ON THE EMPIRICS OF ECONOMIC GROWTH1

A graph of per capita income in the US shows an upward trend, and the longer is the period covered by the graph, the smoother it looks. By changing the scale of the graph, however, it is revealed that the growth is subject to fluctuations. A similar picture, usually with more pronounced variability, is observed for other countries. The increase in income is accounted for in terms of resource accumulation and technical change. This explanation is more applicable to the first graph, but in itself it is insufficient to account for the variability observed in the growth rates. Similarly, it does not explain why some countries grow faster than others and some hardly grow at all. This disparity in performance indicates that the search for explanations is not exhausted. It is commonly agreed that there is no long-term growth without technical change. But if the technical change triggers growth in some countries, why does it fail in others? This question suggests that there are two pertinent concepts of technology: Available technology (AT) and implemented technology (IT). The first concept covers the total knowledge generated everywhere up until the present. The second concept covers that part of the AT that is actually implemented. Knowledge is generated by research which involves human effort and calender time. Past experience suggests monotonic relationship between research inputs and output. However, not much more can be said fo