Skip navigation to content

Quantifying enfranchisement

Principal Investigator: Myles Lavan

This project aims both to solve a long-standing problem in the history of the Roman empire and to model a new approach to quantification in ancient history. The initial work was supported by a Research Fellowship from the Leverhulme Trust in 2014-2015.

The problem

Rome's unusual generosity with its citizenship is often cited as a key factor in its success as an imperial power. The incorporation of outsiders figures prominently in the mythical histories that Romans wrote of their own early history, such as the famous story of ‘Romulus’ asylum’ on the Capitoline hill. The first new citizens were freed slaves (where other cities relegated ex-slaves to the status of resident aliens, the Romans made them citizens) and Italians, as the Romans enfranchised immigrants and annexed some conquered territory. But Roman citizens remained a minority of the Italian population up to the first century BCE, when a war between the Romans and their Italian allies – the so-called ‘Social War’ – convinced the Romans to extend citizenship to all Italy south of the Po. The next step was the enfranchisement of the provinces. Increasing numbers of provincials became Roman citizens by serving in the Roman army, holding magistracies in their own communities or by a special imperial grant to themselves, their family or their city. This process reached in its culmination in 212 CE when the otherwise infamous emperor Caracalla granted citizenship to all free inhabitants of the empire. 

The problem is that we still have a very poor sense of the scale of enfranchisement before Caracalla’s grant. Research to date has focussed on identifying Roman and non-Roman names in inscriptions from the provinces. That is a useful way of getting some sense of local differences in the level of enfranchisement, but it has failed to produce a conclusive picture of the overall situation. Some scholars think citizenship was widespread by 212; others think it remained relatively rare; no one will commit to even an approximate percentage. The majority seem to think the level of enfranchisement is unknowable given the current evidence base and have resigned themselves to writing the social and political history of the empire without a firm understanding of this important process.


This project aims to break the deadlock by taking a new approach. It starts from the fact that we know the mechanisms by which enfranchisement worked, such as service in the Roman army. For each mechanism, we can identify the key parameters that will have governed the rate of enfranchisement. In the case of the army, for example, these include the size of the army, the average length of service and survival rates. It is then a relatively straightforward matter to build a mathematical model that will calculate the progress of enfranchisement based on any given set of values for those key parameters. The challenge is how to deal with the fact that many of the parameters are themselves highly uncertain. The sheer number of uncertain variables involved in calculating the overall level of enfranchisement might seem to doom the analysis.

My solution borrows a technique from other fields with more sophisticated methodologies for managing uncertainty. Traditional modelling techniques in ancient history are deterministic, i.e. they produce a single ‘best estimate’ for the unknown quantity based on inputted most likely values for all the component parameters. More sophisticated approaches to uncertainty rely on stochastic models which allow all the parameters to vary simultaneously. Both inputs and outputs are expressed as probability distributions rather than discrete values. This makes it possible to estimate the likelihood of all possible outcomes based on what we know about the key parameters driving the spread of citizenship.

Results and publications

Lavan, M. (2016), ‘The spread of Roman citizenship, 14-212 CE: Quantification in the face of high uncertainty’, Past & Present 230: 3-46. [Full text]

This article presents the methodology and the preliminary results, which suggest that we already know enough about the mechanisms of enfranchisement and the demography of the empire to rule out much of the possibility space. It argues that it is extremely unlikely that more than a third of the free population of the provinces had Roman citizenship in 212.  The article also discusses the implications for our understanding of the potential impact of Caracalla’s grant, whose effects were probably much more extensive than many scholars have thought. 

leverhulme trust logo

Portrait of the emperor Caracalla from a statue reworked as a bust; picture derived from original by Marie-Lan Nguyen (2011, Wikimedia Commons)
Caracalla (photo: M.-L. Nguyen 2011)

Quantifying enfranchisement project chart
Estimating the likelihood of different outcomes for the level of enfranchisement by simulating thousands of possible scenarios

Find the School on