Skill sorting, inter-industry skill wage premium, and production chains: evidence from India 1999-2000

This paper proposes a mechanism that links industry’s technological characteristics (i.e. quality of non-labor inputs, which is proxied by the length of industry production chains), industry-specific skill wage premium, and skill sorting across industries. It is hypothesized that high-skilled workers are sorted into industries where they can receive a higher skill wage premium, by working with better quality non-labor input. The quality of non-labor inputs is assumed to be worse in industries with longer production chains due to the increased involvement of low-skilled labor and poor infrastructure over the sequential production. By examining Indian wage and employment data for 1999-2000, empirical evidence to support this mechanism can be obtained: First, the skill wage premium is lower [higher] in industries with longer [shorter] production chains. Second, the skill wage premium is lower [higher] in industries with a higher [lower] proportion of low-skilled workers producing inputs outside their own industry. Third, the proportion of high-skilled workers is larger in industries with shorter production chains and lower ratio of low-skilled labor involved, i.e., a skill sorting trend can be observed.

Is there any premium for unobservable quality? : a hedonic price analysis of the Malagasy rice market

This paper examines if consumers pay a premium for unobservable quality in the absence of quality standards and/or quality grading systems and, if so, how they assess that unobservable quality, using a rice retail market in Madagascar as an example. In Madagascar, the lack of quality standards and/or grading systems for rice makes is considered to be one of the causes of the rice market's spatial disintegration. Thus, quality standards and grading systems will be necessary to increase the market's efficiency. We hypothesize that consumers and retailers use product origin and rice name as observable indictors of unobservable quality and test the hypothesis using hedonic price regressions. We find that the interaction terms of product origin and rice name significantly affect the price after controlling for both observable quality and spatial and temporal price variation, but that the contribution of product origin and rice name to rice price variation is smaller than spatial and temporal factors. We thus conclude that consumers pay a premium for unobservable quality throughout Madagascar. This finding implies that quality standards and/or grading systems will work in the Malagasy market and that improving market infrastructure such as roads and storage will make them even more effective.