(1990) and Batten and Karlsson (1996) for a more extensive presentation of the theoretical debate).
More recently a different way analysing this relationship has been formulated that basically involves
the construction of economic models incorporating public capital as one of their basic parameters.
Several such models have been proposed with the objective of combining infrastructure
research with dynamic growth modelling. Barro and Sala-i-Martin have used the production function
approach, as well as attempting to model the impact of infrastructure on different economic growth
rates. Their overall conclusion was that public capital could well have a positive effect on the growth
process (Barro 1990, and Barro and Sala-i-Martin 1992). Similar to these models is that presented by
Alogoskoufis and Kalyvitis (1996), the main difference being that these authors did not assume that
the economy achieves a new steady growth rate1 immediately.
Holtz-Eakin and Schwartz (1994), also using a neoclassical growth model for the (48
contiguous) US states, concluded that infrastructure does not play “an important quantitative role in
the explaining the growth patterns of the states” (p. 20). However, they noted that their results must
be viewed with the caveat that the model does not allow for any interaction between public capital
and private investment incentives.
All the aforementioned models have shed some light on the mechanisms by which public
capital transmits its effects to the economy of the private sector. Some of them are at a totally
theoretical level. Others have been used for empirical research. However, it has not been possible to
use any of them to analyse the different spatial levels of the Greek economy due to the existing data
limitations. Instead, a complete model of a small open economy is has been used here which
incorporates public infrastructure capital, and is similar to that presented by Holtz-Eakin and Lovely
(1996). Section 2 of this paper offers the theoretical construction and assumptions of the model.
Following this, the next part presents the empirical results for the different sectors and spatial
levels of the Greek economy. These results can be classified under two major headings. First, the
findings for the non-manufacturing sector of the economy are provided. More specifically the role of
the public capital in impacting upon the regional Gross Domestic Product (GDP) and its sub-
categories, has been analysed with the use of quasi-production functions.
Secondly, the results for the manufacturing sector of the economy are presented. The Holtz-
Eakin and Lovely model delineates two channels by which infrastructure affects the secondary sector:
one by altering the scale of production (and subsequently the level of total manufacturing output), and
the other by affecting the equilibrium number of manufacturing establishments, perhaps better
described by the phrase ‘returns to variety’. There are again four different spatial levels for the
empirical investigation of the Greek case. The first refers to a regional panel based on the Greek
1 For an introductory analysis of the term ‘steady growth rate’ see Jones (1998) or Mankiw (1997).