World Economics, Vol. 19 No. 3 July-Sept 2018., 23rd October 2018
Key Points
• Piketty claims that capital rises faster than income which assumes that capital and income are independent variables.
• The evidence shows they are not and that the ratio of income to capital is probably stable over time.
• Piketty’s mistake appears to come from confusing depreciation with the cost of maintenance.
• Piketty uses misleading data to support his claims by confusing wealth with capital and income with output.
Summary
Thomas Piketty’s book Capitalism in the 21st Century has the central thesis that unless capital is destroyed by war, it rises faster than income and that this process is inexorable until equilibrium is reached when the value of capital is equal to eight times income.
To justify this claim wealth is used as a proxy for produced capital, national income is used as a proxy for domestic income and national capital as a proxy for domestic capital. At least for the US over the past 88 years these confusions are unnecessary as official data are available for both the value of the domestic produced capital stock and the income it produces. Far from showing an inexorable rising trend the data indicate that the ratio of the value of the capital stock to income is probably stable over the long-term.
The long-term produced capital to income ratio for the US appears to rotate around an average of 2.9 and in 2010 was 3.1, compared with the ratio of 4.1 that Piketty derives for that year. The only other country for which I have been able to find similar data is the UK, but these are only available since 1995 which is too short a period to judge whether the capital/income ratio is mean-reverting around a stable mean. In 2010 the UK ratio was 2.4, compared with 5.2 estimated by Piketty.
The value of the capital stock is the accumulated value of past savings, after deducting capital consumption. Piketty gives examples in which the accumulated value of past savings grows faster than income and the ratio therefore rises. These examples assume that capital consumption, which must be deducted to give the value of past savings, is independent of the rate of growth. But it is not, as the rate of depreciation, which is the major part of capital consumption, varies with the rate at which real wages rise. I conclude that the claim is erroneous and depends on an invalid model in which capital consumption and growth are independent variables. Piketty’s inaccurate definition of depreciation indicates that the basic error lies in confusing maintenance with depreciation.