Some aspects of wage payments and coinage in ancient Rome (1st-3rd century AD), in J. Lucassen (ed.), Wages and Currency: Global and Historical Comparisons, Peter Lang, Bern-Berlin, 2007, p. 77-96.

chapter 3 Some Aspects of Wage Payments and Coinage in Ancient Rome, First to Third Centuries ce∗ Johan van Heesch The Roman Empire stretched over more than thirty modern states, from Great Britain in the west to Jordan in the east and from the Netherlands in the north to Egypt and Morocco in the south. Although the empire was ruled by a single emperor and by a fairly centralized government, there were important social and economic differences across such a wide territory. In that pre-industrial society, mass production and very large factories were exceptional, and few really large towns existed, among them Rome, Alexandria, and Antioch, cities that were essentially parasites on the countryside, which was where at least eighty per cent of people lived, working in small villages or on farms. Coinage use and coinage circulation patterns varied from one area to another and it is certain that barter and other forms of payment were used side by side in more isolated areas of the empire. During the first three centuries of the Roman Empire the Roman imperial government minted nine different denominations of coins in three metals (gold, silver, and bronze),2 which in itself is sufficient reason to believe that the Roman monetary economy was fairly well developed and that coins – at least in major towns – were probably used in almost the same way as today.  2 I would like to express my gratitude to Helen Wang for her comments on this chapter and for her language editing. In gold: aureus and quinarius; in silver: denarius and quinarius; in brass: sestertius, dupondius and semis; in copper: as and quadrans.  aureus = 2 gold quinarii = 2 denarii = 0 silver quinarii = 00 sestertii = 200 dupondii = 00 asses = 800 semisses = ,600 quadrantes. In numismatic studies the word bronze is generally used to denote any alloy containing copper. Most bronze coins from the High Empire are actually made of pure copper (as and quadrans) or brass (sestertius, dupondius, and semis). Brass was much more expensive than copper. For a general introduction to Roman coinage see Burnett, Coinage in the Roman World, and Harl, Coinage in the Roman Economy. For further details, see Wolters, Nummi signati. For special purposes (gifts) gold, silver, and bronze multiples were also coined. Gold and silver multiples are extremely rare before ce 20 as most were melted down (some are known for Augustus, Domitian, Trajan, and Hadrian). See Gnecchi, I medaglioni romani.     ± "" "" Rhine " "" " 8 " Au " Pannonia " " "" Au 6 9 4 " 10 Da lm " Moesia at Au ia "" Spain 7 2 1 3 " 5 Ag " " " " " Syria Au Africa " Egypt " 10 Legions Mining areas (gold (Au) or silver (Ag)) Mints johan van heesch Map 3.1. Location of the Augustan legions, mints and main mining area (design Johan van Heesch) Key to Mints: 1 Ephesus; 2 Pergamum; 3 Samos; 4 “Caetra”-mint; 5 Corduba; 6 Caesaraugusta; 7 Emerita; 8 Lugdunum; 9 Roma; 10 Nemausus 3 some aspects of wage payments and coinage in rome   Wage Labour in Ancient Rome Although wage labour in Ancient Rome never reached the level it has in industrial societies, it existed, and can be thought of as omnipresent in both larger and smaller towns throughout Italy and abroad, as well as in villages and the countryside.The most important group of wage earners were people employed in imperial public service. Agrippa said in his speech to Octavian in 29 bce: “In monarchies one set of people usually engages in agriculture, manufacturing, commerce and politics, and these are the classes from which the state’s receipts are chiefly derived, and a different set is under arms and draws pay.”3 In 2 ce some 00,000 men (almost one per cent of the entire population) were under arms and received a stipendium. That meant annual expenditure of some ,88,000,000 sestertii, and it was normal during the High Empire (from 27 bce to 29 ce) for such sums to be expressed in sestertii, the money of account. But the actual sestertius, a brass coin, was only rarely used for making the payments. Men were paid only three times a year, probably in January, May, and September.6 Even if costs for food, clothing, and armour were deducted from their wages, a considerable amount of money still had to be paid in cash. Several military pay records are known, found in Egypt and written on papyrus. They show that soldiers had to deposit a considerable part of their pay into the camps savings “banks”, a fact well known from other sources. Vegetius in his Epitome of Military Science (II.20) clearly states that “It was a divinely inspired institution of the ancients to deposit ‘with the standards’ half of the donative which the soldiers received, and to save it for each soldier, so it could not be spent by the troops on extravagance or the acquisition of vain things.”7 The most interesting document is the text known as “Geneva 3  Dio Cassius, Roman History, 2.6., transl. Earnest Cary [Loeb Classical Library], (Cambridge, MA, 9-927). The population of the Roman Empire was about 60 million in ce 6, but by ce 200 this number would have decreased as the Empire suffered from several plague epidemics between ce 6 and 80. For demographic estimates see Frier, “Demography”, pp. 83-86. For the impact of the plague see Duncan-Jones, “The Impact of the Antonine Plague”, pp. 08-36, esp. p. 6, note 88, where a figure of between twenty-five and thirty-three per cent for mortality in places affected by the plague is considered to be realistic. Duncan-Jones, Money and Government, pp. 3 and . The Empire’s annual budget is estimated at ,63 million sestertii. See also Wierschowski, Heer und Wirtschaft, and more recently Erdkamp, The Roman Army and the Economy. Watson, The Roman Soldier, pp. 9 and 06, and Fink, Roman Military Records, pp. 266267, document 7. Vegetius, Epitome of Military Science, transl. N.P. Milner (Liverpool, 996), p. .  