Convergence in TFP among Italian Regions - Panel Unit Roots with Heterogeneity and Cross Sectional Dependence

GDP. Very little attention, on the other hand, is dedicated on TFP.³ We be-
lieve that this is an important limitation of the existing literature. TFP
reflects a wide array of both tangible and non tangible factors that influence
the efficiency of the economy, and production in particular. Since the persis-
tence of the spatial differences in Italy can largely be rooted in the efficiency
of the production system, TFP is a variable that requires the necessary at-
tention.

The importance to look at TFP stems also from the need to look at the
persistence (hence the structural nature) of the process of convergence . In
comparison with previous papers, this study captures also the more recent
evidence using data spanning from 1970 until 2001. We root our methodology
in the work of Evans e Karras (1996), who try to analyse the process of
convergence among US states through the identification of a common trend.
Evans e Karras (1996) introduce a particular notion of convergence, claiming
that the different economies converge if and only if there exists a common
trend such that

Et (y-n,t +1 ^_ at+1) = μ-n                            (1)

Moreover, if μ_n = 0, convergence is absolute, and if μ = 0, convergence is
conditional. This methodology has been applied to Italian regions by Mar-
gani and Ricciuti (2001) to analyse the process of convergence in regional per
capita GDPs during the period 1951-1998. These authors estimate a high
rate of convergence for the entire period, but reject the hypothesis of abso-
lute convergence and accepting that of conditional convergence. Moreover,
they break the period into two sub-periods going from 1951 to 1973 and 1974
to 1998, and find evidence of absolute convergence for the first period and
divergence for the second, a result already reached by other studies. How-
ever, with respect to the analysis of clubs of convergence their results are
less conclusive. This is unfortunately a general feature of the literature on

³Only recently we can record some exceptions. See, for example, Di Liberto, Mura,
Pigliaru (2003, 2004).

<<   1   2   3   4   5   6   7   >>

More intriguing information