Dominic w. rathbone
I introduction
This chapter aims to assess the extent, nature, and causes of economic change in Egypt in the first three centuries of Roman rule (30 bc — ad 284). I will argue for significant aggregate and per capita growth in the first two centuries, attributable to the institutional, commercial, and behavioral impact of integration into the Roman world; then, following the Anto-nine plague, some aggregate decline in production but renewed, if more differentiated, per capita growth attributable to internal socioeconomic changes. The issues of continuity and peculiarity are fundamental. First, in any period we face the myth of the unchanging nature of Egyptian agriculture and agrarian culture, which hinders recognition of times of significant change.1 Second, Egypt has long been seen as an untypical province of the Roman empire: politically separate, geophysically distinct, culturally unique. So too, supposedly, it was kept outside the general economic structure of the empire, and is untypical for Roman economic history. Few historians now overtly subscribe to this view, but a sub-theme of this chapter is that Egypt is instead our best documented case ofthe normal regional variation within the empire-wide framework.
These problems envelop the papyrological evidence.2 The 20,000 or so Roman-period texts on papyrus published to date offer a quantity and range of documentary evidence unique in the Roman empire, and permit some quantification of social and economic phenomena. However, the relative wealth is poor and unevenly distributed. Over three centuries we have an average of twenty-one census returns per 14-yearly census, and one price of wine, the best attested commodity, for every two years.3 Half of the papyri
*In Roman Egypt land was measured by the “aroura” (ar.), equivalent to 2,756 m2. The standard dry measure was the “artaba” (art.), equivalent to 4.5 Italic modii, that is 38.78 l.; officially, one artaba of Egyptian wheat weighed 30.3 kg. The standard monetary unit of account was the Alexandrian “drachma” (dr.), notionally equivalent to the Roman sesterce (HS); see further below.
1 Debated in Bowman and Rogan 1999. The myth is in part orientalizing, but also reflects the politico-religious ideology of the Pharaonic state.
2 Montevecchi 1988 is the best guide to the documentary papyri, Bagnall 1995 to historical use of them. Habermann 1998 analyzes their chronological distribution.
3 Bagnall and Frier 1994, adding seventy-eight returns published since they wrote; Rathbone 1997a.
Are of the second century, and the bulk of them come from two nearby areas of Middle Egypt, the town of Oxyrhynchus and some villages on the edge of the Arsinoite nome. From Upper Egypt and the deserts we have some texts, mostly receipts, on ostraka (potsherds), from Lower Egypt (the Delta) some carbonized administrative documents, and from Alexandria almost nothing. We know that the nomes (regional administrative units) of Egypt retained some cultural and administrative differences through the Roman period, and the Arsinoite nome, as a semi-oasis (the Fayyum) developed by the Ptolemies, more than most. The restricted and, within that, haphazard documentation of topics by time and place makes it difficult to assess continuity and typicality. Archaeological evidence in the form of settlement and burial sites, and the material items which survive in them, is far more evenly spread throughout Egypt, but has only been studied patchily.2032
Ii demography
The demography of Roman Egypt is a topic of much recent debate. Josephus (BJ2.385), writing around ad 75, claims a total population of 7.5 million, excluding the residents of Alexandria. Some scholars, myself included, think this a traditional and gross exaggeration, while others insist that Josephus should be believed.2033 Comparison with nineteenth-century census figures and estimates of the ancient carrying capacity of the land are not certain or precise enough to settle the point. Data from a few Arsinoite villages in the mid-second century ad suggest the average rural population density, which excludes the nome capitals and Alexandria, did not exceed 120 persons/km2; although the validity of this crude estimate has been questioned, very much better data from the mid-third century bc, after implementation of the Ptolemaic scheme to develop and settle the Fayyum, point to an average density, including the nome capital, of only 60 persons/km2, which makes the Roman-period estimate look high rather than an underestimate.2034 For want of more persuasive alternatives, I stick to my previous estimates of a total rural population of 3 million, and up to 1.5 million inhabitants of Alexandria and the nome capitals, which, rounded up, gives a total population of around 5 million in the mid-second century. Settlement evidence (see below) suggests this was the peak of Egypt’s populousness in antiquity, apparently the result of a slow but steady rise from 30 bc. No one disputes that the Antonine Plague, which was carried into Egypt in ad 166/7, caused over the next decade a dramatic aggregate population loss, probably of around 20—30 percent to judge from some attested cases, including over twenty Delta villages.2035 Signs of partial abandonment of Arsinoite villages, and the growth of large estates (see below), hint that population regrowth may have been skewed in favour of urban centers. Study of the extant census returns, the richest standardized source of demographic data from Roman Egypt, points to a high mortality and high fertility regime, susceptible to crash in a crisis like the Antonine plague, but also capable of generating significant long-term growth.2036 Whether it was a high or low pressure regime will be considered after discussion of the evidence for production.
