Money Demand and Income Inequality: International Evidence using a Century of Data

Transactions and precautionary theories of money demand imply that aggregate money demand declines as income becomes more unequally distributed. In this paper I test this prediction using money demand regressions and find mixed results. Focusing first in the USA for the 1913-2016 period, I find support of the theory (a statistically significant and negative coefficient on top 1% and top 10% income shares) under the specification in levels for the whole period. For the regression in first difference, I find a statistically significant relationship for the Bretton Woods period (1945-1973), but not for the entire sample nor other historical periods such as the war and inter-war period (1914-1944) and the great moderation (1985-2007). Bootstrap regressions furthermore also provide support under some specifications in levels, but suggest a lack of relationship in the first difference form. I then explore the international evidence, using an unbalanced panel of 18 countries for the 1913-2016 period. The panel regression in levels supports the theory but the first difference specification contradicts it. Furthermore, there is considerable country heterogeneity. I argue that the link between money demand and inequality have important implications for the distributional impact of monetary policy. This is mainly due to the fact non-linearities in money demand imply that the inflation tax is regressive.