Spillover of human capital and physical capital are also important for the poor economies
only.
Overall, the catch-up effect seems to be an important determinant of productivity
growth in almost all sectors. More especially, its effect is much stronger in the mining, FIRE,
manufacturing and government than the other sectors. Moreover, poor economies seem to
show stronger significance on the catch-up effect.
The results suggest three important notions. First, growth process are different
whether aggregate or disaggregate data are considered. Results obtained for the total of all
sectors differ from those obtained in the disaggregate sectors. Therefore, generalizing results
for all sectors based on aggregate data may be misleading. Second, within the same sector the
sign and magnitude of coefficients are not always consistent when distinctions are made
between the low and high GDP states. This suggests that the determinants of the
productivity growth process vary across economies (poor and rich). Third, the effects of
human capital, and its domestic and spillover effects vary across sectors and results are
mixed with regards to the sign of these factors. The negative coefficient on human capital,
and its domestic and spillover effects in several sectors is unexpected. A priori, it was
expected that human capital would be positively correlated with growth. Moreover, the
domestic effect and the spillovers that accompany human capital should be expected to
enhance GDP growth as well. Benhabib and Spiegel (1994) also observed a negative sign on
the coefficient of human capital for a sample of countries in their study.
The estimated growth model seems to indicate that the catch-up with the technology
leader dominates the growth process, mainly for the states that start-off with relatively low
GDP levels. Indeed, the catch-up effect seems to be more consistent across sectors. It is
positive in almost all sectors and more significant in the group of initially low GDP levels.
17