In my last post, I argued that there are significant incentive problems with discretionary central banking. I now turn to an often overlooked, but no less important, class of problems confronting central bankers: information problems.
Money, as an institution, plays an important role in coordinating markets. Money provides a common denominator by which various lines of production can be compared. This allows consumers and producers to communicate and coordinate with each other in a spontaneous, decentralized fashion that maximizes the value of society’s scarce resources. But this tendency toward market clearing requires money prices to reflect accurately traders’ best estimate of relative resource scarcities. Traders’ estimates are accurate when the market for money itself is in equilibrium: when, at the existing “price” of money — frequently proxied by the inverse of the price level — traders want to hold as much money as is supplied. Underlying the desirable properties of market equilibrium is monetary equilibrium; getting the most out of the former requires having the latter.
This description suggests a clear role for society’s monetary institutions: maintaining monetary equilibrium. If a society chooses a central bank as its key monetary institution, and those who staff the central bank act with discretion, then discretionary central bankers are the agents responsible for adjusting the supply of money to meet market demand. Although framing the problem this way makes it sound simple, in practice it is enormously difficult.
First, what counts as money is by no means obvious. In practice, an asset is money if traders treat it as money. In some circumstances, only cash is money. In other circumstances, money-market mutual funds may be treated as money since these funds frequently allow investors to write checks against them. If the demand for any of the various assets people treat as money changes, then money demand can change, either universally or (more commonly) in some circumstances but not others.
Second, even assuming away the above problem, money demand is notoriously difficult to forecast. Empirical monetary economists have long admitted, reluctantly, that money demand is not stable. Money demand can and does change based on traders’ expectations of future economic conditions, innovations in the financial sector, seasonal effects, and a host of other factors. As Dan Smith and I argue in our paper on the information problems confronting central bankers:
Investigations into money demand attempt to discover exploitable relationships between monetary and other macroeconomic aggregates. But these aggregates have no direct causal effect on each other. Instead, changes in the demand for money are made by individuals at the microeconomic level on a daily basis. Macroaggregates, which are probably the best link central bankers can have to actual economic activity, cannot be relied upon to remain constant, precisely because the patterns of individual market actors’ behavior on which these aggregates depend are constantly in flux (Lucas 1976). One of the primary reasons people have for demanding money is uncertainty about the future. As that uncertainty about the future changes, so will individuals’ demand for money. This is the crucial difficulty confronting monetary authorities’ predicting or observing, in real time, changes in the demand for money.
What all this suggests is that the problem facing discretionary central bankers attempting to engineer monetary equilibrium is akin to the problem of a central-planning committee trying to allocate efficiently an economy’s scarce resources. Even assuming away incentive problems — granting, for the sake of argument, that central bankers are angels — the knowledge burden is so immense that the job simply cannot be done well. The best that central bankers can do is try to follow money demand by looking at data about markets in the immediate past. But this is like trying to drive a car on a winding mountain road while looking only through the rearview mirror. At best, it’s going to be a jerky and uncomfortable ride. At worst, we’ll go careening off a cliff.
Discretionary central banking places immense information burdens on central bankers. It isn’t central bankers’ fault that they don’t do their job well. The problem is institutional, not personal. Rather than blame central bankers, we should be looking for alternative monetary institutions that do not suffer from insurmountable information problems. History affords us several. In future posts I will discuss them.