Ludwig Wittgenstein pointed out that many words don’t have a simple precise meaning, but instead refer to a set of concepts with a sort of family resemblance. I thought of this when reading a twitter exchange, which was triggered by a recent post I did over at TheMoneyIllusion:

There are a number of different ways that one could address this question.  For simplicity, I’ll focus on a fiat money system, as I’d probably define monetary policy slightly differently under a gold standard.   

Monetary policy occurs:

1. When a monetary authority . . .

2. uses a set of monetary policy tools . . .

3.  or signals intentions to make future use of policy tools . . .

4.  to impact the supply and/or demand for the medium of account, . . .

5.  which impacts nominal aggregates such as the price level and NGDP.  

Later I’ll return to the question of whether only actions by the monetary authority (such as a central bank) should count.  For now, let’s look at the other parts of my definition.

In the US, the monetary base (cash plus reserves) is the medium of account.  The Fed has traditionally had three policy tools that impact base supply and demand, but those tools have changed over time:

Policies affecting the supply of base money:  Open market operations and loans to banks.

Policies affecting the demand for base money:  Reserve requirements and interest on reserves (IOR).

That’s four tools, but prior to 2008 IOR did not exist and today reserve requirements do not exist.  So mostly three tools. Basically, the Fed adjusts the supply of base money by the purchase and sale of assets, as well as loans of base money.  Demand is impacted via changes in IOR.   More supply of base money is expansionary, ceteris paribus, and more base demand is contractionary.  But, and this is important, policy is not necessarily expansionary when the base is increasing, and it’s not necessarily contractionary when higher IOR is tending to increase base demand.  Lots of other things matter, by far the most important of which is signals about the future path of policy, that is, the future use of the three major policy tools.

You could make an argument that monetary policy should refer not just to actions taken by the monetary authority that impact the supply and demand for base money, but also to any other action that impacts the purchasing power of money. Thus, suppose Bill Gates were to shift $50 billion of his wealth from stocks to US currency placed in safe deposit boxes?  That increases the demand for base money, doesn’t it?  Why not consider his action to be monetary policy?  

David Andolfatto makes a related point in response to Nick Rowe:

I can give you two reasons why I don’t think it’s useful to expand the definition of monetary policy beyond the monetary authority:

1. The behavior of the monetary authority is a very important part of government policymaking.  It is useful to have a term that applies to monetary actions taken by this specific institution, and have other terms (currency hoarding, fiscal policy, etc.) for actions taken by other entities that might impact the value of money.

2.  As a practical matter, I don’t think that currency hoarding by Bill Gates would impact the value of the dollar.  I suspect that the Fed would respond by injecting enough extra base money to offset the effect that Gates’ action might otherwise have on the value of money.  In the same way, I suspect that the Fed would roughly offset the effect of more fiscal spending on prices and nominal spending.  And if it didn’t, I’d still call that passivity an “expansionary monetary policy”.  (A good example is the Fed’s expansionary policy during 2021, which was discretionary, not forced by fiscal policy.) 

I do not believe this monetary offset argument applies in countries like Zimbabwe.  But even in those cases (of fiscal dominance), the first reason I provided is sufficient for having a special term to designate policies of the monetary authority that impact the supply and demand for base money.  I’d prefer to call deficit spending in Zimbabwe “fiscal policy”, while acknowledging that it would likely impact inflation.

PS.  Here’s another tweet by Andolfatto:

I’m not sure how much weight anyone puts on my views, but I will say that Milton Friedman would have certainly viewed that definition as being wrong.  (Although I suppose it could technically be correct if one were to define monetary policy as policy affecting both the path of interest rates and the path of the natural interest rate.)