (2011) developed the Inequality Extraction Ratio (IER) as a metric for pre-modern inequality. The IER focuses on the difference between income necessary to achieve the subsistence minimum and the actual income of non-elites. It does so by measuring the percentage of actual inequality within a given society, as measured by the Gini coefficient, as a percentage of maximum feasible inequality. The maximum feasible Gini for the society in question sets its Inequality Possibility Frontier (IPF): the point at which, if their income were lower, non-elites would fall below the subsistence minimum. As Milanovic, Lindert and Williamson (2011) demonstrated, premodern societies often approach the IPF. If elites extract beyond the level that would leave others in their society at below a 1 S level, the poorest will die from deprivation. The point here is that even a very rapacious elite must ordinarily leave the non-elite with a subsistence minimum in order to avoid destroying elite sources of income. Comparing the actual Gini with the maximum feasible Gini yields the Inequality Extraction Ratio (IER). An IER of 100% indicates that the society has arrived at the IPF, the point at which more inequality implies death of some by deprivation. The median of the Milanovic, Lindert and Williamson study group is an IER of about 75%. Based on my income/population model, the IER for Athens in 330 BC is 48.9%. This makes Athens a low outlier among premodern societies, even lower than England and Wales in 1688, and very much lower than most other societies measured in Milanovic, Lindert and Williamson’s collection of case studies (see Table 1).

Table 1: Historical comparisons of income inequality

Middle, low, and elite population segments. Athenian median per capita income is estimated at 0.023 T (138 dr) per year. Mean Athenian income per capita (0.032) is thus 39% above median income. Coincidentally, this is almost exactly the ratio of mean to median for household income in the contemporary USA.28US Census Bureau, Income Distribution to $250,000 or More for Households 2013 (02.02.2016).Conversely, Athenian median income is 71% of mean income. When median income is translated (per above) into a multiple of bare subsistence, we see that a median family consumed at about 3.5 times subsistence (3.5 S). If we calculate the “middle class” by the ordinary method29Cf. Milanovic 2011, 171-172.of including all those whose income falls within 25% of the median, some 55% of Athenians were middle class; 23% fell below the middle class standard, while 22% are above that standard. By way of comparison, according to the figures cited by Milanovic (2011), in contemporary Latin America the median income is 60% of the mean, and 20% meet the middle class standard. In the average modern developed country, the median income is about 85% of mean income, and about 40% of the population meets the middle class standard.

Calculated another way, by multiples of subsistence, 76% of Athenian households fall within what Scheidel and Friesen (2009) consider the “decent middling” range of 2.4-10 times subsistence; 22.5% of households fall below the decent middling range; 1.3% of households are in the “elite” range of over 10 times subsistence.

The pre-tax income share (compared to the total income of all Athenians) of the top 1% of Athenians (2,228 persons: half of group 17, groups 18-25, 28-34) is about 13%; this falls roughly midway between the top 1% income share in a number of contemporary western European countries and the top 1% share in the contemporary USA.30See Alvaredo et al. 2013, and the data collected in Facundo Alvaredo, Anthony B. Atkinson, Thomas Piketty, Emmanuel Saez, and Gabriel Zucman, The World Wealth and Income Database: http://www.wid.world (02.02.2016).The average per capita income of an Athenian in the top 1% of income earners was about 18 times the income of a median Athenian (in the contemporary USA, it is about 12 times), and about 13 times the mean per capita income. The income share of the liturgical class (groups 15-34: 2.3% of the population, per above) is 16.3%.

Taxation. As the chapters in this volume by Wolfgang Franzen and Dorothea Rohde have clearly shown, taxation is a central feature of the Athenian polis, and the Athenian “tax morale” and tax system are in some ways distinctive: insofar as Athens’ level of income inequality is distinctively low by historical standards, that can be explained by the demands of the Athenian state that the elite pay taxes (or “voluntary” liturgies, see below), and the response of Athenian elites to that demand.

