Abstract

Unemployment problems in urban areas of less developed countries throughout the world have stimulated increasing interest in analyzing the impact of wages on labor supply and demand and in formulating appropriate wage policies. The urban-rural wage differential is a significant variable in attempts to measure the determinants of labor supply through migration from rural areas (Beals et al., 1967; Greenwood, 1969; Knight, 1972; Rourke and Sakyi-Gyinae, 1972). The level and growth of wages relative to capital costs may influence the labor intensity of investment and production through substitution, and hence affect the rate of growth of employment (Fajama, 1973; Frank, 1971; Harris and Todaro, 1969; House, 1973; Maitha, 1973; Roemer, 1971, forthcoming; Senga, 1973). Increasing attention is being paid to inequalities and changes in the distribution of wages and income (Ewusi, 1971; House, 1975; McCabe, 1974) and to the trade-off between unemployment and wage or price inflation (Ajayi and Osayimwese, 1974). Many of these studies stress the importance for employment growth of wage restraint relative to the price of capital and relative to productivity and product price increases over time. Hence it is essential for governments under pressure both from unions and from unemployment to have an accurate measure of the level and growth of the appropriate wage when trying to make wage policy decisions. One difficulty for empirical analysis is that there is in fact no single and a particular measure must be chosen to represent the entire structure of wages or the relevant portion of it. All of the studies cited rely principally on the arithmetic mean wage. Studies of migration sometimes include theoretical discussions of how wages should be measured in analyzing migration decisions (Knight, 1972). Rourke and Sakyi-Gyinae (1972) present minimum as well as mean wage data as an indicator of what migrants might expect to receive. Ewusi (1971) analyzes the distribution of employment by a detailed wage category and refers to the minimum wage in discussing the relative position of low-skilled workers over a period of time. In statistical analysis, however, the mean is invariably used. The median wage is an alternative measure of the average which may be more appropriate for the purpose of studies such as those cited. Researchers as well as government statisticians have paid scant attention to the applicability of the mean or to the possible usefulness of the median. This article analyzes the relative advantages of the median, mean, and minimum in representing the average wage, with reference to their differences and to trends from 1957 to 1971 in Ghana. It is argued that the median is generally the most satisfactory single representation of the wage paid to lower-income workers, and that it is sufficiently different from the other measures to warrant making median wage data more available. AFRICAN STUDIES REVIEW, Vol. XIX, No. 1 (April 1976).

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call