There were always kingdoms and wars in Gaul right up until you submitted to our laws. Although we have often suffered at your hands we have, by right of conquest, imposed only this one thing on you, with which we keep the peace. For peace between nations is impossible without soldiers, and there are no soldiers without pay, and no pay unless taxes are paid. Everything else we share with you.
(Tacitus, Histories 4.74)
The Political Economy of Tributary Empires
Who paid for empire? Like all imperial rulers, the Romans passed the cost on to their subjects. Romans knew this. Tacitus puts this pithy summary of imperial economics into the mouth of the Roman general Cerealis, in a speech aimed at dissuading the Gallic tribes of the Treveri and Lingones from joining the rebellion of AD 69. This was, indeed, the bottom line. By Tacitus’ day the army devoured most of what the emperors raised in taxation. Put like this, the resourcing of Roman imperialism seems very simple. Yet the mechanisms employed were phenomenally complex and in constant evolution. Roman history is, in some sense, the story of unending struggles to balance the imperial budget. Perhaps this was true for all imperial states.
A comparative perspective indicates the tight constraints within which Rome had to solve these problems. Early empires were vast redistributive systems: the key resources on which they depended were land and manpower. Metals, timber, and hard stone were also important; correspondence between the Bronze Age kings of the ancient Near East was already often preoccupied with securing these precious resources. But the basis of all ancient economies was agriculture. Every early empire was, in the final analysis, funded from the agricultural surplus. Given that empires were built on inequality and were very large, they also depended on transport infrastructure. Empires typically spent on soldiers, functionaries, and imperial courts, and none of these groups was evenly distributed among the productive landscapes they controlled. There were not many options. Food could be moved to the consumers; the consumers could go to the food; or else monetary systems could be devised which allowed states to pay in cash but required subject populations to sell surplus on the market to earn money to pay taxes, and consumers to use the market to obtain what they needed. Rome eventually used a combination of these, building roads and ports, levying taxes in kind and in cash, providing incentives for traders to bring their cargoes to the imperial capitals, and expecting provincial populations to supply armies and imperial courts on the move.1
The Roman solution was therefore broadly similar to that employed by other early empires. Great infrastructure projects included the Grand Canal of China, begun in the fifth century BC and nearly 1,000 miles long by the time of its completion a millennium later; the Persian Royal Road that ran over more than 1,500 miles from Sardis to Susa; and the great Inka Road that ran 3,700 miles along the length of the Andes. Rome was especially fortunate in ruling an empire arranged around an inland sea. Moving the consumers to the food was less practical for empires than for smaller states. Early English kings could move their tiny courts around their little kingdom with relative ease, but empires required more complex systems. As a result most created imperial monetary systems with which to pay soldiers and state functionaries. Often this involved extending the uses of an earlier coinage, like the copper coins of the Qin kingdom that became the first imperial coinage of China, or Rome’s silver denarii which in the course of the late Republic gradually replaced all the other precious metal coinage of the Mediterranean world.2 Along with imperial coinages there usually went imperial standard measures. The Athenians’ conversion of the Delian League into something like an empire was marked when the assembly issued a decree, probably in the mid-420s BC, requiring its allies to use Athenian coins, Athenian weights, and Athenian measures.3 The Achaemenid Persian Empire issued coinage and demanded some taxes be paid in coin: Hellenistic empires typically made more used of standardized monetary systems.4
Historians sometimes call this sort of political system a tributary empire, perhaps for obvious reasons.5 Tributary empires may be contrasted with conquest states, polities with institutions and ideologies geared to constant expansion. The Aztec political order depended on annual war, and it had state rituals that demanded constant supplies of war-captives for sacrifice. A run of years without victories would cause political collapse. When conquest states enjoy a sustained comparative advantage over their neighbours, periods of astonishingly rapid expansion might occur. Typically this involved not only rewards for the conquerors, but also means by which new members were recruited to their armies. Conquest states move like tsunamis across political landscapes. The Arab conquest that between 634 and 720 created a caliphate stretching from southern France to the Punjab, and the Inka conquest of the Andes in the century after AD 1438, were both movements of this kind. But this kind of forward drive cannot be maintained indefinitely. All conquest states are doomed either to sudden collapse (like the empire of Attila the Hun) or else to become institutionalized as tributary empires. Achaemenid Persia provides an excellent example. The empire was created between 559 and 522 by Cyrus and his son Cambyses, who together conquered the kingdoms of the Medes, the Babylonians, and the Egyptians: but their empire nearly collapsed in civil war until Darius I took control, creating a single currency, a tax system, a provincial system, and the Royal Road. Greek and Roman writers, spellbound by the horror of civil war, present the achievement of Augustus mostly in terms of establishing civil peace.6 But what really saved the empire from collapse was his success in stopping expansion, and consolidating it around those institutions that were already geared to the sustainable economics of a tributary empire.