6 7     johan van heesch papyrus ” (PGen Lat. recto) and dated 8 or 83 ce. It mentions two legionary soldiers and the pay they received, as well as the deductions for shoes, hay, food, and clothing. As each of them received 27½ drachmae three times a year and since the Alexandrian tetradrachm (a coin worth four drachmae) was the equivalent of one Roman denarius, the 72½ drachmae (27. x 3) represented 8 and /8 denarii, i.e. 82.% of the normal rate of pay at that time (22 denarii).8 As no withdrawals for pocket money are mentioned in the accounts, it is supposed that only part of the pay was deposited and that the rest of their money was paid in cash. The soldiers mentioned in this document then received approximately 39 denarii in cash, which was 7.% of their pay. This amount was probably used for their private expenses and to support their families too, who lived in or near military camps, depending on the rank of the soldier.9 If that was general practice it means that the purchasing power of the ordinary soldier and his direct impact on local markets was quite limited, and that a substantial part of the supplies for military regions (e.g. food) was bought in bulk by the authorities. These transactions were very probably paid for in high-value coins struck from gold and silver (aurei and denarii).The presence of large gold hoards in the fertile rural areas of northwestern Gaul could be explained by such purchases.0 Salaries of ordinary legionary soldiers are fairly well known and the figures, expressed in sestertii, the unit of account and abbreviated HS (the Roman symbol for sestertius, originally IIS or two and a half asses), can be seen in Table 3.. A significant number of civil servants received regular cash payments, and their numbers increased considerably from the Flavians onward (69-96 ce), but exact figures are unknown. Duncan-Jones estimates the budget for civilian employees at 7 million sestertii in the early part of the third century; a figure only one-sixteenth of the military budget.2 Beside the imperial government service, the municipalities too employed salaried men. 8 9 Watson, The Roman Soldier, pp. 06-07, and Alston, “Roman Military Pay”, pp. 67. Before the reign of Septimius Severus soldiers were not entitled to legally marry when in service, and they often lived with their families. Higher officers were allowed to live with their wives, children, and slaves in camp. See Watson, The Roman Soldier, pp. 33-37, and Bowman, Life and Letters on the Roman Frontier, pp. 7-62. Van Heesch, De muntcirculatie, pp. 87-9. For details on the Roman budget see Duncan-Jones, Money and Government, p. . For civil servants see Eck, “The Growth of Administrative Posts”, pp. 238-26. For an excellent survey on wage payments see Corbier, “Salaires et salariat”, pp. 6-0. Also very useful are Garnsey, Non-slave Labour in the Greco-Roman World, and Mrozek, Lohnarbeit im Klassischen Altertum. Duncan-Jones, Money and Government, p. . 0  2 3 some aspects of wage payments and coinage in rome   For instance, a secretary of the colony of Urso in Spain received an annual pay of ,200 sestertii,3 and it is likely that some public works in Rome were undertaken by free labourers who received a salary.That work had an important social function, as Suetonius (Life of Vespasian, 8) states when talking about the emperor Vespasian: “To a mechanical engineer, who promised to transport some heavy columns to the Capitol at small expense, he gave no mean reward for his invention, but refused to make use of it, saying: you must let me feed my poor commons.” Table 3.1. Annual legionary pay in the Roman Empire (in Sestertii) Period (ce) Before 89 89-202 202-212 212Legionary pay 900 HS (or 225 denarii) 1,200 HS (or 300 denarii) 1,600 HS (or 400 denarii) 2,400 HS (or 600 denarii) To most historians that last sentence has been understood as “I must always ensure that the working classes earn enough money to buy themselves food”. However, we do not know if these labourers were actually employed by the Roman government, for it is possible that contractors responsible for public works themselves hired the labour force. Of course slavery, as well as 3   Crawford, Roman Statutes, pp. 00 and 22, document no. 2 (Lex coloniae Genetivae, LXII.32). The document is of the Flavian period (last quarter of the first century ce). Duncan-Jones, Money and Government, p. 3. For the date of the pay rise under Domitian see Campbell, The Emperor and the Roman Army, p. 8. The “famous” coin of Domitian with, on its reverse, the legend stip(endium) imp(eratoris) avg(usti) Domitian(i), dated ce 8 by Kraay (“Two New Sestertii of Domitian”) and also mentioned in Griffin, “Nero to Hadrian”, p. 6, appears to have been wrongly read. Giard proposes Titus (imp) avg Domitian, Giard, Monnaies de l’empire romain, III, no. 29. This coin then cannot be used as evidence for the pay rise. See also note 30 of this chapter. Having examined the coin in Paris, I can confirm that its reading is problematical, though Kraay’s interpretation cannot entirely be ruled out. The letters are very vague and smaller than on most of the coins of this period. It could have been tooled in recent times to match the literary tradition. For army pay see also Zehnacker, “La solde de l’armée romaine”, pp. 9-2. Brunt, “Free Labour and Public Works”, pp. 8-00, and Mrozek, Lohnarbeit, p. 9 (this interpretation is not universally accepted). The first Suetonius translation is by J.C. Rolfe [Loeb Classical Library] (Cambridge, MA, 9); for the second translation see Suetonius, The Twelve Caesars, transl. R. Graves and M. Grant (Harmondsworth, 989). See also Duncan-Jones, Structure and Scale, pp. 7-8.     