1 See briefly W. Scheidel, “Progress and problems in Roman demography”, in id. (ed.), Debating Roman demography (Leiden 2001) 72 and n.294.

2 Critical engagement with my handling of the evidence or the adduction of further information may either corroborate or undermine my interpretation. At best, my model will be judged to have cast new light on the workings of the economy of Roman Egypt; at worst, it will encourage the formulation of more adequate explanations. Either way, deductive reasoning would seem the best way to make sense of scattered and fragmentary data.

3 Duncan-Jones, R. P., “The impact of the Antonine plague,” JRA 9 (1996) 108–36Google Scholar.

4 See Littman, R. J. and Littman, M. L., “Galen and the Antonine plague,” AJPh 94 (1973) 243–55Google ScholarPubMed; McNeill, W. H., Plagues and peoples (New York 1977, repr. 1998) 130–32Google Scholar. The Antonine plague may have been the western arm of a pandemic that struck China between A.D. 161 and 185: cf. Duncan-Jones (supra n.3) 117 and n.102. However, Chinese demographers reckon with a precipitous demographic contraction from 185 to 221, i.e., after rather than during the epidemic ( Chu, C. Y. C. and Lee, R. D., “Famine, revolt, and the dynastic cycle: population dynamics in historic China,” Journal of Population Economics 7 [1994] 354 CrossRefGoogle ScholarPubMed fig. 1).

5 Duncan-Jones (supra n.3) 120-34. For textual accounts see ibid. 118-20. Village head counts were already discussed by Rathbone, D. W., “Villages, land and population in Graeco-Roman Egypt,” ProcCambPhilSoc 36 (1990) 114–19Google Scholar. See also Casanova, G., “Altre testimonianze sulla peste in Egitto: certezzi e ipotesi,” Aegyptus 68 (1988) 93–97 Google Scholar.

6 Duncan-Jones (supra n.3) 136. But see below n. 94.

7 The Egyptian data were presented by Duncan-Jones (supra n.3) 120-25, 133-34. For regional variation, see ibid. 134-36, and now Ehmig, U., “Die Auswirkungen der Pest in antoninischer Zeit,” ZPE 122 (1998) 206–7Google Scholar.

8 Rathbone (supra n.5) 119.

9 Bagnall, R. S. and Frier, B. W., The demography of Roman Egypt (Cambridge 1994) 135–55CrossRefGoogle Scholar; Frier, B. W., “Natural fertility and family limitation in Roman marriage,” CPh 89 (1994) 318–33Google ScholarPubMed.

10 For a summary see, e.g., Hayami, Y., Development economics: from the poverty to the wealth of nations (Oxford 1997) 52–77 Google Scholar. On the chances of economic development under these circumstances, see Lee, R. D., “Malthus and Boserup: a dynamic synthesis,” in Coleman, D. and Schofield, R. (edd.), The state of population theory: forward from Malthus (Oxford 1986) 96–130 Google Scholar; Woods, J. W., “A theory of preindustrial population dynamics,” Current Anthropology 39 (1998) 99–135 CrossRefGoogle Scholar.

11 E.g., Lee, R., “Population homeostasis and English demographic history,” in Rotberg, R. I. and Rabb, T. K. (edd.), Population and economy (Cambridge 1986) 75–100 Google Scholar.

12 Actual population size is unknown, and attempts to derive it from figures reported in ancient sources are dubious at best (for the most recent attempt, see Cascio, E. Lo, “La popolazione dell'Egitto romano,” Stadi Storici 40 (1999) 425–47Google Scholar. Some scholars prefer estimates based on comparative considerations: e.g., Rathbone (supra n.5) 107-10; Bagnall and Frier (supra n.9) 53-56. In a new study, I propose a maximum grand total of 5 to 7 million before the Antonine plague: Scheidel, W., Death on the Nile: disease and the demography of Roman Egypt (Leiden 2002) chapt. 3Google Scholar. A starvation diet would have supported an even larger population.

13 Again, the scale of urbanisation is controversial: modern estimates range up to 37% (Bagnall and Frier [supra n.9] 56). I am inclined to settle for 25% including Alexandria: Scheidel (supra n.12) chapt. 3. It is true that urbanisation might also be taken as a rough index of economic development (independent of population growth): however, the natural fertility régime of Roman Egypt makes it seem unlikely that population and economy would grow at significantly divergent rates. The model of the “low-equilibrium trap” (in which modest increases in output result in population growth rather than higher rates of saving and investment: e.g., Hayami [supra n.10] 113-15) is the most appropriate default position for an ancient population.