Iii land and agriculture
Egypt was famed throughout antiquity for its amazing agricultural output, the result of the annual Nile inundation with its rich silt deposit, which, unlike the Euphrates and Tigris spates, conveniently coincided with the sowing season for arable crops. The standard estimate of the total area potentially under cultivation is 9 million arouras (25,000 km2), based on a Ptolemaic temple text and one set of nineteenth-century figures, which looks optimistic against other late nineteenth-century totals of around
20,ooo km2 actually under cultivation. However, two regional totals for land liable to tax in kind, one for a division of the Arsinoite nome in ad 184/5 and the other a fourth-century summary for the whole Oxyrhynchite nome, backed by much wider archaeological and documentary evidence for rural settlement in middle Egypt and the Delta, indicate that the extent of cultivation, which peaked in the second century, was greater than in the nineteenth century, and not matched until more recent times.2037 There were no grand state-sponsored development schemes on the Ptolemaic or Achaemenid model; expansion must be attributed to private initiative, encouraged by fiscal policy. Plots of uncultivated public and private land which were confiscated, abandoned or otherwise ownerless, a chronically recurring phenomenon, were sold off or leased out by the nome authorities on very favorable terms through a well-documented, albeit unstudied, process of local registration and public offer to any reasonable bid.2038 The state also provided authority and backup for regional maintenance of the public canals and dykes by means of a communal corvee system.2039 Inadequate or excessive inundations were a regular threat to agricultural production, roughly once in every seven years, but, perhaps because of a favourable climatic macro-cycle, only once, in the late 240s to early 250s ad, is there evidence for a run of problems over several consecutive years, and even that, prices suggest, was weathered without serious upset.2040 The one major development in irrigation was the spread of the cattle-powered pot-garland water wheel (Arabic saqiya) to provide perennial water from wells or canals for walled enclosures (ktemata) used for intensive viticulture and the cultivation of fruits and legumes. Although it was a Hellenistic invention, documentary and archaeological evidence both indicate that its diffusion occurred through investment by large private estates in the third century.2041
The Roman government soon subordinated the varied traditional land categories of Egypt to its simple schema of private and public land, and eventually removed the residual restrictions to a free market in private farmland.2042 Even lessees of public land seem in practice to have enjoyed considerable security of tenure, apparently collectively in the Arsinoite nome but as individuals elsewhere.2043 Two Arsinoite village tax registers of the mid-second century show that 53—59 percent of their territory was public land, which may reflect their origin in an area reclaimed by the Ptolemaic state. Notionally public land in the fourth-century Oxyrhynchite nome comprised 19 percent of the total, and 24 percent of the farmland of a Mendesian toparchy, in the Delta, around ad 300. At Naboo, in Upper Egypt, in 118/9 royal and sacred land made up 26 percent of the “mainland” total, although the category of royal land classed as private shows that a total of 39 percent had once been public.2044 These figures imply a steady drift of public land into private ownership through the cycle of abandonment and sale, and we would expect this trend to have accelerated after the Antonine
Figure 26.1 Wallpainting of a water-wheel from the Wardian Tomb (Alexandria) (Reproduced with kind permission from Dr. Mervat Seif el-Din, Director, the Graeco-Roman Museum, Museum Street,
Alexandria, Egypt. Inventory number 27030)
Plague. It has been suggested that the Arsinoite fiscal report of ad 184/5 (see above), where only 27 percent of the wheat taxes due had been paid by the end of the civil year (late August), indicates a drastic reduction in the area actually cultivated, but this is far from certain.2045 By the fourth century, remaining public land had been handed over to private ownership, and was still registered as “public” only because it bore a higher tax rate in lieu of the previous rents (see below). This presumed reform of Diocletian was not radical, but consolidated and confirmed long-standing Roman policy.2046
A fiscal register of private landholdings at the Arsinoite village of Philadelphia in ad 216/7, in which 15 out of 197 landowners are categorized as elite outsiders, appears to show a relatively equal distribution of arable land with a Gini index of 0.532. Tax reports of ad 308/9 from nearby Karanis, after some padding out of gaps, produce a Gini index for private landholding of 0.478. But managers (phrontistai) of large estates are attested at Philadelphia then and soon after, whose holdings must lurk disguised in the village section.2047 Large private estates, with properties dispersed through a nome, begin to appear in the late first to early second century, such as the Oxyrhyn-chite estate of the Tiberii Julii Theones, but the emergence of really large estates, with substantial holdings in many villages of a nome, and comparable landowning probably in other nomes too, such as the Arsinoite estate of Aurelius Appianus (perhaps over 4,000 ha.) or the Oxyrhynchite estate of the Calpurnii Horiones, is a phenomenon of the earlier third century.2048 To some extent these large estates were the result of a slow process of accumulation going back to the first century. However, archaeological evidence for the wine-pressing installations which they used (lenoi, see below) suggests a particularly rapid phase of expansion achieved by acquisition and reclamation of private and public land abandoned as a result of the Antonine Plague. Two survey documents from Theadelphia reflect the change: in ad 158/9, of its 6,300 ar. or more of farmland, 8 percent was taken up by vineyards and orchard/garden land; by ad 216 the total farmland had, it seems, shrunk by 19 percent to 5,100 ar., of which 29 percent was now under vines and the like. The growth of viticulture is attested in the Delta too, in the Mendesian nome, where 24 percent of the farmland in one toparchy by around ad 300 was vineyards and garden/orchard land.2049
