sector specific fundamentals determining (differences in) energy- and labour productivity
developments. Section 6 summarises and concludes.
2. THE ISSUE OF CONVERGENCE
The concept of productivity convergence has its roots in traditional neoclassical growth
theory, with its central notion of a transitional growth path to a steady state. The Solow-Swan
model (Solow 1956, Swan 1956) postulates convergence of per capita income, driven by the
assumption of diminishing returns to capital accumulation at the economy-wide level. The
dynamics of the model imply that initial differences in per capita income and capital
endowments will vanish in the long run, due to declining growth rates as countries approach
their steady state. In the steady state, diminishing returns are offset by technological progress,
the principal source of long-run economic growth. The new or endogenous growth theory
(Lucas 1988 and Romer 1986, 1990) yields a more diverse picture concerning patterns of
convergence. In this view economic growth is ultimately driven by accumulation of
knowledge or human capital, which is (at least partially) a public good. Hence, cross-country
convergence depends on the extent of international knowledge spill-overs, allowing less
productive countries to catch-up with more advanced economies. As such, endogenous
growth theory supports the old hypothesis of the existence of an ‘advantage of backwardness’
(Gerschenkron 1952), suggesting that being relatively backward in productivity carries a
potential for rapid advance (see, e.g., Abramovitz 1986). At the same time, endogenous
growth theory suggests that growth differentials may persist or even increase: learning
effects, externalities and market imperfections allow for economy-wide increasing returns to
capital accumulation and the existence of multiple steady-states. A mixed view on
convergence patterns also emerges if one takes into account the role of international trade: on
the one hand trade will enhance cross-country convergence through knowledge diffusion and
increasing competition, but on the other hand it may contribute to cross-country divergence
since trade advances international specialization (e.g., Grossman and Helpman 1991).
These various approaches generated some degree of controversy around the issue of
convergence and caused the convergence hypothesis to be the subject of extensive empirical
research (see Islam 2003 for a recent survey). The stage for this convergence debate was set
by Baumol (1986), who reported a strong negative relationship between the intial level of
labour productivity and its subsequent growth over a long period (1870-1979), which he