As noted above, the model developed is of pre-tax income and the taxation policy of the Athenian state levied direct taxes (other than on metics) uniquely on the wealthiest (indirect taxes on imports and exports would have affected all income classes). Although estimates can only be very rough (see below), it is certainly the case that Athenian taxation policy served to further equalize Athenian incomes. In the long run, by reducing elite household income over time (and thus the amount of income that could be privately saved), it would also tend to lower wealth inequality. The estimate of the tax burden on wealthy Athenians as a class must be very approximate, even when we assume that the burden is being calculated on actual (hidden and thus untaxed, as well as visible and thus taxed) income. But, again assuming that the figures in the model are within range, we can make some informed guesses.

The biggest “tax” on wealthy Athenians was in the form of trierarchic liturgies. Franzen (this collection) emphasizes the differences between taxes as ordinarily conceived and liturgies. Unlike modern taxes, liturgies were earmarked for specific purposes, involved non-monetary service, and the funds were in some sense controlled by those who paid them. He astutely points out that each of these features can be expected to have had a substantial effect on the tax morale of the liturgist: although the burden was substantial, the resistance to paying a liturgy may have been less than resistance to paying taxes, specifically because the liturgist retained some control over the funds he expended in the public interest. This goes some way to explaining the puzzle with which this chapter concludes: how was the relatively low-inequality Athenian social equilibrium sustained over time, in light of the danger of elite defection? In what follows, I will describe liturgies and other “voluntary” payments by elites to the state as taxes, but Franzen’s emphasis on the differences between modern taxes and liturgies must be kept in mind.

An “active” trierarchic liturgy, according to which the liturgist was required to outfit a warship for a sailing season, varied according to circumstances, but may have averaged about 4000 dr (0.66 T). If, hypothetically, 100 triremes were deployed in a given year, then the total trierarchic liturgy burden would have been ca. 67 T per year. Trierarchic liturgies are generally thought to be restricted to citizens, although the key text (Dem. Or. 20.20) only implies that this is the case and does not say this explicitly. Assuming that it was a citizen-only tax, the trierarchy system alone might produce a tax rate on the liturgical class of citizens of about 8-9% - and more if more ships needed to be manned in a given year.31Trierarchic liturgy amount: Davies 1971, xxi-xxiv; Pritchard 2012, n. 71. Gabrielsen 1994 discusses the evidence for financing warships in detail. His final estimate for the number of active service trierarchies per year, 60, is lower than my estimate, but his estimate for the cost of trierarchy is higher than my estimate, at 6000-7000 dr (ibid., 216). The total trierarchic tax burden comes to roughly the same amount: 60-70 T.

In addition to the trierarchy, there were about 100 festival liturgies, which may have averaged about 1500 dr, for a total of ca. 25 T. Annual property tax on wealthy estates, in ordinary years, totaled about 10 T (eisphora). There were also periodic special levies on the very rich (epidoseis, proeisphora). These taxes were paid by wealthy metics as well as citizens, meaning that the tax burden was spread out. The ca. 300 wealthiest citizen households (corresponding to groups 28-34, minimum estimated income of 1 T per year) were subject to the special assessments noted above. If we estimate that these other (non-trierarchic) taxes added up to a total of about 45 T per year, we may assume that the normal annual tax on wealthy Athenians as a class (citizens and metics alike) was in the range of 112 T, or roughly 10% of gross annual liturgical-class household income. Taxes on individual estates and on the liturgical class as a whole varied substantially from year to year. In years when fewer ships needed to be at sea, the total tax burden on the wealthy as a class would be lighter; in times of military emergency, taxes will have been higher.32See, further Hansen 1999, 110-115. I estimate festival liturgies at 25 T; annual normal eisphora at 10 T; and extraordinary taxes at 10 T.

The model yields an average liturgical-class household income of about 1 T per year, across the groups (15-34) of high-income citizen and metic households.

The impact of a trierarchic liturgy of 4,000 dr on an average estate would be very substantial - amounting to something like two-thirds of gross annual household income. The impact would be even greater, of course, on the lower tiers of the liturgical class, potentially amounting to four times annual household income for a household in group 15 or 25. As Mark Pyzyk (in progress) suggests, there was very good reason for Athenians in the lower end of the liturgical class to seek to avoid taxes by hiding income or foisting the burden onto someone else’s estate via a legal maneuver.33See further Gabrielsen 1986; Christ 1990; Cohen 1992; Kaiser 2007.A series of tax reforms in the course of the fourth century sought to smooth out the tax burden across the wealthy class, in ways that seem to have aimed at lessening the impact of a given year’s tax on the less prosperous liturgical-class citizen households.34See Gabrielsen 1994, ch. 8; Christ 2007.

