Abstract:
Harberger's 'A Vision of the Growth Process', Presidential Address at the 1998 Annual Meeting of the American Economic Association, provides evidence that contributions to aggregate real cost reduction (RCR) are concentrated in a small number of industries. According to Harberger, this is because the effect of broad externalities - such as those linked to the growth of the total stock of knowledge or human capital, or brought about by economies of scale tied to the scale of the economy - is negligible as compared to industry- and firm-specific causes of productivity growth. The argument is that, if growth was driven mainly by broad externalities, contributions to aggregate RCR ought to be quite evenly distributed across sectors. Hence, the growth process looks more like mushrooms than yeast. In this note, we argue that Harberger's evidence is not incompatible with the yeast vision. More specifically, we show that, if there is heterogeneity in the elasticities of sectoral total factor productivities (TFP) to shocks from other sectors, then concentration in the sectoral contributions to aggregate RCR can occur. But this is true even if sectoral TFP growth processes are completely driven by a broad externality stemming from a single sector. To the extent that the response of a sector's TFP to shocks from other sectors can be seen as a measure of absorptive capacity, we suggest that a realistic description of the growth process must go beyond the simple dichotomy yeast vs. mushrooms, and take account of the intersectoral patterns of knowledge transfer.
