Benhabib and Spiegel (1994) also observed positive and significant effect of the catch-up.
The further a state to the technology leader in terms of GDP per capita, the faster is its
productivity growth. This justifies the strong dominance and significance of the catch-up
term for the group of initially low GDP levels. With low GDP per capita at the beginning
period, they converge faster to the technology leader. This is also consistent with prediction
from neoclassical theory where poor economies are expected to grow faster. Due to the
interaction between human capital and the catch-up term in the spatial Benhabib and Spiegel
model, it could be concluded that the effect of human capital on growth is rather indirect,
working through the catch-up term. Human capital by itself does not drive the growth
process, but when interacted with the catch-up term its role becomes more prominent.
Therefore, we could conclude that both geographic and technological proximity are relevant
for the sectoral productivity growth. However, the technological effect is more prominent.
5. Conclusion
In this paper, we have utilized some exploratory and spatial econometric data analysis
techniques to investigate issues of productivity growth, human capital, and technological
leadership in US industries using SIC-based state level data from 1963 through 1997. For
eight industries and the combined total we estimated a simple unconditional convergence
model, an unconditional convergence model allowing for convergence clubs, and an
endogenous growth model incorporating human capital and technological catch-up.
Processes of σ-convergence were not detected in any of the sectors, but all sectors show
strong evidence of β-convergence. Only the wholesale/retail sector exhibits pattern of
convergence clubs with low and high initial GDP states showing different rates of β-
convergence across groups. With regard to the endogenous growth model, results are mixed
for the effects of human capital, and its spillover and domestic effects. However, the catch-
18