Preliminary conclusions: income inequality

The preceding discussion of the population and income model suggests that Athenian income inequality in 330 BC was low by comparison with other documented historical pre-industrial societies and is in some ways comparable to rates of inequality in some contemporary developed states. The Gini coefficient and, even more saliently, the Inequality Extraction Ratio (IER) are both relatively low when compared to imperial Roman or to the most highly developed societies of western Europe before the Industrial Revolution. Athenian income inequality after taxes was lower yet.

The model allows only very rough estimates, but it also shows how far we would have to push the various assumptions on which it is based in order to bring Athens into line with the high IER of other historical societies. Suppose, for example, we were to lower mean Athenian income from my estimated 4.5S to (an implausibly low) 3S, and raise the Gini coefficient from 38 to (an implausibly high) 45. The Athenian IER would, under these revised assumptions, come to 67.5 (rather than, as estimated above, 48.9). Even under these extremely pessimistic assumptions, Athens would be well below the median IER (75) of historical societies studied by Milanovic, Lindert and Williamson. Those sorts of revisions to the model seem implausible because they would require either that our ancient evidence for Athenian wages is very misleading, or that for some reason the evidence is relevant only to a small part of the Athenian population, or that the liturgical class of wealthy Athenians was much larger that the evidence suggests, or that the richest were much wealthier.

Pushing any, or even all, of these underlying assumptions far enough to imperil the result of surprisingly (in comparative perspective) low Athenian income inequality would require dismissing much of our documentary, textual, and archaeological evidence for late classical Athenian society. Thus, while the actual figures derived from the model remain heavily stylized facts, the general conclusion that Athenian income inequality was comparatively low in historical perspective seems a great deal more likely than any alternative description that characterized Athenian inequality as either historically unexceptional or unusually high. Once again, the conclusion is strengthened if we consider after-tax income.

Other measures of inequality

If the population and income model developed above is anywhere close to a reasonable reflection of lived Athenian reality, we might expect to see a somewhat similar pattern in other measures of inequality. We may test the hypothesis that Athenian income inequality was relatively low in comparison with other preindustrial societies by reference to a prediction that inequality would be relatively low in other areas. While this test cannot prove, definitively, that Athenian income inequality was low, it does provide an indirect falsification condition: If other indicators show no tendency to lower-than-expected inequality, we would have reason to be more suspicious about the income results laid out above.

Wealth and land

Geoffrey Kron has measured wealth distribution among citizens in late fourth century Athens by calculating the Gini coefficient based on ancient text evidence. Kron calculates that in late fourth-century Athens the richest 1 percent of the citizen population owned about 30 percent of all private wealth; while the top 10% owned about 60 percent of the wealth. This yields a Gini coefficient of 70.8. Kron then compares this figure to the Gini wealth coefficients for several modern societies. The late-classical Athenian level of total-wealth inequality is roughly comparable to that of the USA in 1953-1954 (71). It is less equal than Canada in 1998 (69), but a more equal than Florence in 1427 (78.8) or the USA in 1998 (79.4). It is much more equal than the USA or England in the early 20th century (93 and 95 respectively).35Kron forthcoming. It is important to keep in mind that the overall Gini wealth index for Athenian society as a whole, including slaves and metics, would surely be substantially higher - I cannot say how much higher because I know no way to calculate the wealth of metics or slaves. Wealth inequality, as measured by the Gini coefficient, is typically much higher than income inequality.