Agricultural experimentation was not as eye-catching as in the Ptolemaic period, because the main obvious changes from closer integration with the Mediterranean world had already been accomplished, notably the switch from emmer to naked wheats as the principal grain crop.2050 The standard, conservative, estimate is that the average wheat yield was tenfold, that is a gross crop of 10 artabas per aroura (roughly i tonne/ha.), as in the Ptolemaic period, although some scholars think this is much too low. There is plenty of direct evidence for a standard sowing rate of i art./ar., but remarkably little for yields. The best set comes from the third-century Arsinoite estate of Aurelius Appianus, where attested wheat yields on directly worked and leased lands ranged from 7.0 to i6.6 art./ar., with an overall mean average of 13.0 art./ar.2051 This supports the obvious expectation that yields will have varied widely depending on a variety of natural and human factors. A normal tenfold yield fits well with the indirect evidence of rents on state land, which averaged around 3 to 3.5 art./ar. and rarely exceeded 5 art./ar. (see below). Private leases of the Roman period, from both the Arsinoite and Oxyrhynchite nomes, tend to specify a biennial rotation of wheat and hay, or legumes, with a high wheat rent, averaging 7.7 art./ar. in the Oxyrhynchite leases but in one Arsinoite group reaching 15 art./ar. in one year and a small cash rent, or no rent, in the other year. The rationale and reality of these rent arrangements remains a puzzle (possibilities include delayed payment in wheat, or part-commutation to cash, and so on), but, because they represent two years’ output, they do not necessarily imply much higher than tenfold annual yields. Biennial crop rotation seems to have been the norm on private land in Roman Egypt, as specified in leases and attested in private accounts for land worked directly. The low taxes assessed in wheat could easily be met from the portion of land under wheat, or by other means. Crop rotation was probably less common on state land, whose annual rents, mostly set in wheat, were much higher and were normally, it seems, paid in wheat.2052 Broadly, we may hypothesize that the average wheat yield on private land was higher than that on public land, say twelvefold rather than tenfold. A crude estimate of the average gross annual agricultural production of Roman Egypt in wheat equivalent (assuming 8o percent of 9 million ar. under cultivation, of which 25 percent was public land, and no fallowing) would be 83 million artabas (2.5 million tonnes).
The evidence for agricultural practices, as for other topics, is patchy and localized, which makes it difficult to discern trends. Some practices may escape documentation. For instance, there is extremely little evidence for double - or catch-cropping in comparison with mediaeval Egypt, but maybe only because it is not mentioned in the lease contracts which are our prime source of information on cropping; in fact hay, which is often specified in leases, was often cropped two or three times a year.2053 The level of wheat rents in private leases rises to a peak in the second century and drops back in the third, but this probably reflects the changing availability of tenants (that is, labor; see below) rather than productivity. Some interrelated developments do seem attributable to the growth of large estates in the third century: greater use of animals in irrigation, cultivation, and transport (see below);
Probably therefore more manuring, especially of the artificially irrigated ktemata (walled enclosures); diffusion of the use of scythes (chortokopa, “hay-cutters”) not just for hay, but for grain crops in place of traditional sickles (cutting near the ear and ploughing in the stalk) or uprooting, perhaps deliberately to harvest more straw for the animals; proportionately greater production of fodder crops, which may also have contributed to the fall in wheat rents and more complex specification of rotation in third-century leases.2054 While we might expect a certain drop in aggregate production due to abandonment of some land after the Antonine Plague, if the earlier third century also saw considerable conversion of public land to private ownership, we could posit a significant increase in productivity due to more diversified and intensive cultivation, perhaps accompanied by a slight rise in the average wheat yield. As already noted, the main development was the enormous expansion of viticulture, which, as the Heroninos archive illustrates, involved investment and the widespread application of technical improvements. Vines were carefully selected, tended, and harvested to produce specific types of wine, they were cultivated intensively in walled enclosures irrigated by cattle-powered water wheels drawing water from stone wells, the grapes were pressed in screw-presses in new pressing rooms (lenoi), built in Roman style of fired brick, mortared and lined with hydraulic cement, and the wine was bottled in clay winejars, which, at least around Alexandria, imitated the Italian form Dressel 2—4. Remains of the virtually indestructible lenoi are found throughout Egypt, confirming the typicality of the evidence of the Heroninos archive, and they were probably common in other wine-producing areas of the later Roman world, for the fourth-century agronomist Palladius explains how to construct one.2055
Iv urbanization and craft production
If we had no papyri, we would still, from the archaeological evidence, unhesitatingly identify urbanization as one of the main socioeconomic developments in Roman Egypt, as it was in most provinces of the Roman empire. The main visible traces on surviving sites of the almost forty metropoleis (nome capitals) are of Roman-period buildings, sometimes as reused in the Byzantine period. Adding the evidence of the papyri confirms that the second century saw a boom in the erection of public buildings, notably theaters, bath-houses and temples, mostly paid for out of civic resources. This continued into the third century, when big programs of repair and embellishment are also attested, notably at Hermopolis Magna.
