The Fed: Tell Spellman It's an Art, Not a Science


Economies mutate. This occurs when responses among variables in the economy start to change. In the terminology of economics, the elasticities have changed and they, in turn, change the behavior of the entire economic system.

These can be due to changes in incentives (or the reactions to incentives), changes in the rules (either regulatory or legal), or changes in supply conditions. And then there is technology, which constantly morphs.

So relationships that one has become accustomed to may not hold up as they did for decades. Ultimately, they were not written in stone.

For the designated managers of the US economy, The Federal Reserve Open Market Committee who influence financial variables as a means to bring about desired changes to a national economy such as for employment and inflation, this is a perplexing task.   To manage it, economists have endeavored to model the economic system mathematically based on a history of economic reactions. They then seek to derive simple rules for adjusting the variables subject to their control to bring about the desired end results.

Therein lies the development of “rules of thumb” for how to manage the economy. One such example is the Taylor Rule for interest rates to smooth economic cycles. Another was Milton Friedman’s dictate to keep the money supply growing at a constant rate. And there are a lot more.

However, once these simple rules of thumb are developed, often mutations creep in and the rules no longer work as well (or at all), and policy becomes a perplexing endeavor — and that’s where we are today.

It’s in this context that we should view the work of the besieged members of the Federal Open Market Committee who make the call as to whether or not to attempt to raise interest rates. It’s clear they have been perplexed for some time as to whether or not to pull the trigger.

Rather than scorn these poor public servants, it would be more appropriate to pity these mere mortals who are tasked to keep the economy and financial markets running smoothly as per their instructions via the Full Employment Act of 1946.

The question of raising interest rates at the moment has become a much ballyhooed event for which every investor, financial writer, and taxi cab driver no doubt has his or her own opinion mostly held with near certainty.

After all, it’s a momentous event when, after seven years of buying securities and adding quantum leaps to its balance sheet, the Fed contemplates switching to selling securities in order to lower financial prices and raise interest rates as a means to glide to a new economic growth path.

To undertake the task of economic management that was thrust upon the government about 70 years ago, economists both inside and outside the Federal Reserve attempt to systematically estimate as exactly as possible the restraining and simulative effects of Fed actions.

That is, the financial variables that are life and death to investors are, to the Fed, a means to an end: economic stabilization. More broadly, the mandate is to adjust the financial variables in order to iron out the excesses of the spending cycles — both the highs and the lows relative to the available resources. The objective is to produce smooth spending growth that matches the level of and the growth of the supply capability.

The supply side target is called Potential GDP and it’s not static. That is, the supply side target moves through time when labor and capital resources grow and productivity advances and the matchup would result in no more than 5% unemployment and an inflation rate of about 2%. Undershoot it and the unemployment rate is higher and over shooting would result in higher inflation.

So this process of matching up spending with the growing supply capability is akin to trying to send a rocket to the moon and produce a soft landing. The policymakers are trying to produce a spending trajectory that neither misses the moving target to either the high or low side, least we have unacceptable unemployment or inflation.

Obviously, this requires a good deal of estimation of not just the path of the target but how the system will move and react to policy controls to reach the target.

To do so, the science (or, some would say, the art) of econometrics was developed to mathematically model the systems reactions to the financial variables that the Fed can control based on past data.

From the point of view of the econometric creators of the mathematical version of the economic-financial system…this was the linkage from the Fed’s controlled variable to the variables that Congressman tasked them to hit …some version of full employment and price level stability.

So for many years since at least the 1960s, econometric models of the system of responses have been built and added to each year. Of course, it’s based on the precious little data. Because there are many variables to be estimated, accuracy requires that economists collect a lot of data.

More important, reducing estimation errors also requires collecting more relevant data. For example, in picking the voters’ choice for the Republican nomination, one would not to care to rely on a sample of 15 voters, which is not much greater than the number of candidates. More accurate estimates of many variables derive from a sample of many thousands.