Kron’s conclusion on the comparatively (in historical terms) equitable distribution of private wealth among citizens in late classical Athens is consistent with estimates of landholding in Athens: Two independent studies by classical scholars in the early 1990s concluded that about 7.5-9% of citizens owned about 30-35% of the land of Attica; some 20% owned little or no land. Excluding those at the top and bottom of the distribution, we are left with roughly 60-65% of the land being owned by about 70-75% of the citizen population.36Foxhall 1992; 2002; Osborne 1992.Ian Morris points out that the resulting Gini coefficient, which I calculate as about 48, is low in comparison to estimated distributions of land-holding for other ancient and medieval societies. Julián Gallego (2016) has recently calculated landholding inequality in late classical Athens (using eight rather than three wealth classes) at 44.1. Although the base line Athenian figures do not tell us anything about some relevant factors affecting the value of land, for example, distribution of especially productive land or financial encumbrances on landholdings, Morris is certainly right to conclude that, “the basic point is clear: landholding among citizens was unusually egalitarian in Classical Athens.”37Morris 1998a, 235-236. Quote, ibid., 36. As Claire Taylor pointed out to me, Morris misreported the Gini coefficient as 38.2-38.6; but Morris’ general point remains valid.

Nutrition

In an important study of a sample of isotopic analysis of 52 human skeletons from three different cemeteries in Attica, Anna Lagia (2015) has shed new light on patterns of ancient Athenian nutrition. The samples, which are sorted into the classical, Hellenistic, and Roman periods, yield several important results. First, nutrition (based on isotopic evidence for protein intake) was overall very good in the classical period. Nutrition levels worsened in the Hellenistic period burials, and declined further in the Roman era. This pattern of decline between the classical to the Roman eras is consistent with isotopic studies of nutrition in several other Greek sites. Lagia concludes, “[t]he homogenous picture of an elevated protein in the Classical period in sites as diverse as Makriyalos, Thebes, and Athens vis-à-vis earlier periods, in addition to a corresponding decrease in the Hellenistic period in Athens and Thebes, suggest that the causes of these changes were transformations in the living standards rather than random factors related to individual physiology or economic strategy.”38Lagia 2015, 140.This pattern of high classical era protein intake is consistent with the evidence for relatively high classical era wages across a substantial part of the population.

Equally important, for our purposes, is the lack of substantial differentiation among the nutrition levels of the individuals buried in the three cemeteries. Lagia argues persuasively that the first cemetery, Kerameikos, hosted generally elite burials; the second cemetery, Plateia Kotzia seems to have been used by ordinary people living in or near the urban center. The four burials in the study set from the third cemetery, located at the deme of Thorikos in the Laurion region of south Attica, are believed to be those of slaves working in the mining district.39Ibid., 121-212.Notably, there were no significant differences in the observed levels of nutrition across the three cemeteries.40Ibid., 127.This lack of differentiation between elite and nonelite nutrition in the classical era contrasts sharply with the very significant differences in nutrition observed (in other isotopic bone studies) between elite and non-elite individuals in Greek burials from the Bronze Age and the Medieval period.41See Knipper et al. 2015; Buikstra and Lagia 2009, 18.Lagia’s conclusions about the striking similarity in nutrition levels among presumed elite and non-elite Athenians of the classical era, based on her study of remains from the three Athenian cemeteries, is consistent with a model of relatively low level of income inequality in the late classical model.

Houses

House sizes can be employed as an indirect proxy of both consumption and inequality. Ian Morris (2004) has demonstrated that, based on the archaeological evidence, archaic/classical era Greek settlements were never characterized by a few mansions and many huts. The average size of houses grew dramatically across this period; in the Early Iron Age most Greeks lived in small and poorly constructed houses. By the late classical period the average Greek family lived in a much larger, and much better-built, house. But across the entire halfmillennium from 800-300 BC, the distribution of Greek houses tended to cluster around the median house size. As the biggest houses in the sample size grew over time, so too did the smaller houses. The size of houses in the top quartile (based on floor plan area) never diverged markedly from that of the bottom quartile. The size of larger and smaller houses grew more or less in lock step across the period: by 300 BC houses in the 75th percentile of the distribution were only about onefifth again (roughly 50 m2) as large as those at the 25th percentile.42Cf. Morris 2004, 722-723.

A comparative survey of classical era house sizes at Olynthos and other well preserved Greek urban areas by Kron (2011) confirms this general picture. Unlike, for instance, nineteenth century England, the distribution of house sizes at mid-fourth century Olynthos describes a bell curve: most houses fall in the middle, rather than on the far left (tiny house) side, of the distribution. Overall inequality among house sizes at Olynthos was very low; Kron calculates the Gini coefficient as 14. In later periods of antiquity, from the Hellenistic through Roman era, house sizes diverged (i.e., the Gini coefficient increased); by the late Roman period the difference in average size between great villas and ordinary houses was immense.43Kron 2014 and forthcoming. For Olynthos houses, Kron 2014, 129, table 2, calculates the Gini coefficient at 14, considerably lower than later Hellenistic and Roman era Greek cities. Cf. Bintliff 2012, ch. 13.