14 London: Wrigley, E. A., “A simple model of London's importance in changing England's society and economy 1650-1750,” Past and Present 37 (1967) 44–70 CrossRefGoogle Scholar. Rome: Morley, N., Metropolis and hinterland: the city of Rome and the Italian economy 200 B.C.–A.D. 200 (Cambridge 1996) 33–54 CrossRefGoogle Scholar; cf. also Scheidel, W., “Germs for Rome,” in Edwards, C. and Woolf, G. (edd.), Rome the cosmopolis (Cambridge, forthcoming)Google Scholar.

15 Scheidel (supra n.12) chapt. 2.

16 For general considerations on the possibility of relative over-population in (parts of?) the Roman empire in the 2nd c. A.D., see Frier, B. W., “More is worse: some observations on the population of the Roman empire,” in Scheidel, (ed.) (supra n.1) 139–59Google Scholar.

17 The most plausible estimates are 200 people per km² in Egypt compared to 15 people per km² in the empire as a whole: for the former, see Scheidel (supra n.12) chapt. 3; for the latter, Frier, B. W., “Demography,” CAH XI (2nd edn., Cambridge 2000) 812, 814 Google Scholar.

18 Sallares, R., The ecology of the ancient Greek world (London 1991) 465 n.367Google Scholar.

19 Littman and Littman (supra n.4) 254-55. In their view, at a “median mortality rate” of 7 to 10%, the initial outbreak in the late 160s would have caused 3.5 to 5 million deaths, with a similar number of people dying up to and especially in 189. Assuming that their calculations are based on a gross population total of 50 million (a variable left unquantified in their article), the cumulative death toll would have amounted to 14 to 20%.

20 Y. Zelener, chapt. 2 of his forthcoming Ph.D. diss. (Columbia Univ.); I thank him for showing me a draft.

21 Rathbone (supra n.5) 119 reckons with losses of 20 to 30%, which cannot be more than a guess. For a discussion of drops in population figures for individual villages and their significance, see ibid. 114-18.

22 This summary is based on Gottfried, R. S., The Black Death: natural and human disaster in medieval Europe (New York 1983) 129-60, esp. 135–45 on the economyGoogle Scholar; Ziegler, P., The Black Death (New York 1969) 232–51Google Scholar; Lütge, F., “Germany: the Black Death and a structural revolution in socioeconomic history,” in Bowsky, W. M. (ed.), The Black Death: a turning point in history? (New York 1971) 80–5Google Scholar; Hatcher, J., Plague, population and the English economy 1348-1530 (London 1977) esp. 31 ff.CrossRefGoogle Scholar; id., “England in the aftermath of the Black Death,” Past and Present 144 (1994) 3-35; Pirillo, P., “Peste Nera, prezzi e salari,” in La Peste Nera: dati di una realtà ed elementi di una interpretazione (Spoleto 1994) 175–214 Google Scholar; Herlihy, D., The Black Death and the transformation of the West (Cambridge, MA 1997) 39–57 Google Scholar; Frank, R. G., “Effects of the Black Death in England: an essay review,” Journal of the History of Medicine 54 (1999) 596-605, esp. 603 Google Scholar. (For some qualifications, Cf. Clark, G., “Microbes and markets: was the Black Death an economic revolution?,” Journal of Economic History 59 (1999) 487 [abstract].)Google Scholar For pertinent sources, see Horrox, R., The Black Death (Manchester 1994) esp. 277 ff., 292 ffCrossRefGoogle Scholar. In general, see also North, D. C. and Thomas, R. P., The rise of the western world: a new economic history (Cambridge 1973) 71–98 CrossRefGoogle Scholar. The impact of the plague on real wages is well brought out by the regional data studied by Hamilton, E. J., Money, prices, and wages in Valencia, Aragon, and Navarre, 1351-1500 (Cambridge, MA 1936) 77, 180, 187 Google Scholar: in the absence of plague, real wages remained stable in Valencia, whereas in Navarre they rose by two-thirds between 1350 and 1445, stimulated by severe epidemic mortality. The Cairo example is taken from Gottfried 138. See also Dols, M. W., The Black Death in the Middle East (Princeton 1977)Google Scholar for conditions in the Middle East (including Egypt), esp. 255-80 on the economic consequences of the plague.

