As we have seen, Appian and Plutarch describe the second century as a period during which Italy’s free population declined as a result of the expansion of rural slavery. Although twentieth-century scholarship has found fault with their accounts in numerous details, their general sketch of the background to the Gracchan land reforms is still very much on the table. Thus Hopkins’ contention that Italy’s free country-dwelling population declined from c.4.1 million to c.2.9 million between 225 and 28, while the slave population increased from c.500,000 to roughly 2 million, is essentially a sophisticated restatement of the old theory that the expansion of rural slavery pushed large numbers of free peasants off the land. Recent research has embellished this picture with many additional nuances, but every attempt to offer a radically new reconstruction has either suffered from serious methodological weaknesses or simply failed to fit the surviving evidence convincingly.
Despite this, even a superficial reexamination of some of the basic assumptions underlying most recent work on this subject is enough to reveal unexplored possibilities, especially with respect to the demographic reconstructions that inform many publications on late republican history. Even if Lo Cascio’s ‘‘high’’ theory of demographic development is rejected, the extant data can be used to support a reconstruction of agrarian development during this period that is substantially different from any put forward so far.
One of the assumptions I have in mind concerns the expansion of rural slavery. As we have seen, Hopkins assumes that early imperial Italy had some 2 million slaves, of whom 1.2 million worked in the countryside. One of the reasons why these figures do not inspire much confidence is that the few pieces of relevant information in the ancient sources are compatible with a very wide range of estimates. Even if Appian’s assertion that some 120,000 slaves took part in Spartacus’ uprising is correct, for example, it does not permit us to infer that there must have been 2 or 3 million slaves in late republican Italy. More generally, the fact that Hopkins’ scheme assumes a third of Italy’s population in 28 to have been slaves raises the suspicion that the figure of 2 million was inspired by the fact that 33 percent of people in the antebellum South were slaves.4 Finally and perhaps most importantly, a labor force of 1.2 million rural slaves would have been wildly in excess of that required to work the wine and olive-oil estates of the elite. Jongman has demonstrated that fewer than 200,000 hectares of Italian land could have kept Italy’s entire urban population fully supplied with wine and olive oil during the early Empire.43 Now, according to the Roman agronomists it took one slave to work 7 iugera (1.75 ha.) of vineyard, while oleoculture required an even lower labor input per hectare (Col. Rust. 3.3.8; Pliny HN 17.215). 4 It seems realistic to assume from this that Italian arboriculture in general required one slave for every two hectares. Combining this ratio with Jongman’s estimate of the amount of land required to keep the cities and towns of Italy supplied with wine and olive oil suggests that this level of production would have demanded only 100,000 slaves. Even if we assume that it took another
100,000 slaves to assist and/or supervise this force and that 50,000 more were used to grow the cereals consumed by all the estate staff, we will still have a figure of only 250,000. The theory that there were 1.2 million rural slaves in late republican Italy can thus only be maintained by assuming that some 80 percent of the rural slave work force was used to grow grain.45 Although slaves are known to have been used in grain farming (cf. above), it seems doubtful whether any ancient historian would be prepared to defend this extreme hypothesis. The only other way to push up the number of rural slaves would be to assume that hundreds of thousands of slaves were employed not on slave-staffed villae but on small family farms. To the best of my knowledge there is no evidence to support this theory.
For the purposes of this contribution it is important not to lose sight of the fact that the production figures underlying my calculations relate to the early years of the Empire, when Italy’s towns and cities probably had about 1.9 million inhabitants.4 Italy must, however, have been far less extensively urbanized in 133 than in (say) 28 BC and the number of slaves needed to grow grapes and olives for the urban market of that time can only have been correspondingly lower. Moreover, if the early imperial figures may require upward adjustment to take substantial wine exports from Italy into account, the same is unlikely to apply to the period 201-133. It may therefore be suggested that even if slaves were widely employed in grain production during this period, the total number of slaves in Italy is likely to have been far smaller than is usually thought. This finding is entirely compatible with the theory that the development of slave-staffed estates reduced the amount of land cultivated by free peasants in certain parts of Italy. It seems, however, far-fetched to suppose that regional developments of this sort brought about a decline in the number of free country-dwellers in Italy as a whole.
If the number of rurally employed slaves has been exaggerated, the question arises whether the data we have on the free Italian population support the traditional view that the free peasantry declined during the second century. As we have seen, Hopkins’ reconstruction of Italy’s demographic history during the last two centuries of the Republic was based on the idea that Italy (including Cisalpina) had some 4 million free inhabitants in the age of Augustus, having had 4.5 million in 225. Although the proper interpretation of the Augustan census figures is a hotly debated issue (see also Chapter 14), the traditional reading on which the former figure is based is almost certainly correct. What, though, of the theory that Italy had 4.5 million free inhabitants in 225? As I have already explained, this figure has been derived by assuming that central and southern Italy had 3 million free inhabitants and Cisalpine Gaul 1.5 million. Both these estimates are, however, open to challenge. The figure for Cisalpine Gaul is based on Brunt’s suggestion that there were probably between
300,000 and 500,000 adult males in this region on the eve of the Second Punic War.47 Extrapolation from the very highest figure in this range suggests that Cisalpina had 1.5 million free inhabitants in all. If the lowest figure in the range is used instead, however, the free population of Cisalpine Gaul drops to around 1 million and that of Italy as a whole to around 4 million (see also Chapter 14).