In the case of houses, the samples are not drawn just from Athens, but from across the Greek world. Nonetheless, there is no reason to believe, based on what is known of classical era Athenian houses, that house-size at Athens was more unequal than the classical era Greek norm. The comparison of relatively equal house size in the Greek classical period, and the suggestion of very considerable growth of inequality in post-classical Greece, fits with the general conclusion of relatively low economic inequality in late classical Athens. Indeed, it suggests that the world of the classical Greek city-states overall may have been characterized by historically low levels of economic inequality.

Conclusions

The evidence for relatively low levels of inequality in classical era Athenian wealth and landholding, Athenian classical era nutrition, and archaic to classical era Greek house size is consistent with the preliminary conclusion, based on the population and income model developed here, of relatively low income inequality in late classical Athens.44Athens, and the classical Greek world generally, appear to be outliers in the long-term global historical pattern of high inequality (except in the aftermath of major war, natural catastrophe, plague, or state collapse): Scheidel 2017.The overall picture is tolerably clear. The historically high level of political equality (among a substantial portion of adult native males) that characterized many Greek city-states, and that was especially characteristic of democratic Athens, is matched by historically low levels of economic inequality. Here, I will not attempt to develop a causal argument connecting democratic political equality and economic equality. But the progressive taxation employed by the Athenian state in the democratic era, which meant that direct taxes (especially in the form of liturgies, eisphorai, epidoseis, and other quasivoluntary donations) were uniquely levied on the richest residents, offers an obvious mechanism for a causal argument in which high levels of political equality explain low levels of economic inequality.45On the difficult question of taxation and state finance in pre-democratic Athens, see van Wees 2013; Bubelis 2014.

Relatively low Athenian inequality, in a context of relatively high levels of overall prosperity, provides a potential (if necessarily partial) explanation for the Athens’ record of overall social stability during the democratic era (the obvious exception is the crisis of the late fifth century). Despite what appear to be historically high levels of chattel slavery, there is no evidence of slave revolts at Athens (or, for that matter, except at Sparta, elsewhere in classical Greece).

The lack of slave revolts might be explained, in part, by the evidence for good nutrition among slaves. Athenian slaves seem not to have been kept at the edge of starvation. Perhaps even more important in reducing slaves’ incentive to revolt was the hope of manumission.46See Roubineau 2015, ch. 10.Aristotle, who notoriously argued that slavery could be the natural (and thereby best) condition for some humans, also promised in Book 7 of the Politics to offer advice on how slaves ought to be treated, and suggested that in his ideal polis “it is better to hold out [the chance of] freedom as a reward for all slaves” (Pol. 7.10 1330a 33). Aristotle did not (in the preserved corpus) fulfill the promise to explain exactly how slaves should be treated and why it was better for them to have the hope of manumission. But we may guess that proper treatment included the chance for a slave to “buy himself” through saving surplus income, and that the lessened threat of slave resistance (through revolt or inferior work) was among the reasons that Aristotle commended the practice of holding out the hope of manumission to slaves.

Moving up the social scale, metics would have more reason to migrate to Athens (and to settle there) if wages were good and if the rules allowed for economic advancement and contract enforcement. Likewise, poorer citizens will have had less reason to seek property redistribution if their wages were high enough to enable them to live reasonably comfortably, and if they believed that their wealthier fellow citizens were doing their fair share in paying taxes for the provision of important public goods. If the poorer citizens had used their majority voting power to set tax rates at an extortionate level, the wealthy would have more reason to seek regime change.47For the association between levels of taxation in democratic societies and social assumptions about fairness, see now Scheve and Stasavage 2016.

