Stanford University economics professor John B. Taylor is perhaps best known as the creator of the Taylor rule, which seeks to determine the Federal Reserve’s interest-rate target according to a simple formula. He has also become an outspoken critic of both fiscal and monetary policy as practiced over recent years.
Taylor writes a lively blog called Economics One, and, early this year, published the book First Principles: Five Keys to Restoring America’s Prosperity (W. W. Norton & Co., 2012). Taylor recently sat down with us for an interview at our offices in New York. Here are excerpts of that interview.
IS THE STIMULUS DEFENSIBLE?
Gene Epstein, Barron’s: What’s the best case you can make on behalf of those who defend the recent fiscal stimulus?
John B. Taylor: The case that has been made for the discretionary fiscal stimulus is based on quite conventional Keynesian theory. It is basically that if the government gives people a lot of extra money on a one-time basis, they will spend it. Not all of it, but most of it. Similarly, when the federal government gives money to states and localities as part of the temporary fiscal stimulus, it will be spent in such a way as to boost gross domestic product. And that will greatly help when economic activity is otherwise either contracting or stagnant.
My own view is that the theory is flawed, and the evidence that the fiscal stimulus achieved the desired result is practically nonexistent. The surge in federal spending only increased the burden of the already burdensome federal debt.
Epstein: Start with the evidence.
Taylor: The attempt to stimulate consumer spending in 2009, or the earlier attempt under President Bush in ’08, showed the expected rise in consumer income as government payments were made, but little or no response from consumer spending. Inconveniently for the advocates, consumer spending actually declined in some of the calendar quarters when it was supposed to have been stimulated. If you use statistical analysis to take into account the factors that would have brought increases or decreases in consumer spending, you find virtually no boost to spending from the stimulus.
As for the money sent to states and localities, economist John Cogan and I found that the funds were either put into financial assets or used to reduce borrowing. The hoped-for increase in infrastructure spending was negligible.
What is also inconvenient for the advocates: according to the national income and product accounts, state- and local-government purchases were lower every quarter in 2009 and 2010 than in 2008.
Epstein: And you would expect these results from the standpoint of economic theory?
Taylor: Let’s start with consumer spending. It’s basic economic theory that most people look beyond the very short term. To expect them to rush out and consume more when the government cuts them an extra check on a temporary basis is not realistic. Instead, they will bank most of the extra money or use most of it to pay down debt. There are exceptions, of course. Some people will feel so pinched, they will need to spend the money. But the data show that the exceptions don’t dominate the story.
Epstein: It has been argued, however, that the Reagan tax cuts of 1981 were an example of effective fiscal stimulus.
Taylor: The Reagan tax cuts were not temporary or targeted, which are defining characteristics of discretionary fiscal stimulus packages. The Reagan tax package was meant to be permanent, and a permanent tax cut can certainly influence consumer behavior for the same reason that a temporary tax cut does not. Consumers can rely on the extra income and might boost spending accordingly. Businesses might then respond by expanding capacity, because they feel they can rely on the extra sales. And the permanently lower tax rates will encourage more hiring and investment.
But let’s even imagine that with temporary tax cuts, consumers do spend all the extra money at the malls. Can we expect business to respond by hiring more workers on a lasting basis? Of course not. Businesspeople will know better than anyone that the pop in spending won’t last.
Epstein: The Obama administration’s “cash for clunkers” program did seem to have the desired effect, however. Consumers in possession of a qualified clunker could trade it in for an inflated rebate, and then apply that government-funded rebate to the purchase of a new car. And yet that was a temporary stimulus.
Taylor: It certainly was, and it taught us the unsurprising lesson that if you offer consumers a financial incentive to buy a new car, people already planning to do so will take advantage of that incentive and make the purchase sooner rather than later.
It also taught us something else about the problem with temporary-stimulus programs: even when they work, there is payback shortly afterward. Cash-for-clunkers caused a pop in new-car sales for only a month or two, soon followed by a downside correction.
Epstein: Let’s return to your point about the states failing to spend the fiscal stimulus money on infrastructure. One theory in support of such policies is that they depend on competent people implementing them, and in this case, we just didn’t have enough competent people. So more competence will make that kind of fiscal stimulus work.
