It occurred to me that the arguments for Supply Side improvements as demand stimulus in a liquidity trap are stronger than I had originally thought.


Well as long as we are in the liquidity trap then the real interest rate is zero. This means that the value of Supply Side improvements after the liquidity trap are necessarily not discounted.

The liquidity trap is like a time warp in which in Discounted Present Value Terms, the end is always tomorrow. Even massively long dated supply improvement should affect demand now.

It means also that the carrying cost of real capital is simply depreciation.

What makes this doubly interesting is that among the slowest depreciating capital goods are structures.

Yet, it is precisely investment in structures that are in the dumps right now.

This wraps around in almost an obvious way, but it suggests that Minsky type moments of the kind Eggerston and Krugman model generally do not persist because if they occurred they would simply induce a boom in the building of structures.

Indeed, that is of course how the business cycle generally operates.

However, that makes bubbles in the price of structures very dangerous. Once the price of structures begins to fall this impedes their use as collateral. This in turn makes borrowing for the purpose of building structures risky and so raises the actual financing cost of structures well above zero even in a liquidity trap.

The key to a liquidity trap then is not that savers cannot be induced to spend, it is that builders cannot post collateral. If they could the slow depreciation of structures and the perpetual tomorrowness of the liquidity trap would induce more building.

This also shows why a currency zone with very slow or negative population growth can so much more easily get stuck in the trap. The demand for structures may actually be falling over time. This means that the distance until the end of the trap matters and you can’t necessarily produce the backwards cascading that would always call for an increase in construction.

What would be interesting is if historically there is a link between the length of a recessions and the speed of population growth/urbanization.