Here too there were only a few fiscal options open to the rulers of tributary empires. Local rulers of various kinds could be recruited to act as agents of empire; tax farmers could be employed; or a tax-gathering bureaucracy could be created. Each method had its disadvantages. Depending on local elites meant surrendering some power over the provinces. A detailed comparison has recently been drawn between Rome’s use of local elites and the situation in the Mughal Empire which extracted tax with the help of localzamindars, in the process conceding them significant autonomy.7 Tax farmers minimized the risks and costs of tax collection, guaranteeing the state a fixed income. But they notoriously had short-term interests, and were the least sympathetic to taxpayers. The long-term disadvantages of public–private partnerships of this kind are now very familiar to us: tributary empires also had to run the risk of revolts and protests stirred up by profiteering. A bureaucracy gave emperors much more control, but it came at a significantly higher cost, one that in the end would have to be recouped from the taxpayers.
The broad path of Rome’s own journey through these tangled waters is simply stated. Roman expansion in Italy conforms closely to the ideal type of a conquest state. But already within the second century BC elements of more stable revenue extraction appear. Indemnities from defeated powers were replaced over that century by regular income, as former Hellenistic kingdoms were absorbed and other territory was won in the west. At first Rome depended on tax farmers: for a state already making the widespread use of public contracts that Polybius observed this was natural. But tax farming was impractical in some regions—notably Spain—and the appalling behaviour of those who held the contract to collect Roman taxes in Asia was widely blamed for Greek support for Mithridates. Caesar entrusted the collection of land tax to the local elites of Asian cities, and that system became widespread during the early first century AD except in Italy, which was exempt, and Egypt where the bureaucracy used by the Ptolemies was preserved, and perhaps some frontier zones where the military acted as different kinds of bureaucrats. The early imperial system remained, however, exceptionally complex and tax farming continued to be used for many indirect taxes.8 The Augustan system had most consistency at the very highest level. Further down, there was little desire to tinker with systems that worked well enough, and only a few attempts to make improvements can be documented. All this changed in the crisis of the third century. The empire needed greater revenue at a time when the economy was for one reason or another weaker, and when local aristocracies were under unprecedented pressure. The result was that as part of the imperial recovery a new system emerged, with new taxes, a new coinage, and a centralized bureaucracy that lasted into the Byzantine Middle Ages.
Good Times and Bad Times
Fiscal systems can be seen as governmental responses to the economic activities of their subjects. Ideally they extract as much as possible without harming that activity. As Tiberius is said to have put it: ‘I want my sheep shorn, not flayed.’ The historical ecology of the Mediterranean basin and its hinterlands has been described already.9 The agrarian regimes of classical antiquity were essentially stable. Roman rule brought a few new farming techniques and a few new crops into some regions, but there was no revolutionary change. That ‘normal’ background had some built-in short-term instability, especially in the Mediterranean part of the empire where food crises were not unusual.10 More generally the ancient Mediterranean was characterized by highly localized cycles of boom and bust that drove peasant cultivators into strategies of crop-diversification, storage, and exchange.11 Growth, where it took place, was the result of intensification. Where landowners had the funds and the desire we can observe them draining and irrigating; planting vines, olive trees, and gardens that would be more profitable than other crops; experimenting with new varieties of trees, with selective livestock breeding; improving the value of their estates by opening up clay-pits, building kilns and olive presses; constructing mill and storage facilities; and improving the transport facilities they used to get their surplus to market. Broadly speaking, the poor feared risk, the rich sought profit, but both pursued their ends by essentially traditional means.