johan van heesch corvée labour, existed, but there is no doubt that free labour played an important role in the Roman economy.6 It was very common to find wage labourers in most major Roman towns, where shopkeepers, traders, and craftsmen employed freeborn people, freedmen, and slaves.7 Itinerant workers, free wage labourers, and day-wage labourers are also attested in rural areas and mining districts in every part of the empire.8 We could quote an example from the gospel of Matthew, which says that a day labourer earned a daily salary of one denarius. Some people in Pompeii were paid one denarius a day, but others there received five asses (6 asses =  denarius).9 It is very difficult to carry out detailed research on wage levels and their importance under the Roman Empire. Most sources are anecdotal, and there are not enough data on the extent of non-slave labour. Egypt is the only region for which a substantial amount of material is available, for thousands of papyri are known from the Roman province, but the interpretation of even these data remains difficult as they do not always come from the same area or even period.20 Just as in modern society, wages varied according to region and occupation and to a great extent depended on demand. In his book on prices and wages in Egypt, Drexhage lists some 27 daily wages covering the first three centuries of the Roman Empire.2 Day labourers working on farms earned less than one drachma a day in the first century ce. On average they received three and a half obols a day (6 obols =  drachma;  drachmae =  silver denarius). Wages were higher in the second century ce, rising gradually from about six to nine obols, and even as high as fourteen towards end of the century. After 20 ce it is not rare to find day wages of four drachmae (twenty-four obols) or more.We do not always know if those wages were actually paid daily or only after some longer period of work. It is perhaps worth remarking that during the first two centuries ce the most common coin denominations were the billon tetradrachm (=  drachmae =  denarius) and the bronze drachma. From the time of Commodus (80-92 ce) onward drachmae became very rare and the tetradrachms dominated coin circulation.22 6 7 8 9 20 2 22 See the recent bibliography on slave labour in H.W. Pleket, “Wirtschaft”, in Vittinghoff, Europäische Wirtschafts- und Sozialgeschichte in der römischen Kaiserzeit, pp. 99-202. Corbier, “Salaires et salariat”, and Treggiari, Urban Labour in Rome, pp. 8-6. Garnsey, “Non-slave Labour”, pp. 3-7, and Mrozek, Lohnarbeit. Mrozek, Lohnarbeit, pp. 0-. For the Egyptian data see Drexhage, Preise, Mieten/Pachten. Ibid., pp. 2-2. See also Johnson, An Economic Survey of Ancient Rome, pp. 30-32. Milne, Catalogue of Alexandrian Coins. 3 some aspects of wage payments and coinage in rome   There are two key questions. Did the denominational structure of coinage reflect changes in wage levels? And can the pattern of coin finds throw any light on the presence of wage earners and the coins with which they were paid? Before addressing these questions it is important to describe briefly how Roman coinage was organized. General Characteristics of Roman Coinage The organization of Roman coinage during the High Empire differed in many ways from that of medieval times. Coins were minted not only by the Roman government but also by a number of local authorities and by a few hundred mostly eastern cities. The Roman Empire’s mint masters minted in gold, silver, and bronze, whereas most of the provincial mints issued only bronze coins. From the reign of Tiberius (-37 ce), the coinage for the whole empire from Gaul to Syria was minted first in Lyons and Rome, but from 6 ce onward only in Rome. But these western coinages rarely reached eastern parts of the empire. In the east, local bronze coinages were used and some important provincial mints like Caesarea in Cappadocia and Antioch in Syria minted silver coins to cover imperial expenditure in the region. In the third century, especially after about 20 ce, the number of Roman mints increased again and local coinages everywhere were gradually replaced by Roman denominations.23 Egypt, one of the richest provinces of the empire, had its own monetary system until the reign of Diocletian (28-30 ce). An imperial mint in Alexandria struck mainly tetradrachms and drachmae for circulation in Egypt alone. Although gold Roman coins are occasionally found in Egypt, it seems that until the reign of Diocletian most “foreign” coins were not allowed to circulate there. Old coins were obviously re-used, but if payments in new coins were made away from minting centres, then it is clear that an organized transport system must have existed, however difficult or costly it might have been. Although gold, silver, brass, and copper coins were issued, not all those metals were always minted continuously, especially in the first century ce. The mint at Lyons, for example, did not coin any gold or silver between 7 and 2 bce but worked in bronze. On the other hand, the Roman government minted no bronze coins between 2 and 6 ce. We remain uncertain why minting was sometimes so irregular, but plausible explanations include lack of precious metals or the limited use of small denominations in certain 23 Van Heesch, “Mints and the Roman Army”, pp. 3-2.     