sequence of low market integration and low substitutability between domestic and imported foods. In addition, if the geographical area over which these multipliers are calculated increases, the share of nontradables in the economy rises and multipliers are correspondingly larger. This leads the authors to praise the potential for a demand-led approach to both agricultural and industrial development, anchored in productivity gains in traditional exports. While the general proposition is reminiscent of the development strategies advanced by Mellor (1986), Adelman (1984), and Hazell (1984), and by now well accepted, the quantification of multiplier effects is inadequate. The model used makes unrealistic assumptions that systematically bias upward the size of the growth linkages induced by technological change in farm or nonfarm tradables. Denoting by T = tradables, T, = farm tradables, NT = nontradables, E = elasticity of supply response, ,qM = elasticity of substitution in consumption between output (q) and imports (M), and m, = income multiplier of an extra $1 of income from farm tradables, the key assumptions made by the authors are T: E = 0, , qM= o; and NT: E = o, YqM = a value which does not matter as long as E = oo. This yields local m, between 1.3 and 1.8 reported in table 2 of the Delgado et al. paper. Consider now the more realistic characteriza-
Read full abstract