Search

Measuring labor's share of an economy's aggregate income seems straightforward, at least in principle. Count up wage and salary income, along with the value of benefits provided to employees, and divide it by total income. However, one fundamental concept of labor's share in macroeconomic theory is not the amount of aggregate income paid out to labor. Rather, it is the share of aggregate production that is attributable to raw units of labor. Or, otherwise stated, it is the share of aggregate income that would have been paid to laborers if they had no accumulated stocks of human capital.1 This share corresponds to an aggregate production function parameter: the elasticity of output with respect to physical (i.e. non-augmented or raw) units of labor (Robert Solow, 1957). In this paper we estimate annual raw labor’s share for the US, 1949 to 1996.

In general, empirical studies on economics rely on the assumption of constant capital share of income both at the aggregate level and at the sector level. However, there is no empirical evidence supporting the constancy of capital share at the sector level. In this paper, using Colombian data, we measure capital share for 48 sectors during the period 1990-2005. We also explore the relation between capital's share and factor prices and the behavior of capital share during the business cycle. The main results are the following: (i) capital share is not constant but, rather, has an increasing trend; (ii) capital shares growth rates positively correlate with sector value-added growth; (iii) the capital shares behave pro-cyclically; and (iv) there is a positive correlation between capital shares and real wages and a negative correlation between capital shares and interest rates. These results suggest that the usual assumption of constant factor shares is not accurate.

The relative stability of aggregate labor's share constitutes one of the greatmacroeconomic ratios. However, relative stability at the aggregate level masks theunbalanced nature of industry labor's shares – the Kuznets stylized facts underlie those of Kaldor. We present a two-sector – one labor-only and the other using both capital andlabor – model of unbalanced economic development with induced innovation that canrationalize these phenomena as well as several other empirical regularities of actualeconomies. Specifically, the model features (i) one sector (goods production)becoming increasingly capital-intensive over time; (ii) an increasing relative price andshare in total output of the labor-only sector (services); and (iii) diverging sectorallabor's shares despite (iii) an aggregate labor's share that converges from above to a value between 0 and unity. Furthermore, the model (iv) supports either a neoclassical steadystate or long-run endogenous growth, giving it the potential to account for a wide range of real world development experiences.