johan van heesch types of payment. The absence of an official bronze coinage from 2-6 ce suggests that for twenty-two years it was mainly gold aurei and silver denarii that were used to cover state expenditure and salary payments. The monetary system of the Roman emperors was reasonably stable. Most of the denominations struck by Augustus at the end of the first century bce still existed in the middle of the third century ce. The relationships among those coins remained unchanged for at least 200 years and coins were rarely demonetized. The only apparent changes were modifications to precious metal content or weight.2 As the metal content of gold and silver coins of the first two centuries was quite stable, and as those coins were probably slightly overvalued, they could justifiably circulate for a very long time. For example, denarii from the second century bce have been found in hoards from the Flavian period (69-96 ce), so it should not surprise us if old coins were used alongside new ones for both private and public expenditure. In short, the only official coinage in circulation in the western part of the empire and from the time of Claudius onward was that struck in Lyons and Rome for the Roman emperors. Coinage was a matter for the government and private minting was highly exceptional.2 Since the Roman government had sole charge of minting coins in the west, it is very important to determine how they were placed in circulation. Issue of coins was effected through government payments and, in exceptional circumstances, through exchanges when the old coinage was demonetized.26 The following expenses weighed heavily on the Roman budget:27 - army pay (stipendia) at regular intervals (three times a year); wages of civilian employees; handouts to civilians (congiaria) and soldiers (donativa); discharge costs (praemia) after twenty-five years’ army service; expenditure on public buildings. As coins were minted in so few places it is highly improbable that private individuals had the opportunity to offer gold and silver objects or bullion to 2 2 26 27 Walker, Metrology. An exception should be made for coin forgeries and emergency issues of small change especially abundant between ce  and 6 and between c. AD 270 and 320. See Boon, “Counterfeit Coins in Roman Britain”, pp. 02-88. For example, under Trajan in ce 07 and possibly in the second half of the third century ce. Howgego, “Why did Ancient States Strike Coins”, pp. -2; Duncan-Jones, Structure and Scale, pp. 33-6, and von Kaenel, “Zur ‘Prägepolitik’ des Kaisers Claudius”, pp. -68 (on the link between coinage and building activities). 3 some aspects of wage payments and coinage in rome   the mint in exchange for freshly minted coins. Some historians have supposed that money changers or bankers (nummularii) organized transport of coins and exchanged gold and silver coins for small bronze coins. But the role of bankers in these processes is not at all clear.28 It is difficult to see how a profit could be made from organizing the costly and dangerous transport of coins from the mint at Lyons or Rome to the most remote corners of the Empire. If merchants or tradesmen needed coins for wage payments, these coins must have come from those already in circulation or from transactions with the government. If there was a preference for a particular denomination, traders could go to bankers, but it is important to bear in mind that there was only one official mint during the first and second centuries. It is unlikely that the Roman mint masters would have been able to satisfy short-term fluctuations in the demands of the “market”. Monetary Policy and Wage Payments By studying the pay of soldiers it is possible to trace a link between monetary changes and wage payments. Between the reign of Augustus (27 bce – ce) and 20 ce the Roman monetary system did not undergo regular or spectacular modifications. However, there were two major types of change: there were several reductions in the silver content of the denarius, the basic silver coin (this was sometimes accompanied by a weight reduction of the gold coin as well), and there was the introduction in 2 ce of the antoninianus (a double denarius), which gradually became the new standard silver coin. Reductions in the weight and fineness of the coinage can be explained in two ways. In a bimetallic currency system with a fixed relationship between silver and gold coins, the government could modify the content of the coins when there was an imbalance between the market value of the two met28 Andreau, La vie financière dans le monde romain, and also Andreau, Banking and Business in the Roman World. This concerns the western part of the empire only. In the east hundreds of towns struck bronze coins until the middle of the third century. The function of these coinages is not completely clear, but I suppose they were also put into circulation in the same way the imperial coinage was. In the eastern cities, the role of bankers might have been more important; they were probably involved in the exchange of the different city coinages and the exchange of Roman gold and silver against local bronze coins. Some private minting of “emergency” issues of small change might have existed in the west during the first half of the first century ce and at the end of the third century. Individuals seem to be responsible for the large quantities of imitative bronze coins that circulated freely. It is possible that these coins entered circulation through local money changers or by wage payments.     