Yet long-term trends did emerge from this activity, trends with which the tributary empire had to deal. Work progresses apace on filling out the details of these trends, and especially on quantifying them.12 Underwater archaeology has shown how the numbers of shipwrecks rose to a peak in the late Republican period, suggesting this was the period of greatest long-distance trade in the ancient Mediterranean.13 Over the same period Italian products, especially wine and ceramic tableware, are found all over the Mediterranean world, and beyond it too. From the early first century AD this evidence diminishes in volume. But there are indications of growth in the provinces, including the production of olive oil for export in North Africa and southern Spain, and the production of wine for local consumption in areas as varied as central France, the environs of Rome, and Egypt. Trade, in other words, boomed first, followed by the capacity to produce the same goods locally.14 New evidence for increased levels of mining and metal production has come from ice cores drilled through the Greenland ice cap. To judge from the levels of atmospheric lead and copper pollution attested there, the production of metals reached a peak in the early Roman Empire not repeated before the Industrial Revolution.15
For the most part, these changes were demand led. For example, one major stimulus to change was the spread of new styles of consumption across the empire, styles modelled on those of the Italian elites. Oil, wine, fish-sauces, textiles, bronzes, ornaments were all in demand among their provincial counterparts. Changes in consumption did not affect commodities alone. Teachers and potters and wall painters found employment in the provinces. Italian architects, engineers, and craftsmen were already at work on public buildings in the western provinces in the early first century AD.16 Once these skills had been taught to locals, quarries had been opened up, tile production had begun, and new carpentry techniques and design features were disseminated, we begin to see the transformation of domestic architecture, first in the cities and then in the countryside. Even more sophisticated engineering skills were needed to build aqueducts that powered monumental fountains and bathhouses, signs of new aesthetic and cosmetic sensibilities. All this was expensive, but if some of Rome’s subjects were exploited more to pay for it, others might make money satisfying these new tastes.17
The greatest change in lifestyle in many parts of the empire was the growth of cities. The number of cities, their density, and the population of the larger centres grew all over the empire. Regions with only villages before Rome—inland Gaul and Spain, central Anatolia, parts of Egypt and the Balkans—experienced the greatest growth. In a few areas, such as the Nile Valley, central Italy, and coastal Asia Minor, the proportion of non-producers who had to purchase their food crept up to as much as 30 per cent of the population. A large part of this growth was the consequence of the emergence of a small number of enormous cities at the top of the settlement hierarchy. The population of Rome reached around one million, and maybe ten other cities crossed the 100,000 mark. The urban system itself was changing, and in some areas small cities grew smaller as the larger ones swelled in size.18 But by most estimates the total urban population of the empire increased to a maximum around AD 200.19 Add to this the creation of a standing army that fluctuated between a quarter and half a million men, and it is clear where the new demand was coming from.