Many of these buildings, like the theater at Oxyrhynchus, magnificent in its size and decor (probably the largest in Roman north Africa), and the urban water-supply system of Ptolemais Euergetis (capital of the Arsinoite nome), represent the large-scale application of sophisticated engineering techniques.2056 We have little evidence for developments in this period at the two main “Hellenic” foundations in Egypt of Alexandria and Ptolemais in Upper Egypt. In the Roman period Coptos seems to have become the chief city of Upper Egypt. Alexandria was no longer a royal capital and focus for lavish royal spending, but instead it became the main transport and trade center of the eastern Roman empire, and the fragmentary architectural remains imply considerable public and private building activity.2057 The rise of the metropoleis, which are normally imagined to have been over-sized villages in the Ptolemaic period, was initiated by the Roman sociopolitical policy of creating an urban “Hellenic” elite to provide the regional and civic liturgic functionaries required by a more Roman administrative system.2058 However, there was from the start an economic element and impact. The definition of this elite was tied to the privatization of catoe-cic land (see above), and the urban residents rapidly developed patterns of consumption, with the associated production and trade, typical of Roman provincial urbanism (see below). Good data for the size of urban populations are completely lacking, but in a few cases we can make guess-estimates from recorded numbers of houses.2059 Some metropoleis were relatively small, like Apollonopolis Heptakomias in Upper Egypt with around 7,000 inhabitants, while in Middle Egypt Ptolemais Euergetis and Oxyrhynchus were probably more typical with about 20,000 each, and Hermopolis Magna unusually large with maybe up to 40,000 residents. Roman Alexandria, the second largest city of the Roman empire, probably had a population of up to 0.75 million (larger than in Ptolemaic times), while villages in Egypt had populations ranging from a few hundred to around 4,000. Omitting these larger villages, although they were as big as “cities” in other provinces, we can estimate that from 20 percent to 30 percent of the population of second-/third-century Egypt lived in cities, which makes Egypt one ofthe most urbanized provinces of the Roman empire. The material vitality of the cities in the third century, allied to the evidence for limited rural repopulation, suggests that the percentage probably increased after the Antonine Plague to the 30 percent end of the range.
The economic life of the cities of Roman Egypt is difficult to reconstruct because the documentation is too particular and fragmentary to permit much generalization. Particularly frustrating is the poor evidence for Alexandria. Textile production is the only craft to have been studied seriously so far, albeit only from the documentary evidence, and there has been very little research into the distribution and marketing of goods. As the normal residence of most large and medium landowners, we would expect cities to have been the major centers of consumption in Egypt. The papyri attest a wide variety of craftsmen and tradesmen, but so sporadically that attempts at quantitative analysis are futile. The basic trades, such as carpenters and wine sellers, are found in villages, but many more, and more specialized, trades were based in the cities. As elsewhere in the Roman world, many “service” occupations are attested, such as the scribes, doctors, musicians, wet nurses, doormen and so on, employed privately and by temples and civic institutions. As elsewhere, textile workers dominate the record. One indication of scale is given by a second-century register of customs duties which shows that over five days in one November 1,956 items of clothing, mostly Greek-style tunics (chitons), were shipped out of Oxyrhynchus, presumably towards Alexandria and the Mediterranean. Egyptian weavers, like their counterparts in other provinces, mostly produced finished items of clothing rather than, as in mediaeval practice, rolls of untailored material. A speciality of Egypt, and probably a major export to the Mediterranean (and India), was linen, and, for what it is worth, there seems to be more evidence for flax cultivation in the third century. In terms of urban employment, unskilled and semi-skilled occupations, such as construction labor, were probably equally or more important, but they are seriously under-represented because they did not generate written contracts or receipts. The Roman period saw a great boom in the quarrying of marble and granite, and some growth in the mining of emeralds and gold, which was all run on private lines, apart from military protection, by the patrimonium (private estate) of the emperor.2060
Retail traders in designated markets were subject to small fees for the right to sell, some flat rate, some pro rata, due to the local civic or religious authorities; for selling in the nomes, a territorial “concession” had to be purchased. Sometimes the authorities also tried to impose conditions about supplies or prices. Artisans, and some other professionals such as prostitutes, were liable to annual fixed capitation taxes for their crafts, called “handicraft-taxes” (cheironaxia), due to the Roman administration. Like the poll tax, with which they were collected, the rates varied from nome to nome; Arsinoite weavers, for instance, paid 38 or 76 dr. a year.2061 Craftsmen often operated from a one-room workshop-cum-shop opening onto the street, the common ancient, mediaeval, and modern habit, which implies small units of production. The new large estates of the third century employed some inhouse craftsmen, especially carpenters, but still relied mostly on ad hoc contracting of independent artisans. If we knew more about Alexandria, where, in broad terms, we know there were important industries producing glassware and perfumes, we might find larger scale enterprises with structured workforces like the workshop with “many” linen workers attested in one isolated text. They must have existed throughout Egypt for construction (no evidence), shipbuilding (one attestation), and so on, and for mining and quarrying, as is now well attested in the eastern desert, especially at Mons Claudianus. Contracts of apprenticeship appear in the papyri for both freeborn youths and slaves, but slave labor seems usually to have been supplementary to free labor, even in the quarries.2062
Some cities had quarters which were named after types of craftsmen or traders, but although a few cases of loose grouping are known, like the weavers around the Serapeum at Oxyrhynchus (and kilns were always on the edges of villages), normally there was no particular zoning of trades. Even small villages had designated and organized areas for selling food, foodstuffs and other items, principally, by long tradition, in and around temple precincts and along their processional avenues (dromoi), such as the big Serapeum market in Oxyrhynchus. Many cities and larger villages also had a Greek-style colonnaded “market” (agora), or several. Annual religious festivals could be accompanied by a special general fair (paneguris), and some villages had periodic specialized markets, for instance for trading camels.2063 Most craftsmen and some retailers seem to have belonged to trade “associations” (koina, or sunodoi, better not translated as “guilds”), which had club rules, annual subscriptions and elected officers. Their best attested functions were social and religious: to hold dinners, to confer a sense of importance, to help with deceased members’ funerals, and so on. However, they also arranged and managed agreements to fix prices or divide market territories, and they often dealt on behalf of their members with the local and provincial authorities, for instance collectively paying their members’ cheironaxia, or providing supplies for the Roman army. In this period, before the fourth century, it seems to have been advantageous, but not a state requirement, to belong to an association, and associations had not yet been made regularly subject to collective obligations imposed by the state.2064