3 2,5 2 1,5 1 0,5 0 89 81 91 83 85 87 93 johan van heesch Figure 3.1. Output of denarii (in millions) under Domitian (81-96 ce) Source: Carradice, Coinage and Finances, p. 8. als.29 In Roman times, a shortage of precious metals was a frequent reason for debasing the coinage. A shortage might be caused by the exhaustion of mines or by diminished income from taxes; by a too zealous hoarding of coins or a rise in government expenditure. In certain cases a reduction in the fineness of the coinage was related to an increase in army pay. Between the reigns of Augustus (27 bce –  ce) and Caracalla (2 – 27 ce), army salaries changed at least four times (see Table 3.). On historical grounds it is assumed that the pay rise during Domitian’s time took place during the revolt of Saturninus or just after (88-89 ce).30 If we look at the hypothetical output figures of the Roman mint, based on coin hoards and die counts, we see a spectacular rise in the output of denarii from 88-89 ce onward (Figure 3.).3 This increase in the production of silver coin could be 29 30 3 Lo Cascio, “State and Coinage”, pp. 76-86. Hasler, Studien zu Wesen und Wert des Geldes, and Pankiewicz, Fluctuations de valeur des métaux monétaires. Campbell, The Emperor and the Roman Army, p. 8, and note  of this chapter. Dio 67.3. (Zonaras) and Suetonius, Domitianus, 7.3. Dio (Zonaras) thinks that the pay rise was related to the German victory of ce 82, though he is not certain. For the date of the revolt of Saturninus see Halfmann, Itinera principum, p. 82. Carradice, Coinage and Finances, pp. 9-92. Domitian raised the fineness of the silver coins between ce 82 and 8; in ce 8 he returned to the pre-82 standards. See DuncanJones, Money and Government, p. 227 (table based on the work of Walker, Metrology). 95 3 some aspects of wage payments and coinage in rome   linked to the pay rise.32 However, we should need to explain why the pay increase of 33.3% corresponds to a rise in coin output of 00%, and even 0%. That suggests the pay rise was probably not the only reason for the observed increase and that other state costs (including bonuses, praemia, payment of civilian employees) rose in the same period. Perhaps we should consider the possibility that the government might have switched from payments in gold or bronze to payments in silver, something that is in any case difficult to verify since production estimates for those metals are really only approximations.33 Moreover, figures for 82-8 ce could well be too low as they were the years when heavy gold coins and better silver denarii were minted.These gold and silver coins of good standard disappeared from circulation pretty swiftly and that might have caused a distortion in our data. If the first of our four pay rises took place in 89 ce and not in 88 ce, then the extra 33.3% the soldiers received is remarkably close to the 36% increase in coin output from 88- 89 ce, as calculated by Carradice (Figure 3.).3 The next pay rises took place under the emperor Septimius Severus. They can be dated to 97 and 202 ce.3 Severus’s great debasement (silver from 8.8% to 6%) dates to 9/9 ce and so preceded by some years the first of these pay rises.36 We may assume that the debasement led to a rise in prices and eventually to the pay rise. A second debasement, dated 97 or 98 ce, can be linked more easily to the first pay increase, although we should not forget that large bonuses (donativa) were distributed to the army during those years.37 No change in the coinage can be detected in 202 ce and even the output of silver (and gold) does not seem to have been influenced by the pay rise of that year.38 32 33 3 3 The ludi saeculares of ce 88 also resulted in significant expenditure. Wars were fought by Domitian before and after ce 88, so this cannot be the only reason for the observed increase in output. Carradice, Coinage and Finances, pp. 08 and 38. Ibid., p. 88. In ce 88, a total of 2,780,000 denarii were minted, and in ce 89 the total was 9,890,000, an increase of 7,0,000 denarii. All these figures are, of course, very rough estimates. ce 97: Herodian 3.8. and Historia Augusta, Severus, 2.2; ce 202: Historia Augusta, Severus, 6.9. See Campbell, The Emperor and the Roman Army, pp. 8-86, who dates the pay rise to ce 97 (after the fall of Clodius Albinus), and Callu, Politique monétaire, p. 30. Gitler and Ponting, The Silver Coinage of Septimius Severus; Walker, Metrology, III, p. 38. For reductions in the silver content in ce 97 or 98 see Duncan-Jones, Money and Government, p. 227 (table based on Walker’s data), and Callu, Politique monétaire, pp. 2-2 and 76. Duncan-Jones, Money and Government, p. 3 (table). 36 37 38     johan van heesch The fourth pay rise, that of Caracalla, can probably be dated to 22 ce, shortly after the murder of his brother Geta.39 The rise was important: the ordinary legionary soldier now received 600 denarii per year instead of the 00 he had received before 22 (and 300 before 202 ce). The effects on the coinage of that rise are much clearer to see than in the previous cases. Caracalla debased the coinage by 8% in the same year and an output peak can be seen in 23 ce (Table 3.2).0 Undoubtedly, Caracalla’s military extravagance and the pay rise of at least 33% were the main causes of the introduction, in 2 ce, of a new and larger coin denomination, conventionally called the antoninianus. Table 3.2. Silver coins in the Reka Devnia hoard struck under Caracalla Period (ce) 211-212 213 214 215 216 217 (Jan.