23 Since farming in the Fayum was more heavily dependent on artificial irrigation than in the Nile valley, it was likely to be more sensitive to changes in the labour supply. For a separate discussion of the Oxyrhynchite, see Rowlandson, J., Landowners and tenants in Roman Egypt: the social relations of agriculture in the Oxyrhynchite nome (Oxford 1996) 248 and 249 fig. 4Google Scholar, noting a 3rd-c. decline in rents following a 2nd-c. peak.

24 From Drexhage, H.-J., Preise, Mieten/Pachten, Kosten und Löhne im römischen Ägypten bis zum Regierungsantritt Diokletians (St Katharinen 1991) 159-63, 169-70, 175-76, 178 Google Scholar. The data for the Oxyrhynchite have been checked against the most recent list in Rowlandson (supra n.23) 332-47 (prefigured in P.Oxy L p. 216 Google Scholar), which yielded only one additional reference ( P.Oxy LV 3800 Google Scholar).

25 From Drexhage (supra n.24) 165-66, 172-73. Rowlandson (supra n.23) 332-47 adds one further reference for the Oxyrhynchite (SB VIII 9918).

26 Neeve, P. W. De, ‘The price of agricultural land in Roman Italy and the problem of economic rationalism,” Opus 4 (1985) 77–109 Google Scholar.

27 Alston, R., Soldier and society in Roman Egypt (London 1995) 108 CrossRefGoogle Scholar Table 6.2, based on Drexhage (supra n.24).

28 Records of land sales only occasionally specify what the plots were used for. In two cases, land sold for less than 600 drachms per aroura was grown with grain; in three cases, it had previously supported viticulture (but was currently unused or used differently); in one case, it was planted with olive-trees. While no arable farming is mentioned in the higher price category, there are three references to active vineyards, one to oleiculture, and one to palm trees. This sample, albeit exiguous, seems to indicate that the more expensive land was commonly planted with trees.

29 From Drexhage (supra n.24) 141-54.

30 The pioneering work was Johnson, A. C., Roman Egypt to the reign of Diocletian (Baltimore 1936)Google Scholar, the second volume of the Economic survey of ancient Rome edited by T. Frank.

31 Drexhage (supra n.24); Rathbone, D., “Monetisation, not price-inflation, in third-century Egypt?,” in King, C. E. and Wigg, D. G. (edd.), Coin finds and coin use in the Roman world (Berlin 1996) 321–39Google Scholar; id., “Prices and price formation in Roman Egypt,” in J. Andreau et al. (edd.), Économie antique: prix et formations des prix dans les économies antiques (Saint-Bertrand-de-Comminges 1997) 183-244. The last-mentioned survey now supersedes the study of wheat prices by Duncan-Jones, R., Structure and scale in the Roman economy (Cambridge 1990) 143–55CrossRefGoogle Scholar. For another brief survey of price trends, cf. also Duncan-Jones, R., Money and government in the Roman empire (Cambridge 1994) 26–32 CrossRefGoogle Scholar.

32 From Rathbone 1997 (supra n.31) 217-23. I follow the conventional assumption that 7 obols equalled 1 drachm, and convert references to drachms accordingly. There is no significant difference between the price of wheat in private transactions and in involuntary transactions with the state: in the first period, 9 prices in the former category (dating from c.100-160) average 8.01 dr (median 8), while 11 state transactions average 8.19 dr (median 8). In the second period, all but 2 prices pertain to private sales. Earlier price records on ostraca from Upper Egypt (Duncan-Jones [supra n.31] 152) were omitted.

