Has monetary policy cared too much about a poor measure of r*?

Measures of the decline in r ∗ are (almost) exclusively based on a fall in government bond yields. Yet, measures of the return on private aggregate capital are roughly constant, or slightly increasing, in most advanced economies. This fact is robust across measures of the demand for capital, using different estimates of capital income or reproducible capital, and across measures of the supply of capital based on the Euler equation. In a neoclassical model with limits to leverage that cause a misallocation of capital, government bond yields only affect affect output via capital returns, and the patterns in the data can be accounted for by an increase in primary deficits and the public debt crowding out private capital. With nominal rigidities and the potential for secular stagnation, the benefits from higher inflation and aggregate demand stimulus are smaller, while the benefits from supply-side reforms that improve the allocation of capital are higher.