Attempts to quantify the percentage of textile workers, for example, or textile output, are illusory at present. Clearly many urban workers, and some in bigger villages, were in apparently “non-productive” occupations which converted the agrarian profits of landowners into items of consumption and display. But this also created wealth, by stimulating competition and demand, and because the processes of transformation (from wheat, say, to painted wallplaster) were complex and monetized. Equally clearly, there was a substantial urban production which created wealth more directly, partly by exports to Alexandria and beyond, and partly by sale to the villagers who produced food. Although loomweights are as commonly found in Egyptian villages as on rural sites in the rest of the Greco-Roman world, the papyri show that peasants purchased clothes (and funerary textiles), and burials and rubbish layers are full of mass-produced jewelry, amulets, figurines, and other small items, which were probably mostly of urban manufacture (see below). Technological developments are hard to trace. Most weavers stuck to the traditional Egyptian horizontal loom, and use of the “Alexandrian” upright loom was a speciality. The quality of terracotta and faience figurines dropped visibly after Roman annexation because of a change to mass production techniques. The skilled techniques used in the glass industry were of Hellenistic origin. Crudely, the scale of urbanization in Roman Egypt, and contemporaneous prosperity ofthe rural population, implies that cities were producers ofwealth as well as consumers ofagrarian surplus.2065
V trade and transport
Ifsurplus production, agrarian and urban, was to create wealth, it needed a distribution system. The number ofspecialist food sellers attested in cities and larger villages show that local retail networks existed, and accounts of personal and institutional expenditure, like those of Kronion the supervisor of the record office of Tebtunis in the 40s, or those of the estate owner Sarapion in the earlier second century, show a daily routine of small-scale purchases for cash.2066 Literary, documentary, and archaeological evidence illustrate regional and interregional distribution through, for example, finds of red slip pottery from Aswan (which imitated North African finewares), the records ofinternal customs posts, shipping contracts, listings ofregional specialities.2067 But the main change to trade in Egypt in the Roman period was the development of Alexandria as the commercial center of the eastern Mediterranean, mediating the east-west flows of goods and wealth, which often, it seems, passed through Alexandria rather than being shipped directly between Italy and Asia Minor or the Levant. Visitors marvelled at the number and varied origin of the ships and merchants in Alexandria’s two enormous harbors.2068 One foundation of this situation was the imperial annona, the grain supply of Rome (see below), which chartered, and thus subsidized, an annual sailing of large private ships from Alexandria to the ports of Rome (first Puteoli, later Ostia). With no official cargo for the return voyage, they carried back Italian wine, Spanish and North African olive oil, and craft products such as North African pottery and Italian metalwork. Some of these wares were re-exported to other provinces, while some, as documents and archaeology show, percolated south through Egypt, and even reached India. However, there is also some evidence for a considerable private export of wheat and other produce to Italy, and presumably there were exports to other areas too where local supplies were unreliable.2069
Most spectacular was the trade with Arabia and India, which is particularly well documented in the first century, but also flourished in the second to third centuries despite the disappearance of texts on ostraka (a change in writing practices?).2070 Although there was some export of produce and manufactured items such as wine and textiles, it is clear that the main exports (say 50 percent?) were gold and silver, to begin with as coin, subsequently mainly as bullion. The principal imports were perfumes and spices (especially pepper), precious stones, ivory, and some textiles, including Chinese silk, and tropical hardwoods. This is normally termed a “luxury” trade to belittle its importance. Admittedly these goods were for consumption and display, but in the developed urban, or rural urbanizing, lifestyle of the empire pepper was a basic condiment masking poor meat, perfumes were equally commonly used in domestic life and in private and public religious rituals, and jewelry inset with gems and small items made of ivory were affordable to reasonably prosperous villagers (see below). The pseudostatistic derived from the elder Pliny that in the later first century Rome was losing not less than 50 million sesterces a year on this trade through Egypt is an academic red herring. The only certain figure is that the return cargo of one ship from India in the mid-second century, after payment of the 25 percent import duty in kind, had an agreed market value at Alexandria of 7 million drachmas. At the contemporary median price this was worth almost 770,000 art. of wheat in Middle Egypt (over 23,000 tonnes), the net product (deducting 2 art./ar. tax and seed) of 77,000 ar. of private land, almost I percent of the productive arable land of Egypt.2071 The wealthy Italians and Alexandrians who could afford to finance trade on this scale will have profited most from it, but there was a substantial trickle-down effect. Some goods were manufactured at Alexandria, Coptos, and elsewhere, for export. The imported perfumes and spices generated a specialized processing industry, including adulteration and faking, for which Alexandria was the main center. The trade employed independent skippers and their staff, shipbuilders and repairers, donkey and camel trains across the eastern desert, and thousands of porters at transit nodes. Some of these worked full time in the trade, but for most it was a seasonal bonus in addition to their normal farming or employment.2072
Transport in Roman Egypt was probably as easy and efficient as at any time before the steam age. The potential of the Nile to link the country was fully realized. Nile ships of the period ranged in size from 1.5 to 150 tonnes burden, with many of 15 to 45 tonnes. Most were Greek-style “galleys” (akatoi), built of wood by the shell method (which economized on wood), powered by sail and oar, though traditional Egyptian reed boats were still also used. The monster grain barges, from 180 to 540 tonnes burden, of the Ptolemaic era had disappeared along with the queens and courtiers who built them as status symbols. In the Roman period most ships, including seagoing ships, were of a medium size which a single owner-captain could finance and manage, and even larger ships of200 to 400 tonnes burden, like some annonal grain freighters and the ships needed for the monsoon ocean crossing to India, were not that expensive. Furthermore, a special contract of “lease-sale” of ships, of Hellenistic origin, enabled an investor to build a ship and “lease-sell” it to a captain for a share in the operating profits.2073 Very little is known about the maintenance of roads. The government did maintain the routes through the eastern desert, and equip them with walled caravanserai, and Hadrian built a new road down the Red Sea coast from his foundation of Antinoopolis, but all this was mainly for access to the mines and quarries and to help police the area. To meet the demands of increased quarrying and mining the camel was used more and special wagons and wheeled cradles were introduced. The use of camels and four-wheeled carts, which could carry respectively twice and four times as much as a donkey, first became common in the Egyptian countryside in the third century on the new large estates.2074 This indicates a growth in the quantity of surplus produce to be transported to market, and implies a greater percentage of land being devoted to fodder crops.