33 In several cases, chronological precision was not feasible; all 3rd-c. prices date from before 274.

34 From Rathbone 1997 (supra n.31) 223-33.

35 From Drexhage (supra n.24) 47-48.

36 From Drexhage (supra n.24) 271-79.

37 From Rathbone 1997 (supra n.31) 234-38.

38 Rathbone 1996 (supra n.31) 332; 1997 (supra n.31) 215. Greater precision in the timing of these price shocks is neither possible nor necessary: to judge from comparative evidence, they may occur with some delay even if they are directly caused by the epidemic event, and intensify incrementally. In Navarre, for instance, it took prices 30 years to double after the initial outbreak: Hamilton (supra n.22) 163. In England, it took 20 years for them to rise by 50%: Farmer, D. L., “Prices and wages, 1350-1500,” in Miller, E. (ed.), The agrarian history of England and Wales vol. 3: 1348-1500 (Cambridge 1991) 437 Google Scholar fig. 5.2. Cf. also Dols (supra n.22) 259-60 (Egypt). For this reason, the objections of Muth, R. F., “Real land rentals in early Roman Egypt,” Explorations in Economic History 31 (1994) 210–24CrossRefGoogle Scholar, against the computation of means and medians for particular periods are misguided: in a case like this, standard measurements (such as polynomial trend-lines) are bound to conceal important discontinuities. Cf. already Rathbone 1997 (supra n.31) 239 n.5.

39 Rathbone 1997 (supra n.31) 215. Compare the charts in North and Thomas (supra n.22) 14, 74; Farmer (supra n. 38) 437,457, and the previous footnote.

40 See Drexhage (supra n.24) 287-94.

41 Rathbone 1997 (supra n.31) 243 n.61.

42 Duncan-Jones (supra n.3) 122; for the Antonine plague, see Ael. Arist., Or. 48.38-39 and Hdn. 1.12.1.

43 Rao, A. R., Smallpox (Bombay 1972) 78–79 Google ScholarPubMed.

44 From Drexhage (supra n.24) 412-25; Rathbone, D. W., Economic rationalism and rural society in third-century AD Egypt: the Heroninos archive and the Appianus estate (Cambridge 1991) 156–58Google Scholar. No data from Upper Egypt appear to be available for the 3rd c.; therefore, comparisons are confined to Middle Egypt. A text from Memphis ( BGU 114 in Drexhage, 423 Google Scholar; A.D. 255) contains a number of exceptionally high daily wages for agricultural tasks. Wages may well have been higher in an urban or suburban setting, and since the other records commonly originate from villages, I have calculated separate rates for the 3rd c. excluding the Memphite data. In the following, these results will be included in parentheses.

45 Already noted by Duncan-Jones (supra n.3) 124.

46 From Drexhage (supra n.24) 425-29; Rathbone (supra n.44) 136-38.

47 See, e.g., Johnson (supra n.30) 301-6; Schwartz, J., Les archives de Sérapion et de ses fils: une exploitation agricole aux environs d'Hermoupolis Magna (de 90 à 133 p.C.) (Cairo 1961) 327–38Google Scholar; Drexhage (supra n.24) 440-54. The most recent contribution, Muth (supra n.38), is marred by incorrect assumptions: see below n.54.

48 Schwartz (supra n.47) 334.

49 Source: Tables 4-7. All data of known origin come from Middle Egypt. The figures in parentheses are based on all daily wages from the 3rd c. excluding the Memphite data (see above n.44).

50 Source: Tables 6-7 (wages); Rathbone (supra n.44) 464 (wheat prices).

51 We must bear in mind that these comparisons draw on data not merely from different locations but also from different social and economic contexts, and from different parts of the year; agricultural prices and wages are sensitive to seasonal fluctuations in supply and demand. There is, however, no compelling reason to reckon with systematic biases in favour of uncharacteristically high (or low) prices (or wages) in either period under review that would distort century-wide averages. The data from the Heroninus archive, which dominate the 3rd-c. record, are the only possible exception, but even they originate from a number of different years and seasons.

52 Under which circumstances would the value of arable decrease relative to the price of wheat, given that land-values are a function of the return on land? The answer is that, in this case, uninterrupted taxation by a strong central state and accelerating urbanisation kept up demand for grain, while the relative shortage of labour made it more expensive to grow. As non-proprietorial producers (tenants and wage-labourers) absorbed a larger share of the surplus than before (indicated by higher real wages and lower real rents), thereby reducing the profit margins of the landlords, the price of wheat and the value of arable moved in different directions.

53 In his recent book, The city in Roman and Byzantine Egypt (London 2002)Google Scholar, R. Alston also reckons with a demographic contraction in the late 2nd and early 3rd c. van Minnen, P., “Agriculture and the ‘taxes-and-trade’ model in Roman Egypt,” ZPE 133 (2000) 212 Google Scholar, observes that the recorded drop in mean agricultural yields after the 2nd c. is consistent with demographic contraction caused by the Antonine plague.

