This book explores central areas of the Roman economy, and ways in which they connect and interact.

In a vast and unwieldy domain like the Roman empire, the speed of communication by sea and the number of shipping movements were obviously important for the processes of government as well as for the economy. What we know of message-speeds is usually disjointed. But more systematic results can be gained from Egyptian documents, which provide thousands of precisely dated co-ordinates identifying the emperor in power. When the emperor changes, the co-ordinates can show how soon this essential fact became known in one of Rome's eastern provinces, and how long the news took to spread inside the province. The results in the period when the evidence is fullest mainly suggest dependence on commercial shipping, with news getting through faster the more closely its date happened to coincide with two main shipping movements in the year. Seasonal differences are very striking in the pattern from the Flavians onwards, and the arrival of news apparently depended on a limited number of shipping-links. The transit-times of government decrees sent to Africa under the later Empire again suggest two main shipping movements during the year (chapter I).

This has some relevance to inter-regional trade. Can substantial trade flows be inferred by arguing that government taxation drew money out of provinces with large tax bills to an extent which only increase in trade could have corrected?

Change of emperor: a worm's eye view

‘The good genius of the world and source of all good things, Nero, has been declared Caesar. Therefore ought we all wearing garlands and with sacrifices of oxen to give thanks to all the gods’ (POxy 1021, 17 November AD 54).

Egypt has left far more documentary evidence than any other Roman province. One way in which its extraordinary wealth of data can throw light on communication-speed is by showing how long it took Egyptians to learn that one emperor had been replaced by another. The results can be compared with later Roman evidence for the travelling-time of Edicts (section 2), and with evidence for voyage-times from the Renaissance period (section 3).

Introduction

The emperor typically lived and died in central Italy. This meant that Egypt was not well placed for hearing of any change at an early stage. Nevertheless journeys to Alexandria of 6 and 7 days from Sicily and 9 days from Puteoli were recorded, though only as exceptional events. Typical delays in Egyptian awareness of change of emperor turn out to have been much longer, suggesting that even urgent news in practice travelled slowly to provinces overseas.

The most direct route from Rome to the Egyptian interior involved one journey by sea and another by river. Occasionally the two stages can be separated. Galba's accession was known in Alexandria by 6 July, within 27 days of Nero's death on 9 June AD 68.

Introduction

Despite prevalent assumptions that its character is well known, the surviving details of provincial taxation under the Principate are remarkably incomplete. Under Augustus some systematisation evidently took place, with the introduction of provincial censuses. The epigraphic evidence does not in itself suggest that these were regular (the surviving references per province are very few save those for Gallia Comata), or that they were introduced in all provinces (the evidence is concentrated overwhelmingly in the ‘imperial’ provinces). Censuses involved, or could involve, surveys of land and property as well as of population. Their occurrence implies that tax was being proportioned in some way to assets, instead of being levied as a fixed amount, which might be arbitrary.

But census-taking was made necessary by poll–tax, if nothing else. Though poll–taxes were levied in money, the occurrence of a census does not in itself indicate that land in the province was taxed in money. The evidence about the way in which the main tax of a province was paid reveals diversity, not consistency.

The evidence of Hyginus

The closest that any ancient source comes to a general description of Roman tax-regimes is in a passage from the second–century land–surveyor Hyginus (205L). Soon after distinguishing between tax–free land (ager immunis) and land liable to tax (ager vectigalis), Hyginus gives a brief description of the way in which direct taxes are assessed.

The Beatty papyri originally consisted of two long rolls containing a day-to-day transcript of high-level correspondence sent to Panopolis in AD 298–300 by the authorities governing the Thebaid. The long series of letters, whose frequent refrain is the failure of subordinates to carry out instructions, gives the most graphic impression of Roman provincial government in action; and the dossier is in its way as important as Pliny's correspondence with Trajan two hundred years earlier.

The letters used here describe payments to ten different army units stationed in the Thebaid, which together made up a large part of the garrison of southern Egypt. Because they contain instructions to pay, the letters specify unit, amount, type of payment and period covered. If in a single case either the rate or the number of recipients can be convincingly identified, much wider information about unit–strengths and military spending under Diocletian potentially becomes available.

In an important pioneering discussion, A.H.M. Jones analysed the payments in money, using smallest round-figure denominators and common denominators. His results suggested that unit-strengths were relatively high, but that the basic rates of army pay (stipendium) late in Diocletian's reign were still what they had been early in the third century.

His interpretation, although based on ingenious analysis, and orthodox in some of its results, is not entirely convincing (Appendix 5).

EARLY ROMAN CHRONOLOGY

The so-called Varronian system of chronology used in this volume effectively placed the foundation of Rome in 753 B.C., the first consuls in 509 and the Gallic Sack of Rome in 390, and was largely followed by the Capitoline Fasti (p. 347). It was, however, a creation of the mid-first century B.C. (perhaps of Atticus rather than Varro) and incorporated the dictator–years, which appear to be a late invention (p. 348); it is not, therefore, representative of the chronologies employed by the Roman historians, to whom the dictator-years were foreign. Those chronologies, however, are imperfectly known since few relevant data are preserved from the lost early chroniclers of Rome and not all the surviving authorities are systematic, accurate or even internally consistent in their chronologies. Thus Livy (probably following the pattern of his Latin predecessors) is generally content to chart the passage of the years merely by recording the successive consular colleges and only occasionally employs dates ‘from the foundation of the city’ (ab urbe condita). As a result, it is very doubtful whether he worked with a clearly defined overall chronological system, particularly since his own narrative omits certain consular years (490–489 and 376 on the Varronian scheme) which his dates ‘from the foundation of the city’ seem to include and some of these latter dates themselves appear to be mutually incompatible. The most satisfactory explanation of these last inconsistencies is that Livy's dates ‘from the foundation of the city’ derive from two different schemes (presumably to be found ultimately in different sources), which placed the foundation of Rome in 751 or 750 respectively, the establishment of the Republic in 507 or 506 and the Gallic Sack in 386; but even this remains hypothetical.