Vi prices, prosperity, monetization
The economy of Roman Egypt was highly integrated and fundamentally free market. Political unity and general peace helped, along with standardized systems of law and justice, and of money, weights, and measures. Major economic trends, such as urbanization, privatization of the land, growing wine consumption, were ubiquitous, like social and cultural trends in, for instance, hairstyles and burial customs (see below). Increasingly, and especially in the third century, the Alexandrian and nome elites merged into a provincial elite, with large landholdings over several nomes. Urbanization implies constant migration from the countryside, and the papyri attest a high level of individual mobility, including permanent settlement, and seasonal and occasional movements.2075 Through the taxman’s eyes this mobility looked like anachoresis, flight to avoid paying taxes, but it made for a free market in labor and contributed to economic growth. Agricultural accounts all show a wide variety of types and rates of remuneration for long-term and casual labor, reflecting a changing and complex market as well as the different requirements of different jobs. Long-term laborers often received accommodation and wheat rations, but most remuneration was in cash, and therefore responsive to short-term fluctuations ofdemand and supply.2076 Slavery existed in Roman Egypt, and was relatively common in the domestic context, but the systematic use of trained groups of slaves in agriculture and urban production which was typical of Italy, is not found in Roman Egypt.2077 Attested prices of wheat, wine, and donkeys, which also mostly come from Middle Egypt, display the characteristics of free market prices: they co-vary broadly in the long term; in the short term they behave differently; those for wheat and wine fluctuate between years and seasons according to harvest and stock levels. No market is perfect, but the round numbers and limited range of wheat prices reflect the stabilizing effect of fairly heavy state intervention in the buying and selling ofwheat (see below). Wine prices, uniquely, have a slow upward trend throughout the period, suggesting that increasing, and increasingly sophisticated, wine consumption was leading wine production. The price bands for other goods and wages display a remarkable stability from the ad 70s to the i6os, and then again from the 190s to ad 274. This indicates maintenance of a long-term balance between internal production and consumption, and unwavering confidence in the token coinage (see below). The sharp doubling of prices and wages in the later second century is best explained as a sign of temporary economic dislocation caused by the Antonine Plague. The system then regained its normal equilibrium until it was rudely disturbed by the coinage reforms of Aurelian and Diocletian.2078
In the absence of decent statistics, attempts to calculate living standards in monetary terms are unsatisfactory. We can estimate that in the second to third centuries it took a rural laborer around seven days’ full employment to earn the median price of an artaba of wheat, roughly equivalent to his food requirement for one month. But his tax dues and other living costs are difficult to price, and most households lived from a diverse portfolio of activities, often founded on farming a plot or two of land. A broader, if more impressionistic, approach can help. Roman Egypt had a small elite, part of the elite of the empire, with similar interests, activities, literature, food, houses. The “ordinary” people are of more interest. There were two serious revolts in Roman Egypt: that of the Jews in ad 115—17, and that of the Boukoloi (“cowherds”) in the Delta around ad 171—2. The causes of the former were religious and ethnic rather than economic; although the latter surfaced in the wake of the Antonine Plague, its motivation is opaque, and may well have been millennarian rather than to do with taxes or oppression.2079 In petitions individuals sometimes complain of being ruined by their enemies or a disaster, but references to chronic destitution are non-existent, in striking contrast to late antique Egypt. Conversely, village cemeteries in the Fayyum show an extraordinary range of grave goods. Even bodies dumped in shallow graves on their palm-frond bier tend to have jewelry. Metropolites and richer villagers constructed plastered brick tombs to house their splendid mummies with painted or gilded masks, or, in particularly Romanized places like Philadelphia, inset naturalistic portraits. Many of the portraits represent these people finely dressed, coiffed, and adorned in the latest (more or less) imperial style, as if for one of the dinners to mark rites depassage, now including birthday parties (a Roman custom), with entertainers, flowers, and refreshments, now principally wine and pork (the Roman taste), held at home, or, more often, in special dining rooms in temple complexes, to which they were constantly inviting each other. Houses of the Roman period, mostly built in the traditional style out of mudbrick (still prized today for its insulation, solidity, and ease of repair), tend to be larger and grander than those of the Ptolemaic period, and the better ones had stone elements and were decorated with painted wall plaster with plant and animal motifs, scenes from Greek mythology, and images of Greco-Egyptian deities like Sarapis.2080 Many domestic utensils were made of wood, fibers, or clay in traditional style. Iron, which Egypt lacks, continued to be rare, to the extent that flint knives and scrapers were also used. The small quantity of imported ceramic finewares, in contrast to the preceding and following periods, implies a widespread use of more expensive glassware, under-represented in the archaeological record because broken pieces were recycled.2081 Change across time is even harder to assess. It has been argued that third-century wages were relatively higher against prices than before the Antonine Plague, because the labor supply was lower, which is possible but not yet convincingly documented.2082 However, the fundamental trend in the third century, seen elsewhere in the empire, was towards greater differentiation between rich and poor, not least as land was privatized more rapidly and absorbed by large estates. In demographic terms, Egypt of the first to third centuries was a low-pressure regime, with a low population in relation to aggregate resources, especially in the second century, and a tolerably open system of entitlement to the basic resource, land. After the Antonine Plague, population seems to have recovered faster than land was reclaimed, and was more urban based (allowing the possibility of some shortage of rural labor), which initially was matched by agricultural intensification (higher per capita output). But, by the fourth century, as access by the poor to land became more restricted and the urban economy began to falter, a higher pressure regime started to develop, one of whose features would be a low but chronic level of urban poverty.