54 In his analysis of smaller samples, Muth (supra n.38) concluded that real wages fell alongside real rents and land values, even though he was compelled to admit that this was not true of agricultural wages and that the statistical basis for this finding was weak. Besides, his statistical technique forced him to discard the valuable evidence from the Heroninus archive. He consequently maintained that overtaxation rather than a shortage of labour was best consistent with the evidence. This reading appears to based on the implicit (erroneous) assumption that taxes were levied in addition to private rents, squeezing the landlord's share if tenants lived at subsistence level and the state raised taxes. For the landlord's obligation to pay land-taxes, see Herrmann, J., Studien zur Bodenpacht im Recht der graeco-aegyptischen Papyri (Munich 1958) 122–23Google Scholar; Rowlandson (supra n.23) 226.

55 Gottfried (supra n.22) 129-34. However, it remains controversial whether further outbreaks or lowered fertility were more important in delaying demographic recovery. In favour of disease, see Hatcher 1977 (supra n.22) 55-62; for the possibility of low fertility after the Black Death, compare Platt, C., King Death: the Black Death and its aftermath in late-medieval England (London 1996) 34–38 Google Scholar. Bailey, M., “Demographic decline in late medieval England: some thoughts on recent research,” Economic History Review 49 (1996) 1–19 Google Scholar, stresses that there is no simple answer to this question.

56 Thus Casanova, G., “Epidemie e fame nella documentazione greco d'Egitto,” Aegyptus 64 (1984) 179–80Google Scholar, on Dio 75.13.1-2.

57 Sallares (supra n.18) 251. Epidemic smallpox affects all age-groups; in its endemic variety, it becomes a disease of early childhood.

58 Zelener (supra n.20). These later secondary outbreaks could have resulted in further population decline but probably caused fertility to rise, eventually creating a new state of equilibrium.

59 Hatcher 1977 (supra n.22) 57.

60 Casanova (supra n.56) gives a good idea of the shortcomings of the available evidence.

61 See Zinsser, H., Rats, lice and history (Boston 1935) 138–41Google Scholar; Casanova (supra n.56) 174-75; for mass mortality in Alexandria, Cf. Parkin, T. G., Demography and Roman Society (Baltimore 1992) 63–65 Google Scholar.

62 For mean life-expectancy at birth in late mediaeval and early modern Essex, with and without plague (c.32-34 years and the mid/upper 30s, respectively), see Poos, L. R., A rural society after the Black Death: Essex 1350-1525 (Cambridge 1991) 118 CrossRefGoogle Scholar. These rates are higher than for Egypt around 1900: see Scheidel (supra n.12) chapt. 3.

63 See Scheidel, W., “Roman age structure: evidence and models,” JRS 91 (2001) 21–24 Google Scholar; id. (supra n.12) chapt. 2.

64 Scheidel (supra n.12) chapt. 2, contra Bagnall and Frier (supra n.9) 177. Moreover, demographic modelling suggests that following a catastrophic epidemic, the intrinsic growth-rate will be higher than usual for the first 15 years or so, before dropping below the long-term average for the next 30 years — in keeping with changes in the relative distribution of different age-cohorts: Paine, R. R., “If a population crashes in prehistory, and there is no paleodemographer there to hear it, does it make a sound?,” American Journal of Physical Anthropology 112 (2000) 189 fig. 53.0.CO;2-9>CrossRefGoogle Scholar.

65 North and Thomas (supra n.22) 14 fig.2.1; Farmer (supra n.38) 437 fig. 5.2. For nominal wages, cf. also Poos (supra n.62) 211 fig. 10.2.

66 For the long-term inelasticity of real prices and rents in Roman and early Byzantine Egypt, cf. J.-M. Carrié , “L'arithmétique sociale de l'économie agraire: prix de la terre, rente foncière et prix des céréales dans l'Égypte romano-byzantine,” in Andreau et al. (supra n.31) 105-46.

67 Penn, S. A. C. and Dyer, C., “Wages and earnings in late medieval England: evidence from the enforcement of the labour laws,” Economic History Review 43 (1990) 356–76CrossRefGoogle Scholar cast doubt on the value of real wages as an index of true earnings and living standards.

