Geoscience Reference
In-Depth Information
2009
; Bradley et al.
2003
) or by mixing both actual interregional transfers
and the territorial incidence of interpersonal ones into the indicator (Rodden
and Dragu
2010
), the field continues to neglect the multidimensionality in the
architecture of redistribution. However, insofar as the conclusions of this study
hold, there is much to gain in the future by analytically decomposing the differ-
ent tools through which redistribution occurs in both developed and developing
nations. This will require not only a massive data collection effort on the ter-
ritorial incidence of different policy tools but, more importantly, a theoretical
effort to better understand the incentives of incumbents in centralized regimes
to target funds to different regions and/or population subgroups.
In addition, the full understanding of the multidimensional nature of redis-
tributive conflicts in complex societies requires to take on several other chal-
lenges. These relate directly to some of the simplifying assumptions adopted
in this study. The first one concerns the interplay between identity and redis-
tribution. In many political unions, identity is a major factor driving political
preferences, party alliances, and public policy (Fearon and Van Houten 1998;
Horowitz
1985
). In the theoretical argument of this topic, I have treated iden-
tity both partially and indirectly. But this by no means implies that identity
based dimensions do not matter. Quite to the contrary, as reflected by a grow-
ing literature on the role of second dimensions in the formation of preferences
for redistribution (Austen-Smith and Wallerstein
2008
; Glaeser
2005
; Roemer
and Woojin
2006
) this is hardly the case. While the need to prevent additional
complexities in an argument already involving several dimensions justifies its
exclusion from the analysis of the relationship between economic geography
and preference formation, a thorough analysis of the interplay between identity
and inequality during the process of preference formation for fiscal structures
remains a major task ahead in comparative politics.
Identity also matters for the process of preference aggregation through the
system of political representation. On this point, I have limited my approach
to the assumption that the presence of regional political parties mobilizing
ethnic identities is a major factor contributing to the centrifugal nature of
political representation. The Spanish experience offers an illustration of how
the presence of nationalist parties as major players in the political system
triggers and sustains a centrifugal spiral in the representation of contending
political views. Identity based parties limit the nationalization of parties and
alter the feasible set of political coalitions in the lower house. Moreover, as
reflected by the Spanish and other experiences, this tendency tends to reinforce
itself over time.
Within such a process, the boundary between contentions about fiscal struc-
tures and the politicization of the federal contract itself becomes increasingly
blurred. A full grasp of this process requires taking yet another step: tackling
the endogenous nature of political representation over time. An important con-
cern in this topic was whether political representation was
fixed
at the time
an exogenous alteration of the patterns of economic geography triggered the