Surely some Athenian elites must have hoped for a different government, one in which the ordinary citizens did not set the rate of elite taxes. But revolutionary action to change the democratic government would require coordinated action across the ranks of elite Athenians. As Ps.-Xenophon (Ath. Pol. 2.20; 3.13) had noted in the later fifth century, many elite Athenians chose to cooperate actively with the democratic regime. The same seems to be the case in the fourth century. Tax rates are hardly the only factor that came into play in regulating relations between the mass of citizens and the classical era Athenian elite.48See Ober 1989.But it appears likely that, especially in light of the distinctive “tax morale” that Franzen (this collection) has argued was associated with the liturgical system, tax rates were, in the later classical period, overall low enough to discourage elites from coordinating on revolutionary plans to overthrow the democratic government.

Appendix Table. Athens, 330 BC. Model of population and income.

Bibliography

Acemoglu, D. and Robinson, J. A. 2012.

Why Nations Fail: The Origins of Power, Prosperity and Poverty. New York: Crown Publishers.— — 2016.

“Paths to Inclusive Political Institutions.” Working Paper.

Acton, P. H. 2014.

Poiesis: Manufacturing in Classical Athens. Oxford: Oxford University Press.

Akrigg, B. 2015.

“Metics in Athens.” In: C. Taylor and K. Vlassopoulos (eds.), Communities and Networks in the Ancient Greek World. Oxford & New York: Oxford University Press, 155-176.— — 2011.

“Demography and Classical Athens.” In: Holleran and Pudsey 2011, 37-59.

Allen, R. C. 2009.

“How Prosperous Were the Romans? Evidence from Diocletian’s Price Edict (AD 301).” In: A. K. Bowman and A. Wilson (eds.), Quantifying the Roman Economy: Methods and Problems. Oxford & New York: Oxford University Press, 327-345.

Alvaredo, F. et al. 2013.

“The Top 1 Percent in International and Historical Perspective.” Journal of Economic Perspectives 27/3: 3-20.

Bintliff, J. 2012.

The Complete Archaeology of Greece: From Hunter-Gatherers to the 20th Century A.D. Chichester & Malden, MA: Wiley-Blackwell.

Boix, C. 2015.

Political Order and Inequality: Their Foundations and their Consequences for Human Welfare. Cambridge Studies in Comparative Politics. Cambridge: Cambridge University Press.

Bresson, A. 2015.

The Making of the Ancient Greek Economy: Institutions, Markets, and Growth in the City-States. Princeton: Princeton University Press.

Buikstra, J. and Lagia, A. 2009.

“Bioarchaeological Approaches to Aegean Archaeology.” In: L. A. Schepartz, S. C. Fox and C. Bourbou (eds.), New Directions in the Skeletal Biology of Greece. Hesperia Supplement 43. Princeton, NJ: American School of Classical Studies, 7-30.

Bubelis, W. 2016.

Hallowed Stewards: Solon and the Sacred Treasurers of Ancient Athens. Ann Arbor: University of Michigan Press.

Cartledge, P. 1985.

“Rebels and ‘Sambos’ in Classical Greece: A Comparative View.” In: id., F. D. Harvey and G. E. M. de Ste. Croix (eds.), Crux: Essays Presented to G. E. M. de Ste. Croix. London: Duckworth, 16-46.

Christ, M. R. 1990.

“Liturgy Avoidance and Antidosis in Classical Athens.” Transactions of the American Philological Association 120: 147-169.— — 2007.

“The Evolution of the Eisphora in Classical Athens.” Classical Quarterly 57: 53-69.

Cohen, E. E. 1992.

Athenian Economy and Society: A Banking Perspective. Princeton, NJ: Princeton University Press.

Davies, J. K. 1971.

Athenian Propertied Families, 600-300 B.C. Oxford: Clarendon Press.

Fawcett, P. W. 2016.

“‘When I Squeeze You with Eisphorai.’ Taxes and Tax Policy in Classical Athens.” Hesperia 85: 153-199.

Foxhall, L. 1992.

“The Control of the Attic Landscape.” In: B. Wells (ed.), Agriculture in Ancient Greece: Proceedings of the Seventh International Symposium at the Swedish Institute at Athens, 16-17 May, 1990. Stockholm: Paul Åströms Förlag, 155-159.Franzen, W. 2019.

“Tax Morale in Classical Athens.” [this volume]

Gabrielsen, V. 1994.

Financing the Athenian Fleet: Public Taxation and Social Relations. Baltimore: Johns Hopkins University Press.

