Henrik Mouritsen has sent me a summary of his paper given at the Cambridge Epigraphy Day in February, which I post below:
Henrik Mouritsen (King’s College London) discussed the possibility of quantifying Roman manumission using epigraphic evidence. While acknowledging that most inscriptions are of little help in establishing hard statistics in this area, he drew attention to two types of document which may provide more reliable information. The first are the epitaphs of the familial columbaria from the early empire, esp. those of the Statilii and the Volusii, where the ratio of slave to freed suggests a very high manumission rate in elite households. The second type is the municipal alba and particularly CIL X 1403 from Herculaneum. This inscription, long believed to contain the names of the Augustales, is unique in its scale. Even a cautious reconstruction of the fragments entails a total of around a thousand names, the large majority being those of local freedmen, which–given the overall size of Herculaneum’s population–would suggest that a substantial proportion of the free adult males were former slaves.