The Roman Empire promoted these processes accidentally and indirectly. By supporting urbanization and creating a standing army, the empire increased net demand. The models of civilized life that, under Roman rule, spread beyond the Mediterranean basin provided opportunities for merchants, craftsmen, and architects. Roman rule favoured the rich, and to the extent that they became richer their aggregate purchasing power increased. Standardized currencies and weights and measures, and investment in transport infrastructure—all designed to serve administrative and military ends— must have made trade easier. Law, common languages, and peace must also have made their contributions. Perhaps taxation played a part as well, stimulating landowners to produce greater surpluses, whether to be supplied in kind to the army or sold to pay taxes in case.20
Intensification of this kind powered growth from at least the second century BC until some point in the third century AD. But then the process went into reverse. Cities shrank, investment in production seems to have diminished, and certain kinds of long-distance trade declined in volume. There are many uncertainties. In some cases a decline in trade reflected the growing success of local producers: cycles of production and consumption, in other words, were becoming more localized. Archaeological studies of some commodities, for example ceramic products from North Africa, suggest long-distance trade continued up to and beyond the collapse of the western empire. Africa remained a major exporter of grain too, even under the Vandals. Most indicators of economic decline also suggest a more dramatic decline in the north and west of the empire than in its southern and eastern provinces. The local economies of Syria and parts of Asia Minor actually seem to boom in late antiquity. Even less consensus exists about the reasons for the reversal of many of these trends, or indeed about the point at which this reversal began. The peak in shipwrecks is actually in the late Republic, and Italian wine exports reach fewer and fewer regions in bulk from the same period. Could improvements in shipping or increased domestic consumption of wine explain part of this? The urban apogee, however, is around two centuries later.
Once the urban peak was passed, collapsing demand certainly did have a major effect. Many western cities, Rome included, shrank dramatically in population between 200 and 300 AD. The occupied areas of some cities in the north-west provinces were reduced to a quarter of their second-century maximum by the end of the third century. A few cities were even abandoned. Many cities must have contained vast areas of crumbling buildings and empty plots in the late empire, and quite a few huddled around a fortified castle in old monumental centres. Hardly any new urban monuments were constructed after the 230s anywhere. Rome itself dropped to a third of its size in the same period. All this must have affected the market for foodstuffs and textiles, for ceramics and fuel, and also the construction industry. The rich remained rich and some became richer: some of the most splendid rural and suburban villas were constructed in the fourth century all over the empire. But their spending alone could not absorb the mass productions that agricultural intensification had aimed at generating.
Many factors have been suggested to explain these changes. It is sometimes suggested that as economies became more and more regional, some parts of the empire had simply opted out of its expensive urbanized civilization. But it is difficult to see the articulation of alternative value-systems in the literary texts of the fourth century. Indeed they are often so nostalgic in tone that they are described as classicizing. The barbarian invasions of the late third century cannot have done this much damage, which is also evident in areas unaffected by those raids such as Britain. Nor had the empire (yet) increased the tax burden to the point where the productive base was under pressure, nor were the wealthy (yet) so wealthy that they had crippled public finances.
Recently explanations in terms of plague and climate change have been revived. Does each age visit its own fears on the fall of the Roman Empire? A terrifying plague certainly did grip the empire in the middle of the second century AD.21 Vivid descriptions have survived, including one from Galen, the leading physician of the early empire. It may have been smallpox, or measles, or a disease that no longer exists. It came from the east, brought into the Roman Empire by an army that encountered it while campaigning against the Persians in the 160s AD, and spread rapidly along the militarized areas of the Danube and Rhine and eventually reached Rome driving refugees (and at one point emperors) before it. But it is difficult to estimate its effect on the long-term operation of the economy: comparative evidence shows plagues can have a range of effects, some even positive, on economic growth.22 The question of climate change is even less certain, and hinges on very large-scale estimation of fluctuations of mean annual temperature. If there was indeed a slight decrease in temperature around the middle of the first millennium BC, this could have affected agricultural productivity. Both these ideas see the high point of the Roman economy as resting on a rather fragile basis, and that thesis is genuinely difficult to assess in the current state of the data. And there are other alternatives. One strand of argument suggests that the ancient economy had reached its maximum carrying capacity even earlier, and that the growth of the last centuriesBC was a final spurt generated by the incorporation of new regions into the Mediterranean system. It is difficult at the moment to decide between these hypotheses.
Whatever the reasons, the consequences of economic contraction for a tributary empire are clear. Indirect taxes tapped the profits of trade, auctions, and manumissions. If fewer of these took place, the revenue decreased. Direct tax was based on the land, but if land became less profitable there was a limit to what could be raised. This is, however, clearer to us than it was to them.