Gallego, J. 2016.

“El campesinado y la distribución de la tierra en la Atenas del siglo IV a.C.” Gerion. Revista de Historia Antigua 34: 43-75.

Hansen, M. H. 1986.

Demography and Democracy: The Number of Athenian Citizens in the Fourth Century B.C. Herning: Systime.— — 1988.

Three Studies in Athenian Demography. Copenhagen: Det Kongelige Danske videnskabernes selskab.— — 1999.

The Athenian Democracy in the Age of Demosthenes: Structure, Principles and Ideology. Norman, OK: University of Oklahoma Press.— — 2006.

Studies in the Population of Aigina, Athens and Eretria. Copenhagen: Royal Danish Academy.

Harris, E. M., Lewis, D. M. and Woolmer, M. (eds.). 2015.

The Ancient Greek Economy: Markets, Households and City-States. Cambridge: Cambridge University Press.

Holleran, C. and Pudsey, A. (eds.). 2011.

Demography and the Graeco-Roman World: New Insights and Approaches. Cambridge: Cambridge University Press.

Jones, A. H. M. 1964.

Athenian Democracy. Oxford: B. Blackwell.

Kaiser, B. A. 2007.

“The Athenian Trierarchy: Mechanism Design for the Private Provision of Public Goods.” Journal of Economic History 67/2: 445-480.

Kamen, D. 2013.

Status in Classical Athens. Princeton, NJ: Princeton University Press.— — in progress.

Manumission in Ancient Greece: Modes, Meanings, and Metaphors.

Knipper, C. et al. 2015.

“Superior in Life - Superior in Death: Dietary Distinction of Central European Prehistoric and Medieval Elites.” Current Anthropology 56: 579-589.

Kron, G. 2014.

“Comparative Evidence and the Reconstruction of the Ancient Economy: Greco-Roman Housing and the Level and Distribution of Wealth and Income.” In: F. de Callataÿ (ed.), Quantifying the Greco-Roman Economy and Beyond. Bari: Edipuglia, 123-146.— — forthcoming.

“Growth and Decline. Forms of Growth. Estimating Growth in the Greek World.” In: E. Lo Cascio, A. Bresson and F. Velde (eds.), The Oxford Handbook of Economies in the Classical World. Oxford: Oxford University Press.

Lagia, A. 2015.

“Diet and the polis: An Isotopic Study of Diet in Athens and Laurion during the Classical, Hellenistic, and Roman Periods.” In: A. Papathanasiou, M. Richards and S. C. Fox (eds.), Archaeodiet in the Greek World: Dietary Reconstruction from Stable Isotope Analysis. Hesperia Supplement 49. Princeton, NJ: American School of Classical Studies at Athens, 119-146.

Lyttkens, C. H. 2012.

Economic Analysis of Institutional Change in Ancient Greece: Politics, Taxation and Rational Behaviour. London: Routledge.

Milanovic, B. 2011.

The Haves and the Have-Nots: A Brief and Idiosyncratic History of Global Inequality. New York: Basic Books.

Milanovic, B., Lindert, P. H. and Williamson, J. G. 2011.

“Pre-Industrial Inequality.” The Economic Journal 121/551: 255-272.

Morris, I. 1998.

“Archaeology as a Kind of Anthropology (A Response to David Small).” In: id. and K. A. Raaflaub (eds.), Democracy 2500? Questions and Challenges. Dubuque, IA: Kendall/Hunt Pub. Co., 229-239.— — 1998b.

“Remaining Invisible: The Archaeology of the Excluded in Classical Athens.” In: S. Murnaghan and S. R. Joshel (eds.), Women and Slaves in Greco-Roman Culture. New York: Routledge, 193-220.— — 2004.

“Economic Growth in Ancient Greece.” Journal of Institutional and Theoretical Economics 160/4: 709-742.

Ober, J. 1989.

Mass and Elite in Democratic Athens: Rhetoric, Ideology, and the Power of the People. Princeton, NJ: Princeton University Press.— — 1998.

Political Dissent in Democratic Athens: Intellectual Critics of Popular Rule. Princeton, NJ: Princeton University Press.— — 2008.

Democracy and Knowledge: Innovation and Learning in Classical Athens. Princeton. NJ: Princeton University Press.— — 2015a.