68 Johnson (supra n.30) 305.

69 Dyers, C., “Changes in diet in the Late Middle Ages: the case of harvest workers,” Agricultural History Review 36 (1988) 21–37 Google Scholar. On the higher increase in the real wages of urban workers, see, e.g., Lütge (supra n.22) 82 (Germany); Farmer (supra n.38) 437 fig. 5.2 (England); or Ashtor, E., Histoire des prix et des salaires dans l'Orient médiéval (Paris 1969) 373 (Mamluk Egypt)Google Scholar.

70 Wiet, G., “Le traité des famines de Maqrizi,” JESHO 4 (1962) 75 Google Scholar; Udovitch, A. L., “Egypt: crisis in a Muslim land,” in Bowsky, (supra n.22) 125 Google Scholar; Dols (supra n.22) 271. At the same time, because of the expansion of urban labour, artisans prospered: Dols 267-70.

71 See above nn. 69-70.

72 Braunert, H., Die Binnenwanderung: Studien zur Siedlungsgeschichte Ägyptens in der Ptolemäer-und Kaiserzeit (Bonn 1964) 252 Google Scholar (large villages), 176-79 (migration to cities), 193 (tenants), 266 (population exchange between villages), 269, 273-4 (epibole and pressure on élites). Cf. also his discussion of P. Oslo 111, indicating population loss in Oxyrhynchos (270-73). Notwithstanding earlier claims, land flight may have been relatively rare up to the mid-2nd c. A.D. (thus Link, S., “Anachoresis: Steuerflucht im Ägypten der frühen Kaiserzeit,” Klio 75 [1993] 306–20CrossRefGoogle Scholar), and may have increased from the time of the plague onwards (Braunert 164-83).

73 Rathbone 1990 (supra n.5) 123; R. Alston, “Urban population in Late Roman Egypt and the end of the ancient world,” in Scheidel (ed.) (supra n.1) 167, with refs. Flowever, Alston deems it unlikely that the documented pattern of urban élite expenditure was compatible with demographic contraction, assuming that workers rather than élites would benefit from population losses (167 and n.27).

74 Sharp, M., “The village of Theadelphia in the Fayyum: land and population in the second century,” in Bowman, A. K. and Rogan, E. (edd.), Agriculture in Egypt: from Pharaonic to modern times (Oxford 1999) 185–89Google Scholar, esp. 188: a drop from 1600-1730 hectares of arable and c.140 ha of vineyards and orchards in the year 158/9 to 990 ha of arable and 415 ha of arboricultural land in 216 (based on P. Berl. Leihg. I 5 Google Scholar; P. Stras. VII 688 Google Scholar). (In fig. 1, I have chosen the lower total for the first figure in order to minimise the extent of subsequent changes.) The supposed population total of c.2500 residents in the 120/130s (Sharp 164) tallies well with the total amount of arable before the epidemic, assuming that c.90% of all teenage and adult men cultivated c.2.5 ha per person, a ratio applied to 19th-c. Egypt by Cuno, K. M., “A tale of two villages: family, property, and economic activity in rural Egypt in the 1880s,” in Bowman, and Rogan, 316 Google Scholar. This calculation suggests that, prior to the epidemic, the land was exploited at or near full capacity; subsequent population loss would have alleviated this pressure.

Comparative data from the late Middle Ages suggests that reports of land abandonment should be taken with a grain of salt. On English estates, contraction in the scale of peasant cultivation was significantly less pronounced than the rise in ‘vacant’ holdings. It seems that ‘unused’ land could still be productive, especially when converted into pasture, even if this did not register in official records: Titow, J., “Lost rents, vacant holdings and the contraction of peasant cultivation after the Black Death,” Agricultural History Review 42 (1994) 97–114 Google Scholar. Similar developments are at least conceivable for Roman Egypt.

75 Bagnall, R. S., “ P. Oxy 4527 and the Antonine plague in Egypt: death or flight?,” JRA 13 (2000) 291 Google Scholar.

76 Gottfried (supra n.22) 139.

77 We must allow for the possibility that the increase in viticulture was a function of the growth of large estates: see van Minnen, P., “Dietary Hellenization or ecological transformation? Beer, wine and oil in Later Roman Egypt,” in Andorlini, I. et al. (edd.), Atti del XXII cong. int. di papirologia, Firenze 1998 (Florence 2001) 1279 Google Scholar. Even so, the availability of spare land and higher demand were necessary preconditions for an expansion of arboriculture.

