| dc.description.abstract | This dissertation consists of three essays on political economics. First, it studies how changes
in the electoral design affect politicians and voters alike. Second, it studies the impact of
criminal organisations on both the legal economy and the development of society.
The first essay (Chapter 2) investigates the impact that the simultaneity of national and local
elections has on the local electoral stage, in terms of civic and political participation, political
competition and electoral results. Methodologically, it employs a difference-in-differences de-
sign that exploits the staggered nature of the local elections in Italy, where some municipalities
experience the simultaneity, while the rest do not. Voters? participation is positively impacted,
with homogeneous effects along the gender dimension. Political participation mildly increases,
while its competition is affected in a dual and opposite way. Mayors winning in this simulta-
neous setting are more likely to have previous office experience, while no pattern emerges in
terms of educational attainment and out-of-office job characteristics. Estimates survive a series
of robustness checks, including a propensity score matching estimating-sample procedure to
account for possible bias in the simultaneity occurrence, a full-dynamic specification employed
to assess the parallel trend assumption, and a change in clustering unit, Lastly, using the si-
multaneity as a sample restriction, it documents the existence of a split-ticket voting behaviour.
It is worth mentioning that the simultaneity is crucial to investigate such behaviour. Namely,
naively comparing the election results of a local election and a national one with a time dif-
ference may just express a change in policy preferences or a change in ideology. However, the
simultaneous elections setting nullifies these concerns. In this regard, this essay finds support
for the divided government hypothesis as a mechanism.
The second essay (Chapter 3) investigates the impact of criminal organisations on the legal
economy, by studying the effect of the collusion between organised crime and local politicians
in the South of Italy on outcomes related to public procurement. To identify the presence of
organised crime, this essay takes advantage of the staggered application of a national law that
enables the dissolution of a local government when evidence of complicity between local offi-
cials and the criminal organisation. It then measures and quantifies the consequences of this
collusion by using unique data on public procurement and firms. Results from difference-in-
differences estimates show that infiltrated governments present lower average rebates and are
more likely to have non-public negotiations. Results are heterogeneous across procurement
macro-categories and political party colours. The essay employs a series of robustness tests
that mainly include the estimation of a full dynamic specification to test the parallel trend as-
sumption, a range of novel tests related to the staggered difference-in-differences design, and checks related to the nature of the municipal government dissolution and to the local political
stage. Further, employing an intention-to-treat type of setup, it finds that firms winning a
procurement contract in an infiltrated municipality exhibit an increase in the cumulative prob-
ability of victory in a public procurement contract. These firms show an increase in balance
sheet values related to profits and labour, and a reduction in capital, leaving unaffected total
factor productivity. Again, a full dynamic specification is estimated to assess the validity of
parallel trends.
The third essay (Chapter 4) studies the long-term effect of organised crime presence on social
capital. This study combines multiple historical data sources, especially novel ones related to
social capital and organised crime, to study this relationship. Focusing on the expansion of
the criminal organisations in the Centre-North of Italy, it provides two alternative empirical
identification strategies, based on an instrumental variable approach. First, it exploits a law
("soggiorno obbligato") that forced organised crime members living in the South of Italy to re-
settle in the Centre-North area. Second, it exploits the combination of migratory movements
from the South to the Centre-North of the country and the allocation of Marshall plan funds.
Crucially, both instrumental variable strategies show strength and relevance in their first-stage.
Primarily employing a measure of tax compliance, which proxies for civic awareness, instru-
mental variables estimates show that long-term exposure to mafia presence depresses social
capital accumulation. Part of this result is found in other measures of social capital. The es-
say contains a series of robustness checks, including a propensity score matching estimating-
sample procedure for the forced resettlement law, an alternative set of instruments instead of
the combination of migration and public funds allocation, and formal tests for pre-trends for
both. | en |