“Classical Athens [State Finances].” In: W. Scheidel and A. Monson (eds.), Fiscal Regimes and Political Economy of Early States. Cambridge: Cambridge University Press, 492-522.— — 2015b.

The Rise and Fall of Classical Greece. Princeton, NJ: Princeton University Press.— — 2016.

“Institutions, Growth, and Inequality in Ancient Greece.” Working Paper. — — 2017.

“Inequality in Late-Classical Democratic Athens: Evidence and Models.” G. C. Bitros and N. C. Kyriazis (eds.), Democracy and Open-Economy World Order. Cham: Springer International Publishing, 125-146.

Osborne, R. 1992.

“Is It a Farm? The Definition of Agricultural Sites and Settlements in Ancient

Greece.” In: B. Wells (ed.), Agriculture in Ancient Greece. Stockholm: Paul Åströms Förlag, 22-27.

Patriquin, L. 2015.

Economic Equality and Direct Democracy in Ancient Athens. London: Palgrave MacMillan.

Pritchard, D. 2012.

“Costing Festivals and War: Spending Priorities of the Athenian Democracy.” Historia 61: 18-65.

Pyzyk, M. in progress.

“Onerous Burdens: Liturgies and the Athenian Elite.”

Rohde, D. 2019.

“For Everything to Remain the Same, Everything Must Change! Private Wealth and Public Revenues.” [this volume]

Roubineau, J.-M. 2015.

Les cités grecques (VIe-IIe siècle av J.-C.): Essai d’histoire sociale. Paris: Presses Universitaires de France.

Scheidel, W. 2005.

“Real Slave Prices and the Relative Costs of Slave Labor in the Greco-Roman World.” Ancient Society 35: 1-17.— — 2007.

“Demography.” In: id., I. Morris and R. P. Saller (eds.), The Cambridge Economic History of the Greco-Roman World. Cambridge: Cambridge University Press, 38-86.— — 2008.

“The Comparative Economies of Slavery in the Greco-Roman World.” In: E. Dal Lago and C. Katsari (eds.), Slave Systems: Ancient and Modern. Cambridge & New York: Cambridge University Press, 105-126.— — 2010.

“Real Wages in Early Economies: Evidence for Living Standards from 1800 BCE to 1300 CE.” Journal of the Social and Economic History of the Orient 53: 425-462.— — 2017.

The Great Leveler. Violence and the Global History of Inequality from the Stone Age to the Present. Princeton, NJ: Princeton University Press.

Scheidel, W. and Friesen, S. J. 2009.

“The Size of the Economy and the Distribution of Income in the Roman Empire.” Journal of Roman Studies 99: 61-91.

Scheve, K. and Stasavage, D. 2016.

Taxing the Rich: A History of Fiscal Fairness in the United States and Europe. Princeton, NJ: Princeton University Press.

Silver, M. 2006.

“Slaves versus Free Hired Workers in Ancient Greece.” Historia 55: 257-263.

Taylor, C. 2017.

Poverty, Wealth, and Well-being: Experiencing penia in Democratic Athens. Oxford: Oxford University Press.

Taylor, T. 2001.

“Believing the Ancients: Quantitative and Qualitative Dimensions of Slavery and the Slave Trade in Later Prehistoric Eurasia.” World Archaeology 33/1: 27-43.

Van Wees, H. 2011.

“Demetrius and Draco: Athens’ Property Classes and Population in and before 317 BC.” Journal of Hellenic Studies 131: 95-114.— — 2013.

Ships and Silver, Taxes and Tribute: A Fiscal History of Archaic Athens. London: I. B. Tauris.

Whitehead, D. 1977.

The Ideology of the Athenian Metic. Cambridge: Cambridge Philological Society.— — 1984.

“Immigrant Communities in the Classical polis. Some Principles for a Synoptic Treatment.” L’Antiquité Classique 53: 47-59.

Zelnick-Abramovitz, R. 2005.

Not Wholly Free: The Concept of Manumission and the Status of Manumitted Slaves in the Ancient Greek World. Leiden & Boston: Brill.— — 2009.

“Freed Slaves, their Status and State Control in Ancient Greece.” European Review of History 16/3: 303-318.