78 Bagnall (supra n.75) 292.

79 Cf. Rathbone (supra n.5) 124-37; on Theadelphia, see above n.74.

80 Bagnall (supra n.75) 288-92.

81 van Minnen, P., “P.Oxy. LXVI 4527 and the Antonine plague in the Fayyum,” ZPE 135 (2001) 175–77Google Scholar.

82 Bagnall (supra n.75) 291. For population pressure, see above n.16.

83 E.g., Mate, M., “Agrarian economy after the Black Death: the manors of Canterbury Cathedral Priory, 1348-91,” Economic History Review 37 (1984) 341–54CrossRefGoogle Scholar.

84 Hatcher 1977 (supra n.22) 32. Again, this effect is explained by previous over-population (33).

85 Hatcher 1994 (supra n.22) 33-34.

86 For the demographic conditions, see above n.64.

87 van Minnen (supra n.81) 176.

88 Bagnall (supra n.75) 291.

89 For the impact of the Black Death, see above, nn.22, 38, 69-70. For price shocks following the epidemics of the 17th c., see Fischer, D. H., The great wave: price revolution and the rhythm of history (Oxford 1996) 94–102 Google Scholar.

90 Prices of commodities and services may rise for a variety of reasons. Mass mortality induced an increase in the stock of money per capita, helping to push up prices: Farmer (supra n.38) 441. It is true but also too simplistic to say, as Dols (supra n.22) 255 does (with regard to Mamluk Egypt), that “the substantial modifications in prices and salaries were symptomatic of a sustained dislocation of commerce and labor by depopulation”. The Black Death showed that price effects could be more persistent than actual production and distribution failures. Psychological factors come into play: unstable price signals and the rupturing of trust-based exchange-networks would increase the perception of risk and unpredictability, causing prices to soar.

91 After the Black Death, it generally took commodity prices more than a generation to start falling again, and sometimes much longer: e.g., North and Thomas (supra n.22) 14 fig. 2.1; Herlihy (supra n.22) 47.

92 Rathbone 1997 (supra n.31) 216. Pace Rathbone 1996 (supra n.31) 328, it may not be true that coin issues following the debasement of 176/77 were small: compare Duncan-Jones (supra n.31) 31 fig. 2.2, for a substantial spike in coin output around 190. Even so, it seems doubtful whether coin debasement in the late 170s could have resulted in full-scale inflationary price-adjustment by the 190s. Cf. also van Minnen (supra n.81) 176-77.

93 For general accounts of the bullion shortage, see Day, J., “The great bullion famine of the fifteenth century,” Past and Present 79 (1978) 3–54 CrossRefGoogle Scholar; Spufford, P., Money and its uses in medieval Europe (Cambridge 1988) 356–62CrossRefGoogle Scholar; Fischer (supra n.87) 48. For a shortage of copper in Egypt in the first half of the 15th c., see Shoshan, B., “Money supply and grain prices in fifteenth-century Egypt,” Economic History Reviezv 36 (1983) 60 Google Scholar; silver stocks fell from c.1380 ( Ashtor, E., A social and economic history of the Near East in the Middle Ages [Berkeley 1976] 305 Google Scholar), resulting in a liquidity crisis c. 1400/1415 ( Ashtor, E., Les métaux précieux et la balance des payments du Proche-Orient à la basse époque [Paris 1971] 44 Google Scholar). For the contraction of money supply as a driving force behind price depression, see, e.g., Mayhew, N. J., “Numismatic evidence and falling prices in the fourteenth century,” Economic History Review 27 (1974) 1–15 CrossRefGoogle Scholar; Farmer (supra n.38) 441-42. Credit could not compensate for the lack of cash because its availability was tied to that of bullion: Nightingale, P., “Monetary contraction and mercantile credit: later medieval England,” Economic History Review 43 (1990) 560–75CrossRefGoogle Scholar.

94 Postscript: In a forthcoming paper, J. Greenberg challenges Duncan-Jones’ contention that numerous samples of documentary evidence reflect the impact of the Antonine plague. To my mind, his critique further underscores the need for comprehensive deductive models that reduce the likelihood of ad hoc reasoning.