impact growth and convergence. This paper therefore revisits the convergence debate for
US industries, and extends previous studies by investigating economic growth and the
process of catch-up to technology leaders for several economic sectors, using data for the
lower 48 US states from 1963 through 1997.
The analysis starts with a standard convergence model that explores convergence
patterns for different sectors in the lower 48 states, using well-known spatial econometric
techniques. Next, a spatially explicit growth model in which technological progress is
endogenously determined is applied to data for nine US industries, categorized as Mining,
Construction, Manufacturing, Wholesale/Retail trade, Transportation and Utilities, Services,
Finance Insurance and Real Estate, Government, and the combined sectors labeled Total.
The remainder of this paper is structured as follows. Section 2 reviews some of the
recent literature on sectoral convergence of productivity, and technological leadership.
Section 3 presents the spatial endogenous growth model, and discusses the estimation
results. Section 4 provides a summary and some concluding notes.
2. Sectoral convergence of productivity levels
The economic growth literature devotes substantial attention to the study of economic
growth or total factor productivity in a cross-country setting. Most studies focus on
aggregate data for national economies, although a few utilize disaggregate levels. For
instance, Dollar and Wolf (1993) examine the productivity growth in individual industries
and the process of convergence of overall productivity growth for a set of developed
countries. They observe that in 1963, the US led in labor productivity for all manufacturing
industries, but over the period 1963—1986, labor productivity of the other countries
converged to the US level in virtually every industry at different rates of convergence. Other
studies concentrate on sectoral convergence within specific regions or countries. For