The demography of Italy in the Roman Republic has been heatedly debated in recent years. Roughly, scholarship is divided into two camps, the so-called ‘low count’, which assumes that the population of Italy in the early Augustan period amounted to about 7 million people (supported by scholars such as Brunt (1971) and Scheidel (1996)), and the ‘high count’, which argues that the number may have been as high as 17 million (as suggested by Lo Cascio (1994) and Kron (2005)). These two theses seem mutually exclusive, and although some have suggested ‘middle counts’ (e.g. Hin 2013), the impasse seems irresolvable. The main problem is that the available source material can be manipulated to support either theory, and that there is no decisive evidence that excludes one of the hypotheses.
One may ask, therefore, whether another book on the topic is necessary and what it could possibly have to add to all the existing theories. Let me state right away that this is one of the most important books on the demography of Republican Italy and that it should be read by anyone interested in the Republic, not just demographic historians. Obviously, the size of the population of Republican Italy is extremely important for any reconstruction of social and economic developments, so that it must be taken into account by anyone studying the period. This book not only gives a clear overview of the current debate on Republican demography, but also argues convincingly that the ‘low count’ is the more likely reconstruction of those suggested so far.
Much of the argument depends on intricate and detailed analysis of the Roman census figures, and this book is no different. However, De Ligt is right in making full use of the archaeological evidence, in particular information on the size of towns and their expansion over time, the clearing of marginal land, and archaeological surveys suggesting a continued presence of small and medium farms throughout Italy. This would suggest that a population decline, which was postulated by some scholars (e.g. Toynbee (1965)) is unlikely, at least in most regions; in fact, De L. suggests that the population of Italy continued to grow slowly until the Antonine plague in the mid-second century a.d. In the ‘high count’ theory, stagnation would already have set in during the Augustan period, brought on by a Malthusian crisis due to population pressure. De L. shows that in fact there was space for population growth in Italy, which is reflected in the growth of towns during the first century a.d. In the Appendix, he collects all available information about the size of all Italian towns; apart from being a very useful overview of recent archaeological research, the results show that the urban population was not as large as would be necessary to make a ‘high count’ possible.
An important part of De L.'s argument focuses on the population of Cisalpine Gaul. Whereas this area was the most densely populated region of Italy in the Early Modern period, it seems to have been relatively empty in the Roman era. Since the central and southern parts of the peninsula were also not as densely populated as in later periods, there is simply no possibility that the population was as high as the ‘high count’ suggests. However, there were indeed some problems with access to land pointing to some population pressure, but these were especially prevalent in central Italy and not representative of Italy as a whole (Roselaar (2010)). An Italy-wide Malthusian crisis is not likely. De L. also shows most of the other arguments of the ‘high count’ to be improbable; for example, the fact that the census figures were taken seriously by Roman politicians shows that there cannot have been a 30 per cent undercount and that we must therefore take them seriously as evidence.
Of course there are weaknesses in the argument; for example, it is not made clear why De L. does not consider a ‘middle count’, which would only need to be one or two million persons higher, to be a valid option (it receives only cursory treatment on pp. 126–8). Furthermore, many of the arguments brought forward hinge on rather small differences in the interpretation of the available census figures, so that a slight change may make the difference between population growth and decline. For example, on pp. 239–41 De L. presents veracious scenarios for the urbanization rate and population density in peninsular Italy that would be needed to achieve the high or low count models. For the low count an urbanization rate of between 12.2 and 21.1 per cent would be needed, but the high count might be possible if the population density was high (which recent research suggests it might have been, see Flohr forthcoming), and an urbanization rate of 11.7 to 19.8 per cent prevailed. There is actually quite some overlap between the possible rates suggested for the high and low counts, suggesting that some reconciliation might be possible; and since all the figures in all reconstructions are hypothetical anyway, the reader is left not quite convinced by the argument. Therefore, since more decisive evidence does not exist, there will always remain considerable room for debate, and this book does not fully clear all doubts. For the moment, however, this book has argued as convincingly as possible that the ‘low count’ — or at most a lower middle count — of the Italian population is far more likely than any higher estimates. It is now the turn of ‘high count’ scholars to argue their case; if they do not manage to do so in a similarly well-argued and well-presented way as De Ligt has done, his book will remain the standard work on Republican demography for many years to come.
