Age heaping in the 1841 Tuscany census: a reappraisal

The present dissertation discusses the appropriateness of using the age-heaping method as a measure of numeracy. Age heaping is the tendency to round one’s age when declared. Such a tendency is typical of developing countries today and most countries till the nineteenth century. The hypothesis that underlies the line of research based on the age-heaping method is that people who are able to count should declare their ages accurately. By using a large sample of data of the 1841 census of the Grand Duchy of Tuscany, which contains information at the individual level, it will be shown that several anomalies arise by the use of this method. The investigation will be conducted at diverse levels of aggregation, from the individual to the parish to the entire sample. Firstly, widowed people, in particular women, round their ages more frequently. The same occurs for literate women with respect to illiterate women. Secondly, there exists a strong correlation between the Whipple index, measuring age heaping, separately computed for literates and illiterates at the municipal level. This element suggests that there may exist a common factor of age heaping for both parts of the population, which is different from education. Thirdly, with a logit model which computes the probability of declaring a round age, it emerges that there exists a relationship between rounding one’s age and the capacity to read (but not to write) in municipalities with fewer than 9,000 inhabitants. These facts, along with other critical contributions on this topic, suggest at best that the relationship between numeracy and age heaping is more complex than it is usually affirmed. The identification of relevant variables representing institutional, demographic and cultural factors may prove necessary to understand more fully the nature of age heaping