-Apr.) Denarii 251 928 276 369 171 59 Antoniniani 102 39 25 Source: Duncan-Jones, Money and Government, p. 139. The introduction of the antoninianus meant a real gain for the government, which had tried to cover its growing expenditure by cheating both soldiers and civilians. Although the antoninianus had a value of two denarii, it contained only one and a half times the amount of silver (Table 3.3). The profit for the emperor is clear: the pay rise of about 33% was actually covered by the debasement of 8% and the introduction of the antoninianus (Table 3.). 39 0 The pay rise for the praetorians is dated ce 22. See Herodian ..7. See also Dio 78 [79], 36.3. From .% to 0.% silver. See Walker, Metrology, III, pp. 30-3, and for output figures see Duncan-Jones, Money and Government, pp. 39 and 2. These figures on the precious metal content of the coins are certainly too high and thus wrong. Though Walker’s metal analysis provides us with percentages that are systematically too high, I believe that his general conclusion regarding the evolution of the metal content is valid and should still be accepted. On the work of Walker and the margin of error see, for example, Gitler and Ponting, The Silver Coinage. Nineteenth-century analysis of the metal content of antoniniani issued by Caracalla (wet method) gave the following figures: 62%, %, and 2%. See Hammer, “Feingehalt der griechischen und römischen Münzen”, pp. 02-03. 3 some aspects of wage payments and coinage in rome  Table 3.3.Weight of Caracalla’s denarii and antoniniani in 215 ce Denomination Denarii Antoniniani Weight 3.17 g 5.13 g % silver 50.5 50  Weight of silver 1.60 g 2.55 g Source:Walker, Metrology, III, pp. 19-20. However, the introduction of this larger denomination was not a success and was soon ended (under Elagabalus, 28-222 ce). It was only from 238 ce onward (during the reigns of Pupienus, Balbinus, and Gordian III) that the antoninianus effectively replaced the denarius in the soldier’s pay. Table 3.4. Caracalla’s pay rise and the effects of the monetary reforms - Pay rise (212 ce) - Monetary reforms by way of reductions in costs of production: c. 33% Profit of 8% Profit of 25% 33% Debasement (212 ce) Introduction of antoninianus (215 ce) Total Source: see text. Coin Circulation and Wage Payments Coins were put into circulation through government spending and remained in use for a very long time. They travelled, became intermingled, and were used in all sorts of subsequent commercial transactions. It is not easy to link the different coin denominations in circulation to a specific type of government spending or to wage payments. However, in some instances we can demonstrate that the circulation pattern in one area was determined by the presence of wage earners. In such cases, coin finds throw light on the coins actually used to pay salaries. Our first example comes from northwestern Gaul and concerns the small non-Roman bronze coins, traditionally called “Celtic bronzes”. Several series of small bronze coins (the so-called “rameau” bronzes) were issued in the territory of the Nervians in northern Gaul, probably during the reign of Augustus (27 bce – ce). It was the first time that small change became available in     johan van heesch Figure 3.2 Sestertius (brass) of the emperor Nero (54-68 ce). On the reverse the representation of Nero’s second congiarium or coin distribution. On the ground an attendant holding a rectangular counting device and a citizen of Rome holding out fold of toga to receive the coins (Brussels, Coin Cabinet, du Chastel collection no. 411) that region. These coins were found mainly in Roman village settlements (vici). As their types and occasional inscriptions prove, they were local Gallic productions, intended for use in the limited area of the civitas Nerviorum. As the typical Roman military coinages of the same period (discussed below) were almost entirely absent from the region, it is clear that Roman military payments did not influence the circulation pattern of this region. I believe that these series of coins can be linked to local expenditure, in particular to wage payments made by the leading classes of the civitas. Leaving aside the possibility of having to pay a local militia, it is still easy to suppose that a large number of workmen were required to construct the road from Bavay to Cologne, to build the civitas capital in Bavay (note the visit of Tiberius in  ce, mentioned in an inscription found in Bavay), and to build the important religious centre at Blicquy.2 The second example concerns the Rhine limes, or frontier. Although Julius Caesar conquered Gaul between 8 and  bce, permanent army camps along the Rhine and in Germania were built during only the last few decades bce of the reign of Augustus. In these military centres lived thousands of soldiers all needing to be paid – assuming all went well – three times a year. Undoubtedly, in the camps and their immediate surroundings coins were used almost exactly as they were in the Mediterranean urban economy. Although the Gauls of the north had a long monetary tradition, with gold and even potin coins minted in the second century bce (gold even earlier), the “monetariza 2 Van Heesch, De muntcirculatie, pp. 0-3 and 70-72. See Thollard et al., Bavay antique, Demarez et al., Le sanctuaire gallo-romain de Blicquy, and Tacitus, Agricola, 2., for the importance of building activities and early Romanization. 