8 Dynamic Externalities and Structural Change in Kenya

Economic development and structural transformation are dynamic processes in which sectoral interactions are numerous and multidirectional. Recent growth literature has noted that many of these effects occur in the form of externalities not captured in a neoclassical accounting of intersectoral resource flows and price adjustments. This work has emphasized, for example, that increasing returns phenomena such as demand externalities (Murphy, Shleifer, and Vishny 1989a) and human capital spillovers (Romer 1986) can play an important role in leading the growth process. The assessment just presented by Bigsten and Collier in chapter 7 takes a primarily static view of agricultural linkages in Kenya. This chapter points out ways in which such a perspective may overlook important aspects of linkage effects and structural transformation in Kenya. The discussion draws heavily on the successful agricultural and rural development experience in Taiwan to illustrate the potential importance of intersectoral linkages and certain dynamic externalities. While not claiming to disprove the existence of strong growth linkages between agriculture and other sectors, Bigsten and Collier argue that the only significant connection between agriculture and other sectors is through construction. By way of evidence in support of their view they document the effects of two important agricultural shocks, a coffee boom in 1976-79 and a drought in 1983-84. However, their examination of the short-term macroeconomic transmission of temporary shocks to other sectors of the economy provides little insight into the importance of agricultural development to long-run structural transformation, a question of central and pressing concern. It is not surprising that the results of reduced-form empirical specifications and Granger tests that attempt to encompass complicated and simultaneous underlying structural relationships are not robust. For one thing, the time series is too short and considerable noise is created by a sequence of well-documented macroeconomic shocks during the time period analyzed (1965-87). The perception that gains are transitory, which is especially likely to be true of price shocks, will result in far