By pre-modern standards the economy of Roman Egypt was highly monetized, and probably increasingly so from the first to third centuries. Wheat was still also used to store wealth and as a medium of exchange, principally in peasant autarky, taxation in kind of land, private “giro”-payments of wheat using deposits in state granaries, and rations for longterm employees on private estates.2083 But this was an adjunct to the money economy, often interlinked with money payments, or commutable into them. Cash, credit, and monetary valuations permeate the papyri. Coin finds on rural sites are sparse but ubiquitous (there were few low-value coins in circulation to be dropped and not recovered), and the large dumps of demonetized late third-century tetradrachms show that coins had been in widespread use. Villagers paid some taxes in cash, rented some land or accommodation for cash, labored for cash, bought goods, materials, and foodstuffs for cash, and often took out small loans, or sold crops in advance, to meet cash-flow problems.2084 There was never enough coin in circulation to cover these transactions. The Romans had adopted the coinage system of the late Ptolemies, based on the Alexandrian tetradrachm, a billon (silver-tin) coin, with a secondary series of copper drachma units, never minted in large quantities.2085 The tetradrachm was a token coin, with much less silver content than the denarius, but notionally equivalent to it. Hence the tetradrachm circulated exclusively in Egypt, but physical transfers of money between Egypt and the rest of the empire were possible using mainstream imperial gold coins. Minting at Alexandria, as at Rome, was very irregular. It is normally supposed that the quantity of coin in circulation grew gradually but steadily, but it seems there were only two major bursts of output, under Nero and under Trajan to Hadrian, which included much reminting of earlier issues, and new minting may have achieved no more than to replace previous cumulative coin loss. In the late second and third centuries minting of copper coins faded out, and the silver content of the tetradrachm was reduced in steps to a mere wash, parallel to developments in the coinage of the rest of the empire. Prices, however, did not rise, probably because the government did not raise tax rates and took back its own coin at its face value, thus maintaining confidence in its token value. Monetized exchange functioned, and grew across time, through credit arrangements. At a basic level purchases and payments, including tax payments, were recorded on account and partially settled at irregular intervals in multiples of the tetradrachm. For larger transactions credit transfers could be made between individual accounts in private banks. Private and state banks, the latter purely for receiving tax dues and making public expenditure, existed in every city.2086 Contracts could be registered with private banks in place of the normal notarial authorities, presumably when the banks handled the money-transfers in the contracts, which implies common use of banks. Although most private banks were local to one city and there was no general clearing system, it seems to have been possible to transfer money from an account in one bank to an account in another.