3 some aspects of wage payments and coinage in rome   tion” at the time of Caesar’s intervention was rather limited.3 Local Celtic coinages circulated in some religious and political centres (the Titelberg in modern Luxembourg for instance), but only from the reign of Augustus (27 bce – ce) did the use of coins increase to unprecedented levels. The finds from these Roman camps are highly instructive and it is reasonable to suppose that the coins found there originally entered circulation through military pay in the form of salaries and bonuses (donativa). Two of the many German sites are of particular interest because they reflect coin loss over a very short period of time: the camp at Oberaden (occupied between  and 8/7 bce) and the famous Varus battlefield at Kalkriese near Osnabrück, dated 9 ce, where three Roman legions were lost (Table 3.). Table 3.5.The coins from Oberaden and Kalkriese45 Denomination Aureus Aureus quinarius Denarius Denarius quinarius Sestertius Dupondius As Semis Quadrans Celtic silver Celtic bronze Oberaden 31 3 2 317 2 9 Kalkriese 19 384 8 1 2 484 1 2 Sources: Ilisch, “Die Münzen aus den Ausgrabungen im Römerlager Oberaden”, and Berger, Kalkriese . 3   Heinrichs, “Überlegungen zur Versorgung”; N. Roymans, “The Lower Rhine Triquetrum Coinages”; Wigg, “Die Rolle des Militärs bei der Münzversorgung”. Ilisch, “Die Münzen aus den Ausgrabungen im Römerlager Oberaden”, and Berger, Kalkriese 1. The denomination of the bronze coins struck at Nîmes (found in large numbers at Oberaden) is disputed. Some numismatists describe these as dupondii, but given the very limited role of dupondii in this early period we defend the traditional view that these coins are asses. Halved coins were included in the numbers. Hoards and stray finds for Kalkriese are in the same column (these finds can be considered as one large hoard).     johan van heesch As we see in Table 3., only three denominations (out of nine) occur regularly in finds: gold aurei, silver denarii, and copper asses. Calculated in sestertii, the Kalkriese finds are easier to compare: 19 aurei 384 denarii 484 asses 1,900 sestertii (19 × 100; 1 aureus = 100 sestertii) 1,536 sestertii (384 × 4; 1 denarius = 4 sestertii) 121 sestertii (484 ÷ 4; 1 sestertius = 4 asses) Apparently, during the reign of Augustus, these three denominations were the main coins used in pay for troops. The coins from Oberaden are stray finds, so it is quite normal that the high-value denominations, such as aurei and denarii, are absent or rare as in normal circumstances one could expect them to be lost only rarely. By contrast, in Kalkriese, where coins were lost during the battle, the finds reflect what was kept in the soldiers’ purses. We do not know whether, on pay day, the soldiers received coins of one metal only or a mixture of coins of different metals. As pay days came round only three times a year and as their wages were kept in the camp treasuries, both circumstances are possible, but as the Roman mint did not continuously issue all metals (see above) it is likely that the soldiers received their pay occasionally in silver or bronze only. The Roman historian Tacitus tells us an interesting story about a mutiny that took place after the death of Augustus ( ce).The soldiers in Germany and Pannonia were complaining about their pay. They received ten asses per day but claimed one denarius ( denarius = 6 asses). The switch made by Tacitus from the word “asses” to the word “denarius” might imply that they were asking not only for a pay rise, but also for more silver rather than bronze coins. In fact, gold and silver coins were seldom minted towards the end of the reign of Augustus, but the output of bronze coins was considerable.6 Our next example of a monetary pattern directly influenced by the presence of a large body of wage earners comes from modern Switzerland. Marcus Peter showed that the legionaries stationed at the camp of Vindonissa during Trajan’s reign had access to freshly minted bronze coins.7 Vindonissa 6 7 Tacitus, Annals, 7.6-8 and 3.. For this interpretation see Zehnacker, “La solde de l’armée romaine”, p. 9, and van Heesch, “Some Considerations on the Circulation of Augustan and Tiberian Bronze Coins”, p. 6. Peter, “Bemerkungen zur Kleingeldversorgung”, and Peter, Untersuchungen zu den Fundmünzen aus Augst, pp. 9-92. On coin transports see Millar, The Emperor in the Roman World, pp. 0-2, and Harl, Coinage in the Roman Economy, p. 226. 3 some aspects of wage payments and coinage in rome   Figure 3.3. Denarius (silver) of the emperor Hadrian (117-138 ce). Hadrian and Gallia on the reverse (Brussels, Coin Cabinet, du Chastel collection no. 563 (enlarged)) Source:Walker, Metrology, III, pp. 19-20. was abandoned in 0 ce, which explains why 3 Trajanic coins for the years 98-00 ce were found, compared with only one from the years 0-03 ce. Normally, the ratio of coins from 98-00 ce to coins from 0-03 ce is 60:0, probably reflecting the original output of the mint in Rome. But sites near Vindonissa have a much higher percentage of coins from 98-00 ce than those in other areas (Figure 3.), which demonstrates that the presence of a large concentration of soldiers influenced the circulation pattern of bronze coins in a wide area and that the departure of the soldiers from Vindonissa deprived the entire area of an important source of freshly minted coins. Table 3.6. Silver and bronze coins from Trajan (98-117 ce) to Gordian III (238-244 ce) found at Augst (Switzerland) and Ordona (South Italy) Emperors Silver Augst Bronze 277 317 268 200 105 22 10 2 13 1 7 23 24 22 18 11 62 9 Ordona Silver 1 2 4 2 2 2 1 1 Bronze 19 17 39 12 8 2 4 2 9 2 11 Trajan (98-117) Hadrian (117-138) Antoninus Pius (138-161) Marcus Aurelius (161-180) Commodus (180-192) Septimius Severus (193-211) Caracalla-Geta (211-217) Elagabalus (218-222) Severus Alexander (222-235) Maximinus Thrax (235-238) Gordian III (238-244) 27 53 6 20 Sources: Peter, Untersuchungen zu den Fundmünzen aus Augst, pp. 28-29, and Scheers, “La circulation monétaire à Ordona”, p. 302.     