The theory that central and southern Italy had about 3 million free inhabitants in 225 is equally dubious, being based on a controversial reading of the Polybian manpower figures. The essence of this reading is that these figures included every adult male Roman citizen regardless of his age, but counted only men aged between 17 and 45 in the case of the Latin and other Italian allies. If, however, we assume that all free adult men were counted whatever their age and origin, the free population of central and southern Italy drops from 3 to 2.5 million.48 In short, since the significance of the figures for 225 is a matter of interpretation it is entirely possible that Italy had just as many free inhabitants in 28 as it did in 225. We cannot even rule out the possibility that the combined impact of the Second Punic War, the civil wars of the first century, and the emigration of large numbers of Italians to colonies outside the Italian peninsula was insufficient to prevent the free Italian population from growing from around 3.5 million in 225 bc to roughly 4 million in 28 bc.
A closely related question concerns the quantitative fate of the Roman citizen body. As is generally known, every reconstruction of the demographic trajectory of this Italian subgroup has to rely upon the surviving census figures, whose correct interpretation remains controversial. Even if not all of these figures can be explained, however, there seem to be no good grounds to dispute the widely held view that in principle at least the Roman censors were supposed to register all male citizens age 17 or over, including proletarians and the so-called cives sine suffragio (‘‘citizens without the vote’’).49 Building on this interpretation, many scholars have observed that the census figure for 164/163, when around 337,000 citizens were registered, is substantially higher than any of the figures relating to the third century. This would seem to indicate that the citizen body recovered quickly from the terrible losses suffered during the Second Punic War.50
Interpreting the census figures for the next five decades is far more difficult. The main problem is that the figures for the period 159-130 are slightly lower than that for 164/163 (only around 319,000 citizens being counted in 131/130), but also much lower than the figures for 125/124 and 115/114, when the censors managed to register some 395,000 citizens. Unsurprisingly, the 159-130 figures have been interpreted as illustrating the decline of the free rural population that allegedly lay behind the Gracchan land reforms. A major weakness of this theory is, of course, that it cannot explain the sudden increase reflected in the figures for 125/124 and 115/ 114. The only way around this problem is to follow Beloch’s suggestion that the figures for these years should be amended to around 295,000.51 In other words, the theory of population decline can only be maintained by manipulating the surviving evidence!
For this reason alone it seems preferable to seek an alternative explanation for the figures for the three decades preceding the Gracchan land reforms. One strong possibility is that Rome’s prolonged overseas campaigns (especially the unrewarding wars in Spain) may have made the prospect of military service far less appealing to many adult citizens. Since the most obvious way of dodging the draft would have been to avoid being registered by the censors, such a change in attitude could very well account for the relatively low census figures of the mid-second century.52 In short, while the high census figures for 125/124 and 115/114 are completely incompatible with any theory of population decline, the low figures for the period preceding the Gracchan land reforms can easily be accommodated within a scenario of continuing demographic expansion.
In order to put some flesh on the bones of this alternative reconstruction we must also take into account the growth of the city of Rome (see also Chapter 16). A dearth of reliable data makes it impossible to follow this process in detail, which is why the few estimates of Rome’s population on the eve of the Gracchan land reforms that have been attempted range from around 200,000 to 400,000.53 Despite the uncertainty surrounding them, even these very approximate figures have interesting implications for the numerical fate of the free country-dwelling population during the second century. Even if we allow pre-Gracchan Rome the largest population ever suggested, it is unlikely to have been inhabited by more than 100,000 adult male citizens. Combining this figure with our earlier finding that Italy had about 400,000 adult male citizens in the late 130s suggests that some 300,000 male citizens age 17 or over must have lived in smaller towns or in the countryside. Interestingly, this rough estimate more or less equals Italy’s total number of adult male citizens on the eve of the Hannibalic War.54 If the census figures for 125/124 and 115/114 are correct, then we must conclude that the theory of a drastic decline in the number of free country-dwellers is completely untenable, at least with regard to the second century. This conclusion fits in quite nicely with our earlier finding that the expansion of rural slavery must have been much less dramatic than is often assumed.
Why, then, do our sources explain the Sempronian Land Law of 133 as an attempt to halt a decline in the rural population caused by the steady expansion of slave-staffed estates? Part of the answer to this question may be that this law was prompted by developments in the coastal districts of Etruria (cf. Plut. Ti. Gracch. 8.7) and other parts of central Italy where slave-staffed estates are indeed likely to have pushed a significant number of free peasants off the land (see also Chapter 28). It may, however, also be suggested that Tiberius Gracchus was reacting to an increase in rural poverty without realizing that this was being caused by an ongoing process of demographic growth.55 The census figures for the 140s and 130s, moreover, may well have led him to believe that the free citizen population had begun to decline: not until 125/124 BC would the figures reveal that the Republic now had more citizens than ever before.
If this interpretation is correct, the period 201-133 witnessed both the expansion of rural slavery and an increase in the number of Italy’s country-dwelling citizens. Initially the coexistence of these processes resulted in intense competition for access to public land in Italy.56 When the Roman elite came out on top, political stability could only be preserved by finding an alternative outlet for the expanding free population of Italy. The sending out of some 265,000 adult male citizens to colonies in other parts of the Mediterranean should be seen as part of the solution to this problem.
World History