Harnessing the Ancient Economy
This picture of the economy that is emerging from the very latest research would have been incomprehensible to the Romans themselves. Their very practical understanding of economic activity did not include the use of predictive or descriptive modelling, they gathered little data from which they could have analysed trends, and ancient science had no concept of the economy as a separate entity. In that sense the emperors were flying blind as they designed and modified their tax systems.
But their solutions to problems were not foolish. When earthquakes devastated the cities of Asia, Tiberius remitted tax income for five years. When Augustus needed more money for the military he created new taxes. When prices began to soar at the end of the third century Diocletian tried to fix legal maximums. Historians understood that when new silver mines were opened up the price of silver would decrease, and also how changes in rules governing interest rates could provoke a shortage of coin (even if they did not have specialized terms for ‘supply and demand’ or ‘liquidity crisis’). It follows that the devices that the emperors employed to tax this vast economy were pragmatic, if also fundamentally reactive and adaptive. Not much effort was put into smoothing out differences between provinces taxed in different ways: uniformity and consistency were not sought in themselves, and establishing equity between different groups of taxpayers was never a concern. The emperors harvested the economy opportunistically, aiming to take a share of whatever profit was being made. As a result the early imperial system preserved institutional fossils of every stage of Roman imperialism, and of some earlier ages too.
The political economy of the Roman Republican state before the Punic Wars was almost non-existent. Successful campaigns brought some booty, especially chattel slaves and bullion, most of which was divided between the allies, the citizen soldiers, the general, and the gods. The state had in any case few expenses. Most of the monuments built in this period were temples constructed in fulfilment of battlefield vows and paid for out of the general’s share of the booty.23 The Roman census did allocate tax obligations according to wealth, but we know very little of this direct taxation, except that it was abolished for ever after the defeat of Macedon in 168 BC. Warfare up and down the peninsula extended Roman control over land and manpower. Conquered manpower it exploited through enslavement and those alliances that required subject states to provide troops to support Roman armies. Captured land became ager publicus and was used either to found colonies or else rented out to Roman citizens. Those rents (vectigalia) became one of the state’s first regular and predictable sources of income.24
Overseas wars with Carthage and Macedon brought other sources of income. Indemnities were imposed on defeated Carthage in 241 and 201, in both cases spread into a series of annual payments. Macedon and Syria too had to pay massive indemnities after their respective defeats in 196 and 188 BC. During this period great public building works were initiated in Rome.25 Other public contracts were issued for the provisioning of armies. Only Roman citizens with funds to guarantee them could take public contracts, which were a means of spreading the proceeds of empire among the propertied classes.
The political economy was transformed when Rome began to acquire territory overseas. The first province was Sicily. Carthage had taxed her possessions in the west of the island and perhaps Rome took over their fiscal system after the first Punic war. After the capture of Syracuse in 211 during the second Punic war, Rome adapted the tax system created by King Hiero. What became known as the Lex Hieronica in effect imposed a tithe on the agricultural produce of most of the Greek cities of the island. Rome allowed a few cities exemption, and deprived some others of their land. Using taxation to reward allies and punish enemies became a standard Roman technique. Absorbing this small Hellenistic kingdom probably opened the Romans’ eyes to the possibilities of using tax as a means of generating a regular income out of their military supremacy. Certainly when Tiberius Gracchus successfully urged the takeover of the kingdom of Pergamum in 133 the key motive was to provide an income stream for his own project of land distribution. The Roman people received the income from royal taxes and the royal lands. Tax farmers (publicani) made a profit from its collection, and this bought Gracchus political support in Rome.
Tax farming of one kind or another remained important long into the empire for the collection of indirect taxes.26 An inscription of Nero’s day found recently at Ephesus shows that the various customs duties created by the kings of Pergamum were still being levied in exactly the same way they had done more than two centuries after the Roman takeover.27 Tax farmers remained in charge of collecting internal tariffs long after they had lost the lucrative contracts to collect the land tax in the last years of the Republic. The tariffs levied at the frontiers of the old Pergamene kingdom were probably the model for new internal tariffs, like the 2.5 per cent tax on goods going in and out of the Gallic provinces created in Augustus’ reign. To begin with this was levied by publican companies, then by individual contractors (in both cases under the supervision of equestrian procurators, but the latter much easier to control), and finally was handed over to state officials in the late second century.28 That evolution perhaps illustrates a wider trend towards increased central control of taxation, even before economic contraction and the onset of the military crisis in the middle of the third century.