johan van heesch Figure 3.4. Sestertius (brass) of Commodus Caesar (166-177 ce). On the reverse a coin distribution scene with a Roman citizen mounting the steps to receive the coins falling out of the rectangular counting device (abacus) in the hands of Liberalitas.. 100 90 80 70 60 50 40 30 20 10 0 Sl Ca Al Vi Au Wü Da Fr Figure 3.5. Trajanic bronze coins from 98-100 ce and from 101-103 ce. Vindonissa (Vi) (Switzerland) compared with other regions: Slovenia, Carnuntum, Swiss Alps, Augst, South Würtemberg, Darmstadt, and Frankfurt am Main Source: Peter, Untersuchungen zu den Fundmünzen aus Augst. 101-103 98-100 Our last example concerns the first half of the third century. From Septimius Severus onward (98-2 ce), bronze coins tended to become extremely rare finds on the Rhine and Danube frontiers, as well as in Britain and Gaul. They were clearly outnumbered by the now debased denarii. However, in Italy, Spain, and North Africa, third-century bronze coins were much more 3 some aspects of wage payments and coinage in rome   common.8 That is illustrated in Table 3.6, which shows the coin finds from two civilian sites, Augst in Switzerland and Ordona in Italy.9 That bronze was still needed in the northwest in the third century is clearly demonstrated by the presence of cast forgeries of copper coins.0 However, it seems that official bronze was no longer reaching this area. This can be explained by accepting that Roman government payments (civilian and army pay) in those regions were made almost exclusively with silver coins. Why bronze continued to be struck for the less militarized zones is more difficult to explain. Conclusions We know little about how coins were actually paid out. Some bas-reliefs found in the land of the Treveri have been explained as scenes of payment to labourers in the wool industry, but that interpretation is not accepted by everyone. Depictions on bas-reliefs and especially on coins more often attempt to indicate the generosity of the various Roman emperors.2 Though not actual wage payments, they do represent the distribution of coins to Roman citizens, in which an attendant close by the emperor distributes coins using an account board, probably made of wood and containing a certain number of holes, each one filled by a coin. It was a simple device for very quickly counting a fixed number of coins (Figures 3.2, 3. and 3.6). In this chapter, I have focused on certain aspects of wage payments and coinage and discussed some links between monetary changes and pay rises. For clarity I have presented arguments very simply; the Roman economy was very much more complex. Changes of coinage and alterations to the monetary system were not, of course, linked exclusively to payments to government servants, whether military or civilian. Large cash holdings were needed to cover many other expenses, such as those related to construction work, warfare, payments to foreign kings, purchases of luxury products, and offi8 9 0  2 Bost, L’épave Cabrera III, pp. 06-6, and Duncan-Jones, Money and Government, p. 08. Peter, Untersuchungen zu den Fundmünzen aus Augst, pp. 28-29, and Scheers, “La circulation monétaire à Ordona”, pp. 293-37, 302. Lallemand, “Les moules monétaires de Saint-Mard”, and Boon, “Counterfeit Coins in Roman Britain”. Drinkwater, “Die Sekundinier von Igel”. See Campbell, The Emperor and the Roman Army, pp. 83-8, for a payment scene to soldiers on Trajan’s column (doubtful), and, for example, Mattingly, Coins of the Roman Empire, plate 7 no. 6 and plate 03 no .     johan van heesch Figure 3.6.Bas-relief on the triumphal arch of Constantine the Great in Rome with the representation of a coin distribution scene or congiarium.The coins falling out of the wooden (?) counting device can be seen in the fold of the toga of the recipient (extreme left). In the middle a box containing the money and on the right an administrator checking the name of the citizen. (Photo Collection Brussels, Coin Cabinet) cial largesse.3 With the exception of the doles to the citizens of Rome, most of these aspects are not well known because contemporary sources are very rare. Since the value of coins depended to a great extent on their metal content, the government could – if enough bullion were available – strike all the money it required. Yet scarcity of precious metals, exhaustion of mines, and sometimes their loss to invaders are just some of the reasons why the Roman government could not always mint as much gold and silver coins as it wanted. No relation between private wage payments and the minting policy of the Roman emperors during the High Empire has been detected. That should not come as a surprise, for currency was the concern of the emperor and his government. The position and power of the autocratic ruler depended heavily upon his financial independence and capacity. Money was crucial to guarantee the loyalty of his army and to demonstrate his generosity towards the people. 3 For a detailed description of a pay day in the Roman army see Fl. Josephus, Iud , 39 ff.