So as to be able to obtain as much data as possible to make estimates of how the variables of the economy react to each other, the economists had to go back to the data bin of past experiences and recreate the data starting in 1929.

But herein lies the rub of the scientific approach: to enlarge the sample size, which is still at bare minimum levels relative to the number of variables that are involved, they looked backwards and relied on history — but history can be fickle. Mutations to the underlying elasticities create a misleading database.

So they have been in a quandary as to whether or not to raise interest rates because the system they had thought they understood suddenly stopped working in its usual way.

Stimulus was applied in some great multiple of any previous stimulus, but the reaction has been weak and late to materialize.

The source of the uncertainty is the new context of the economy as it labors to produce positive results within an open global system of not just foreign goods competing with domestic goods but also with the wild card of foreign capital flows that are capable of gushing in or out of a country and offsetting or magnifying Federal Reserve changes in available market funding.

Moreover, exchange rates have become flexible and no longer fixed by governments, and there is no insulation from foreign capital inflows or outflows. So when the US dollar strengthens, as it has, the products of US multi-nationals get priced out of foreign markets. Further, the changing exchange rates drive capital in and out of countries, which accounts for far more financial buying power than the Federal Reserve would dare employ.

To make matters more difficult, foreign central banks are motivated to protect their own end of the global bargain and offset Fed actions that are self-serving to the US. Furthermore, the US banking system has not responded to the availability of cash reserves remotely near past responses.

One should understand the complications are not just determining the path of the “moon” that our rocket is chasing but also the responsiveness of the rocket to the Fed’s control tower.

So the question is, is policymaking today an art or a science given that the models they have to go on fail to capture the essence of today’s elasticities?

This idea causes me to harken back to an experience I had in the l960s when I was an economist at the Federal Reserve. At the time, economists (myself included) were agog over the science of being able to model the responses to policy.

To do so, I constructed a model of the economy on an analog computer, which was ideal for tracking the interactions of the financial and economic variables over time. There was a lot of experimentation to set the elasticities so that the resulting system reactions followed patterns that were similar to the then-current economy.

To give the experiment some life, I set about simulating what the Fed’s Chairman at the time described to be his method of stabilizing the economy. He called it “leaning against the wind.” While highly suggestive, it needed to be fleshed out. I came up with a few versions of practical implementations of what could arguably be a policy rule for “leaning against the wind.”

I presented the methodology and results to a staff economist colloquium that also included most of the Governors who sit on the Federal Open Market Committee.

The response was dramatic among those tasked with deriving a mathematical response to policy stimulus. Alas, the paper and news of the presentation reached the Chairman whose considered response was that someone should “tell Spellman that monetary policy is an art, not a science.”

Over the following decades, the scientific approach to modeling has attempted to be more precise, but during those same decades, the underlying system mutated and there is insufficient historical data that’s relevant to today’s mutated economy.

So now we have come full circle. The Fed, in making its historic interest rate decision, is left only with art and models built from less relevant data, which for them is an uncomfortable place to be as the policymakers these days are trained scientists, not artists.

Let’s hope they have the courage to move on even though they lack the scientific proof that, indeed, raising interest rates would push the economy toward absorbing the last of the unemployed while not trigging incipient inflation, nor which would send the economy back into the Great Recession.

At this point in life, decades later, I’ve come to agree with the Chairman of bygone days. Policymaking is an art, not a science. This is not to say they shouldn’t lean on what science would suggest.

This is not a comfortable place to be when having the responsibility for the outcomes ahead, so I do watch for nervous twitches as they testify before Congress and global financial opinion.



The views expressed are those of the author and not necessarily The University of Texas at Austin.

About The Author

Lew Spellman

Professor, Department of Finance, McCombs School of Business, The University of Texas at Austin

Lewis Spellman received his BBA and MBA from the University of Michigan and his MA and Ph.D. from Stanford University. His research interests include...

Leave a comment

We want to hear from you! To keep discussions on-topic and constructive, comments are moderated for relevance and for abusive or profane language.
Login or register to post comments