Fig 15. The tax law of Ephesus (now in Ephesus Museum)
Yet local diversity persisted, wherever local systems worked. The tax systems created in Egypt by the Ptolemies were far more complex than those of Sicily, and were administered by a bureaucracy headed by a chief minister in the palace. That system too was simply swallowed up after Actium. The royal Idioslogos was now an equestrian official who reported to the prefect of Alexandria and Egypt rather than to the monarch. Other examples abound. Neither Republican generals nor the emperors seemed to feel much need to impose uniform fiscal systems across their domains, so long as the provinces provided what was needed.
Not all parts of the empire had fiscal systems that could so easily be plugged into that of the empire. It is not clear that private property, in the sense that we or Romans would have understood it, even existed in some areas of the north and west before the Roman conquest. Caesar mentioned some customs dues levied by Gallic tribes, but most exchange was not monetized. Here the Romans were forced to develop new mechanisms. A partial exception in the west was the former Carthaginian territory in Spain and North Africa, acquired by Rome after the second and third Punic wars respectively. The silver mines created around Cartagena by Hannibal were confiscated as state property: they were then exploited by hundreds of small-scale contractors using slave labour. Equally the tribute in olive oil paid by the cities of Tripolitania to Carthage was diverted to Rome. But in many areas Romans had to improvise, usually driven by the need to feed and pay armies that might be in the field for much longer than one campaign and might have lesser expectations of booty. It was during the long campaigns in Spain that the pressure was first felt. Local populations had been required to provide subsistence for the armies since the Punic Wars: irregular levies were also made on allies to pay the troops. At some point during the early second century BC these irregular levies became formalized into regular annual levies of cash and grain.29
The reign of Augustus was a period of fiscal reorganization. Even before the military disasters in Germany towards the end of his reign, it was clear that Rome could not keep expanding forever. And the Augustan solution to civil war was expensive. At the centre a military treasury was created in AD 6, funded by a 1 per cent sales tax and a 5 per cent inheritance tax introduced to provide hypothecated income from which veteran discharge bonuses were to be paid. Eventually around 75 per cent of imperial revenue would be spent on the army.30 So great property assessments were made in many provinces during the 20s BC, renewed in theory at regular intervals thereafter, and permanent tax liabilities were fixed on entire communities. The language used was that of the census, but this was not an exercise in distributing political rights and responsibilities of the kind conducted by Republican censors. Many areas were given a cash assessment, others were instructed to pay taxes in kind. Particularly common were taxes of grain, most likely delivered to nearby military camps. One procurator was based in Trier on the Moselle Valley and coordinated imperial finances across northern France (the province of Belgica) and the Rhineland where around a third of the imperial army was based in the first century AD. It looks very likely he was responsible for local provisioning from the grain lands of the Paris basin. Some tax assessments were more unusual. The Frisians at the Rhine mouth were assessed in cattle hides: this made perfect sense in terms of the dependence of their local economy on stock raising and also in terms of the military need for leather. Other pragmatic idiosyncrasies existed in other parts of the empire.
Vast tracts of new territory were conquered during Augustus’ long reign. Much of it was simply subjected to taxation, with new civic territories created in tribal land rather as Pompey had done in Pontus and tribal leaders given the same governmental responsibilities as the civic elites of Asia and North Africa. A certain amount was expropriated to settle troops. Augustus also retained land for himself, especially former royal lands. Increasingly, the same equestrian procurators who supervised provincial taxation oversaw this imperial property in the provinces. Emperors controlled the greatest quarries, the mines, and much else besides. Some of their vast agricultural holdings had been confiscated from senators, some inherited from their predecessors. At Rome, a scrupulous distinction was maintained between public and imperial finances, but there was no independent oversight of either and the distinction was most useful because it allowed emperors to pose as personal benefactors.