Vii taxation and the state
No plan for the economic development of Egypt was ever drafted by the Roman government, but it had beliefs, practices, and interests which favored general prosperity and encouraged some economic developments, and which, whatever their immediate motivation, were recognized by rulers and subjects to have these effects. Taxation in Roman Egypt is an enormous and complex subject which is overdue for renewed study.2087 Arable land was assessed in kind, mainly wheat, on a system based on that elaborated by the Persians and Ptolemies from Pharaonic practice. The normal rate for private land was i art./ar. Public land attracted a rent which varied, in theory, according to the Nile flood. A normal range of rents from 2 to 7 art./ar. is attested, with a median of around 3 to 3.5 art./ar. The general trend in Roman times was to simplify by assessing average rates, probably first in target totals for nomes and villages, by the fourth century directly for all landholdings.2088 Tax on garden land was paid in cash. All male inhabitants aged fourteen to sixty-five of Egypt, excluding citizens of Alexandria and of Rome, paid an annual cash poll tax (laographia), a novelty introduced to Egypt, and the rest of the empire, by Augustus. The basic rate in most nomes was 16 dr., but there were inexplicable variations, including the uniquely high 40 dr. in the Arsinoite, and metropolites paid half the basic rate. Urban crafts and trades were liable to trade taxes (cheironaxia), which were a set annual sum per person.2089 There were also imperial customs dues, which were farmed to private contractors: a 25 percent levy in kind on imports into the empire, an internal cash duty of around 1.5 percent on goods moved between the Nile valley and the Delta (the “Memphis harbor” tax), and, probably, a 2.5 percent cash duty on goods moved between Egypt and other customs zones of the empire. There were also various sales taxes, often of i percent, which mostly, it seems, went to the civic authorities.2090 Fragmentary evidence and local variation make it impossible to estimate the tax burden by itemization, but some gross figures suggest that, despite major fiscal reforms in the early second and mid-third centuries, the total burden remained remarkably stable, and relatively low, from Roman annexation through early Arab rule. It also seems that much of the theoretical imperial tax-take was spent within Egypt on maintaining officials, army units and temples, or just went uncollected, and at most half was exported to Rome, of which about half was wheat for the annona, whose importance historians habitually exaggerate, and half in cash.2091 Civic revenues were spent locally on building and repair programs, the running costs of gymnasium complexes and civic temples, mounting theatrical, athletic, and other shows and spectacles, and providing distributions of wheat, and sometimes oil and pork, to privileged groups of residents.2092
Roman taxation in Egypt was relatively fair in that only the poll tax was, in very broad terms, regressive. Both land taxes and trade taxes were fixed in practice, which gave an incentive to increasing production. The state never managed, or even sought, to collect all that it needed in kind: throughout the period it resorted regularly to compulsory purchases of wheat, and of all sorts of supplies for the army, at fixed but not unfair prices.2093 Government edicts, echoed by petitions from private individuals, constantly claim that the agricultural productivity of Egypt was its aim, and auctions of abandoned lands and other ad hoc tax remissions were practical manifestations of this.2094 This reinforced the general Roman ideology of good government, protecting individual property rights against officials and other individuals, whose effect was that documents about legal disputes constitute a high percentage of the surviving papyri. This should not be mistaken for a policy to protect peasant farmers; as had already happened in Italy, the Roman belief in private ownership facilitated the growth of large estates by legal acquisition and a consequent decline of independent peasant farmers. Apart from its compulsory purchases and control of mines and quarries, the Roman government intervened very little in the economy. There were imperial and civic checks on weights and measures, wheat and other distributions at Alexandria and some metropoleis, and various attempts to limit profiteering in times of dearth, but only a single case is known of the governor threatening to purchase all private stocks of wheat at a set price for re-sale.2095 Attitudes to trade were neutral: the 25 percent levy in kind on imports at the imperial frontier was high, but internal customs dues were few and low; transport was facilitated by the maintenance of roads, and suppression of piracy and banditry, even if these had other aims too; apart from mines and quarries, there was no attempt to create and profit from monopolies in production or trade (trade taxes and retail fees did not confer exclusive rights).
Roman provincial justice was flexible and, under the umbrella of Roman legal institutions, recognized local legal traditions and the hybrid contracts which often resulted, and was prepared to judge cases by precedent and common sense.2096 In modern terms, transaction costs were low.
Viii conclusion
Like every province of the empire, Egypt had its peculiarities. The most striking economic oddities were the size and wealth of Alexandria (whose impact, through lack of evidence, is always underestimated, here too), the agricultural fecundity gifted by the Nile inundation, and the closed silver and copper coinage systems. But none of these cut Egypt off from the rest of the empire; indeed Alexandria and the wheat surplus were integrating factors. Rome drew Egypt in further. A simplistic “taxes-and-trade” model, whereby Egypt was stimulated to produce surplus goods for export to the rest of the empire to pay for the taxes extracted by Rome, does not explain the situation. The eastern trade alone probably gave Egypt an inflow of wealth several times greater than the notional tax-take, of which at most a half actually left Egypt. Far more important than taxation was the more indirect, highly variegated and equally pervasive institutional, commercial, and behavioral influence of Rome. It changed consumption patterns in Egypt: by the mid-first century even richer villagers were adopting Roman tastes in dress, food, funerary commemoration, and so on. Over the second to third centuries most metropoleis turned themselves into Romano-Greek civic centers with the appropriate monuments, entertainments, and wheat distributions on the Roman model. The sizeable Roman army presence in Egypt introduced to its many local recruits, and the surrounding population, Roman management practices, including paper salary accounts, Roman tools, equipment and techniques, especially in construction and mining, and the Roman economic mentality. It is noticeable that veterans and their families typically had complex financial affairs, making and taking loans of all types, assuming tax-collecting and public supply con-tracts.69 Roman taxation expected high liquidity from top to bottom of society, and the Roman administration accepted its duty to provide and maintain an adequate coinage. It also deliberately fostered a general climate of probity and equity in public and private business. The Roman government imported its age-old preference for private ownership over public control of land, which enabled, while their administrative reforms encouraged, the formation of large private estates. Roman influences picked up by these estates included army-style salary accounts, barrack accommodation for workers (kella, from the Latin cella), the use of four-wheeled
Carts {karron/karnon, from currum), and wine-pressing installations made from fired brick and hydraulic cement. Brick and cement construction techniques influenced public building too, and if we knew more about the urban economy we would doubtless find more influences like the mass production techniques used for clay figurines. However, the main stimulus to economic development in Roman Egypt came from the Roman creation of a peaceful and open Mediterranean market, and the boom in demand caused by empire-wide urbanization. The market for Egyptian wheat was, in effect, limitless (in that the normal price in Italy was at least twice that in Egypt), and it was mass consumption patterns, not just elite profligacy, which drove the commerce with the east, and all the crafts and trades which rode on it, from which Egypt, especially through Alexandria, profited so greatly.
World History
