ABSTRACT
This chapter focuses on the large empirical literature on the determinants of growth. This has produced a number of contradictory findings. For instance, Levine and Renelt (1992) have argued that the aggregate physical investment ratio is one of the more robust cross-sectional determinants of economic growth. By contrast DeLong and Summers (1991, 1992, 1994) dispute this role for aggregate investment and controversially argue that equipment investment is the only form of physical investment that can improve the growth record of an economy and that investment in private and public non-residential structures make no contribution. This chapter addresses this issue by undertaking cointegration analysis of output per capita and the components of investment using a long span of data for six OECD countries. It is argued below that the equipmentgrowth nexus is not a unique association and is not based on a simple, uni-directional, causal relationship. The empirical methodology of DeLong and Summers is questioned and an alternative methodology based on time series techniques is suggested. Empirical results are then reported, followed by a conclusion.