Augustus did not change the taxation of the empire overnight. The terms ‘conquest state’ and ‘tributary empire’ describe ideal types in a sociological sense. From the second century BC, Rome exhibited features of both, but it moved steadily in the direction of a sustainable tributary economy. Booty remained a major objective of campaigning well into Augustus’ reign and occasionally beyond: Trajan’s great building projects in Rome were funded by his wars in Dacia. But the long-term trend was towards the sustainable income that only taxation could provide.
The early imperial Roman tax system strikes us as unnecessarily complex, and yet it did certain jobs well. Being built up of smaller regional tax systems it was able to accommodate the major variations in the economic life of Italy and the provinces, including differences in the level of trade, in the nature of land tenure, and even in farming practices. Uniformity matters most either in terms of distributional justice—not a major concern for the emperors—or if it enables centralized bodies to act more efficiently. But little was centralized. However, there were drawbacks. For a start, there was constant uncertainty over which taxes were current: this is useful for historians who rely largely on monumental epigraphic clarifications to reconstruct the system, but it must have wasted much time. Second, there was a tendency in all systems for the weight of taxation to be pressed on the poorest, since exemptions were commonly given to already privileged groups and individuals. Given the vast inequalities of wealth in the empire, the emperors could only really afford to let the rich off lightly when times were good. Lastly, there was little flexibility. Medieval European monarchs could call a parliament to get new taxes, modern governments adjust rates very regularly, but Roman emperors had no way of doing so; indeed they were vulnerable in tithed areas to poor harvests, and through indirect taxes to decreases in the volume of trade. When times were not good, the fiscal inflexibility of the imperial tax system would catch them out. As the economy went into contraction things got considerably harder.
The Roman economy has attracted some of the most imaginative work of recent years, and it has also been able to draw on large amounts of new archaeological data of very high quality. The most recent global assessment is contained in the Cambridge Economic History of the Greco-Roman World (Cambridge, 2007). These essays have not superseded the excellent chapters on economic matters and imperial finances in volumes x–xii of the Cambridge Ancient History. William Harris has made major contributions to both projects, and the subject more generally: the most important are now collected in his Rome’s Imperial Economy (Oxford, 2011). The first products of a major new investigation have been published as Quantifying the Roman Economy(Oxford, 2009), edited by the project directors, Alan Bowman and Andrew Wilson. Archaeological contributions to the debate are surveyed in Kevin Greene’s Archaeology of the Roman Economy (London, 1986) and also a collection edited by David Mattingly and John Salmon entitledEconomies beyond Agriculture (London, 2001). It is probably fair to say the highly original theses about economic life included in Peregrine Horden and Nicholas Purcell’s The Corrupting Sea (Oxford, 2000) have yet to be fully assimilated into these debates. Christopher Howgego’s Ancient History from Coins (London, 1995) is not just about Rome or just about the economy, but has fascinating things to say about both, and says them with wonderful clarity.
The study of taxation is based above all on papyrological and epigraphic material. Perhaps for that reason discussion has been mostly technical and mostly published in journals and conference proceedings, although a marvellous recent exception is provided by Michel Cottier’s The Customs Law of Asia (Oxford, 2008). The best introductions are Peter Brunt’s article ‘The Revenues of Rome’ originally published in the Journal of Roman Studies 1981 and reprinted in his Roman Imperial Themes (Oxford, 1990) and Dominic Rathbone’s chapter on ‘The Imperial Finances’ in Cambridge Ancient History volume x. Probably the most influential contribution to this subject was a paper by Keith Hopkins entitled ‘Taxes and Trade in the Roman Empire’ published in Journal of Roman Studies 1980, arguing that the tax system of the early empire actually stimulated the economy. Hopkins returned to this theme on a number of occasions: his final version is most easily to be found (along with other important essays) in Walter Scheidel and Sitta von Reden’s collection The Ancient Economy (Edinburgh, 2002).