When teaching undergraduates or talking with non-economists, I usually argue that, while the Fed can affect aggregate demand (or nominal spending), it has no effect whatsoever on aggregate supply. If economic output declines because productivity growth is lower, there isn’t anything the Fed can do. Output is lower than it would have been with higher productivity growth. But we are doing as well as possible given the growth of productivity actually realized. And, even though we might prefer greater productivity growth and economic output, the Fed is unable to improve matters along these lines.
It is convenient to make such arguments when considering short run macroeconomic policy. Indeed, such a simple view might prevent us from attempting to offset low productivity growth with expansionary monetary policy, generating an unsustainable boom in the process and, ultimately, reducing economic output even further. But such a view is not, strictly speaking, true. There are long run benefits of sound monetary policy. And the economy will underperform in the long run in the absence of sound monetary policy. When the Fed credibly commits to a monetary policy rule, it anchors long run expectations, enables long run contracting, and promotes long run economic growth.
Proponents of discretionary monetary policy usually maintain that, with discretion, the monetary authority can do whatever one’s preferred rule prescribes or something better. And, to some extent, they are right. Rules are binding. Strict adherence to a rule would prevent the monetary authority from taking some courses of action. And, unless the rule is perfectly specified, a course of action that is not permitted by the rule might be preferable to the rule at a given point in time.
While deviating from the course of action prescribed by the rule might seem like a good idea at a point in time, there are benefits to adhering to a rule over time. The problem with discretionary monetary policy, as pointed out by Kydland and Prescott among others, is that it leaves everyone guessing what the monetary authority will do. We spend valuable resources trying to inform our guesses—resources that could be put to other uses. We engage in shorter-term contracting—and incur the costs of re-contracting—to account for new information that becomes available over the course of a project—information that would be irrelevant if we knew how the monetary authority was going to behave. And we forego some long term projects altogether because uncertainty about the future stance of monetary policy renders them too risky. A rule—to the extent that it is credible—anchors expectations. It let’s us know what to expect. And, if we know what the monetary authority will do in the future, it will be much easier for us to decide which production processes to take on. In other words, the benefits are in the binding.
The relevant question, then, is whether the static costs associated with adhering to a less-than-perfect rule at a point in time are greater than the dynamic benefits associated with adhering to a less-than-perfect rule over time. I doubt it. For one, we can (and should!) choose a pretty good rule to minimize the static costs. Moreover, it is not clear we have the information required to successfully conduct real-time countercyclical monetary policy—in which case, the static costs are zero. Strict adherence to a good monetary rule is probably the best we can do.
Given the Fed’s commitment (or lack of commitment) to sound monetary policy, there isn’t much it can do about changes in aggregate supply. Some macroeconomic fluctuation is unavoidable and, indeed, is desirable given the economy’s real constraints. However, credibly committing to sound monetary policy makes us more productive than we would be in the absence of such a commitment. By anchoring expectations, it can make it cheaper to engage in long term projects—freeing up some resources to produce more stuff. From that perspective, the Fed can affect economic production in the long run.