The goal of this paper is to measure the correlation between intergenerational mobility and a series
of signi cant macroeconomic variables such as growth, income, crime or the degree of inequality. Our
knowledge of these correlations at the empirical level is in-existent as measurements of intergenerational
mobility obtained from traditional methods (based on panel data) are scarce, dicult to compare across
countries and almost impossible to get across time.
For example, we do not know how intergenerational mobility (IM) correlates with income. This
is, we do not know if it is larger in richer or in poorer countries. Similarly, we do not know how IM
correlates with growth, independently of their level of income. We do not know how it correlates with
crime, corruption, etc. These are obviously relevant questions for our understanding of how economies
work as well as for the design of social policies.
In this paper we apply a novel measure of IM, developed by Guell, Rodrguez Mora and Telmer
(2007), to a very rich set of Italian data and we are able to to produce comparable measures of IM at
the province level (there are about 100 provinces in Italy). We then exploit the large and signi cant
di erences across Italian provinces to explore how IM correlates with in a large array of socio-economic
variables.
The measure of IM proposed by Guell, Rodrguez Mora and Telmer (2007) is based on the idea
that surnames are informative about family links. Since the distribution of surnames is necessarily very
skewed, with many relatively infrequent surnames, it can be exploited to extract longitudinal information
from a cross section of data. Surnames are largely inherited from parents to children together with other
characteristics that matter for the children’s well-being (such as having a certain occupation or belonging
to a certain socioeconomic group). Hence, the more surnames are informative about the outcomes of their
holders, the more important the characteristics inherited along with surnames must be in determining
such outcomes and the less mobility is there.
Our main data consists of the complete Italian tax records for the year 2005, where we observe
each and every person who submitted a tax form for personal income taxation in Italy, together with
their names and surnames (recoded with numerical ids for anonymity), their taxable incomes and their
province of residence (plus a few other characteristics). Furthermore, we combine these tax records with
data the complete registries of lawyers and politicians, whose actual names and surnames are publicly
available. We are thus able to compute our measure of IM based on di erent outcomes: not only income,
but also the probability of begin a lawyer or a politician.
We explore the correlation between our measures of IM and several socio-economic outcomes of the
province: per-capita income, growth, employment and crime. Further, we also look at some occupation
speci c outcomes, like the eciency of the legal system (measured as the average duration of trials) and
of the political system (measured as the ability of the local administration to spend pre-allocated funds).
Our preliminary results suggest that more social mobility is associated with more value added per
capita, more exports, lower unemployment, more voters turnout and quicker trials.