Public infrastructure capital, scale economies and returns to variety

Note: t-statistics in parentheses (and henceforth in all tables)

***Statistically significant at 1% level **Statistically significant at 5% level *Statistically significant at 10% level

In tables 2 and 3 the total infrastructure variable has been replaced by its productive and social
components. The regional GDP categories, as well as the regression tests remain the same. Table 2
gives a similar picture to that of table 1. The magnitude of the coefficient for public capital, in the
regression where the dependent variable is the total regional GDP, is small (0.031) and statistically
insignificant. Again the coefficients for public capital are statistically significant, although with
different signs, in cases where the GDP sub-categories, Agriculture, Banking, Housing, Public
Administration, Services, Health were used as the dependent variable. In all these cases the Hausman
specification tests indicated that the infrastructure variable is probably endogenous.

Table 2 The effect of productive public capital (G productive) on the GDP of the non-manufacturing
sectors of Greece, 1982-1991

Equation for per capital regional income (ln)

***Statistically significant at 1% level **Statistically significant at 5% level *Statistically significant at 10% level

The situation is not different when the social part of infrastructure is examined. The results,
given in table 3, show that for the total regional GDP the social infrastructure coefficient is small
(0.015) and statistically insignificant. Here, only when the categories of Agriculture, Mines, and
16

<<   <   13   14   15   16   17   18   19   >   >>