Efficacy of monetary policy under uncertainty

In this post, I am going to talk about the efficacy of monetary policy in the face of uncertainty. 

In an earlier post, I have talked about uncertainty reducing interest rate sensitivity – but does that mean that the efficacy of monetary policy has declined? No, as ultimately, we need to think about how any investment response translates into a change in output!

The key idea here is that an uncertainty shock may do more than just reduce investment for a given interest rate, or reduce the interest sensitivity of investment. It may also change consumption or export activity in a way that makes a given change in investment MORE potent at increasing output.

This may be a bit confusing at first glance, so lets think through an example.

Talking through the idea

Let’s go back to our simplified investment function I = a – b*r.

A reduction in b implies that, for a given change in r (interest rate), investment function “I” will change by less. I have shown this in a graph in an earlier post where the I was unchanged following the shock, and still corresponded to the optimal level, I*, at the same interest rate – instead all that was different was how responsive investment was.

But we know that monetary policy is about the change in Y (output) from the change in r.

In that post we noted that if this investment level was unchanged then investment isn’t the cause of a downturn. However, uncertainty can function in other ways. Previously we talked about it reducing A (a lower amount of investment at the same interest rate). But what about, for example, if we imagine that uncertainty also made some proportion of consumers credit constrained due to the response of financial intermediaries.

This credit constraint would reduce consumption below its desired level, would reduce demand, and would lead to an economic cycle where the central bank would like to stimulate demand.

So do we have a problem here?

Now when they cut interest rates, due to lower interest sensitivity investment will not change by as much! As a result, interest rates need to be cut by more to make the investment shock the same size as the consumption shock … which would imply that there is a problem.

But it isn’t just the shocks that matter – the answer comes from thinking about the actual change in output/income (Y). The change in Y = change in I * multiplier.   

The multiplier is 1/(1-MPC) [MPC=marginal propensity to consume] if we use a simple model where consumption is the only variable that changes as current income changes, and that consumers consume some “fixed proportion” of a given change in income.

So, we have a shock that reduces b. But that same shock also increases MPC. Why? Previously the MPC for a change in income was some proportion of income based on peoples expectations of future income – now it is close to 1 as these same individuals are credit constrained and would be willing to spend any increase in their incomes.

Here this smaller lift in I increases expenditure, and translates into a proportionally large increase in consumption as it circulates around the economy. As a result, the relatively smaller investment response ends up with a large change in Y due to the same underlying shock – namely the response of the economy to investment uncertainty!

In summary we have a situation where firms and banks are “uncertain” and less responsive to interest rates. Households who wanted to borrow to consume are currently unable to due to that uncertainty. As they are credit constrained , their spending out of an income boost will be close to MPC = 1. As a result, such a shock could reduce the interest sensitivity of investment (b) but increase the multiplier, thereby not making monetary policy less effective.

If we view uncertainty as an intrinsic part of the economic cycle, then this concept mirrors recent literature asking about credit constraints and how they generate asymmetries in the response of the economy to policy change. 

As a result, this discussion tells us two things:

  • we need to be careful linking a decline in interest rate sensitivity with a decline in the ability of central banks to increase output
  • the specific characteristics of shocks are important, and can create asymmetric and counter-intuitive responses to policy changes.