How should we measure inequality?

At least since Kaldor (1961), the constancy of the labour share of income has been considered one of the key foundations underlying macroeconomic models. In Karabarbounis and Neiman (2014a), we documented a global decline in the share of labour compensation in gross income (‘gross labour share’) since 1975 and emphasised the role of declining investment prices for this trend. Piketty (2014) and Piketty and Zucman (2014) also discussed this factor share movement and linked it to increases in the capital–output ratio. Their theory focuses on labour compensation in net income (‘net labour share’), which excludes depreciation. The rationale is that depreciation is not consumed and, therefore, the net labour share may more closely approximate inequality between workers and capitalists.

Summers (2014) and Rognlie (2014) emphasise that the net labour share is likely to increase even if the gross labour share decreases. If capital and labour are sufficiently substitutable in gross production, a decline in the real interest rate will reduce the cost of capital and cause a large enough increase in the real capital–output ratio to reduce the gross labour share. Summers and Rognlie correctly note that if the value of depreciation increases in proportion with the capital–output ratio, it will cause the net labour share to increase relative to the gross labour share. For realistic values of parameters, this likely causes the net labour share to rise even when the gross labour share falls.

The data, however, do not bear this prediction out. In Karabarbounis and Neiman (2014b), we measure trends in gross and net labour shares for a large number of countries. We start with our national accounting dataset, which allows us to compare these trends in the corporate sector as well as for the overall economy. The US is a bit of an outlier and exhibited a meaningfully smaller net labour share decline of 2.6 percentage points compared to the 4.7 percentage point decline in its gross labour share. More generally, however, countries experienced highly similar declines in both labour share measures. We additionally analysed factor shares and depreciation in the newly released Penn World Tables 8.0 dataset. Figure 1 plots trends in the gross labour share since 1975 on the x-axis against trends in the net labour share on the y-axis and paints a similar picture. The two labour share measures by and large moved together.

If one believes the gross labour share declines resulted from reduced interest rates, this should come as a surprise. The Rognlie (2014) and Summers (2014) logic would in this case apply and the points in Figure 1 should be clustered along or above the x-axis. By contrast, in Karabarbounis and Neiman (2014a), we argue empirically that reductions in the price of capital were the critical drivers of gross labour share declines. And in Karabarbounis and Neiman (2014b), we show theoretically that in response to reductions in the price of capital, the gross and net labour shares will always move together. Our result follows because, unlike shocks to the real interest rate, reductions in the price of capital not only reduce the gross return to capital but also simultaneously reduce the value of depreciation.

Figure 1. Gross and net labour share trends in the Penn World Tables 8.0

Finally, in Karabarbounis and Neiman (2014b), we ask when one should prefer using gross or net labour share measures to study inequality. We simulate the transition of an economy in response to a number of shocks and compare the behaviour of the gross and net labour shares with welfare-based measures of inequality. Workers in our economy consume their wages each period, whereas capitalists own the entire capital stock and make dynamically optimal consumption and saving decisions. We illustrate that during an economy’s transition, the gross labour share is sometimes more informative about inequality than the net labour share, depending on the nature and timing of the underlying shocks.

Published in collaboration with VoxEU

Author: Loukas Karabarbounis is an Associate Professor of Economics at the Booth School of Business, University of Chicago. Brent Neiman is an Associate Professor of Economics at Booth School of Business, University of Chicago.

Image: The heat impressions of people waiting for a bus are left in a cold morning frost on the glass of a transit shelter in Vancouver, British Columbia February 6, 2012.   REUTERS/Andy Clark