Taylor: But the competence you are assuming is based on the unrealistic idea that we have a completely different federal-state political system. For example, some observers think China’s government was recently successful in boosting infrastructure spending to spur economic activity. We don’t know for sure, but let’s assume they’re right. There is no lesson in that for the United States because we can’t choose China’s system—and wouldn’t want to even if we could.
For one thing, local governments in China have no access to capital markets. They basically take what they get from the central government, and spend what they get. And since they always have a list of projects on the shelf, all the central government need do is lower the hurdle for what is a good project, and suddenly the project gets implemented.
By contrast, we have a decentralized system in which state and local governments do have access to borrowed funds. That is what you have to take largely as a given.
Epstein: You are saying we don’t really have shovel-ready projects that can be quickly implemented by government?
Taylor: Back in the 1970s, the Carter administration made stimulus grants to the states, and that did not work, either. There are some who say “just use the military—they can always spend the money somewhere.” But how useful, or politically realistic, is that? It would probably be worthwhile to have more information technology in medicine, and better record keeping. Part of the stimulus went to that. But how could that possibly stimulate the economic recovery from recession, when it should all be planned and implemented carefully over time?
Government officials imagined they had a little textbook model. You boost the “G” that stands for government spending, and voilà, it works. But that is not the reality.
Epstein: But if, as you say, consumers and local government used the money to reduce borrowing and pay down debt, could that be a completely bad thing?
Taylor: It can’t be bad to reduce the debt of consumers and of state and local government. But it can’t be good to boost federal debt in the process by the same amount. So that’s a wash. It’s really hard for me to see it as a net positive.
Epstein: Is there any part of the huge increase in the federal deficit during this period that you do not oppose?
Taylor: Absolutely—the part that is normally referred to as “automatic stabilizers.” When the economy contracts or barely grows, certain fiscal responses happen automatically. Revenues from taxes decline as profits, capital gains, and wages and salaries decline or barely grow. Spending also increases—on welfare programs, unemployment insurance, and when some people retire earlier than planned and claim Social Security payouts. So you automatically get a decline in revenue and increase in spending.
I emphasize the “automatic” quality of these events. They are predictable, and they occur in a timely fashion. What is unpredictable and not timely is the futile attempt to engage in proactive, discretionary fiscal policy.
Epstein: Right now, the Federal Reserve’s official interest-rate target on federal funds is zero percent to 0.25 percent. Is that where the target should be, according to the Taylor rule?
Taylor: No, the interest-rate target would be at 1 percent, or a little higher. That is based on starting with a multiple of the inflation rate, and then adjusting up or down for the growth rate of the economy compared with its potential growth rate. Right now, the growth-rate part of the formula would call for a downward adjustment, so you end up around 1 percent.
The Taylor rule is based on empirical research about what tends to work best for the short-term interest rate. It is far from perfect. It just works better over the long run than the discretionary tinkering of imperfect human beings. And in fact, through most of the period from the early 1980s through 2002, when the economy performed relatively well, the central bank’s interest-rate target tracked the Taylor rule fairly closely.
Epstein: But not since then. You have said that during the period from 2003 through 2005, the Fed-funds rate violated the Taylor rule by being too low.
Taylor: Yes, at times by as much as 3 full percentage points. And that helped cause the financial crisis, as I have argued in my book Getting Off Track.
Epstein: Federal Reserve Chairman Ben Bernanke, who was a Fed governor under Chairman Alan Greenspan when this policy was implemented, has responded publicly to your criticism. What would you say is his best argument?
Taylor: His argument has been that at the time, they were forecasting a slowdown in the rate of inflation. And that if you used this forecast, the target they maintained would have been consistent with the Taylor rule.
Epstein: And what’s your response?
Taylor: My main response is that the Taylor rule does not, and should not, depend on forecasts. Forecasts are subjective—and therefore too discretionary. Based on what was known at the time about inflation and economic growth, the target was far too low.
Epstein: Bernanke once claimed that economists generally agree that the Fed’s policies in 2003 and ’04 did not contribute to the housing bubble.
Taylor: Yes, but Bernanke mainly cited himself or the Fed board staff, and did not cite others who concluded just the opposite: that the Fed’s policies did help fuel the housing bubble. More recently, Bernanke has retreated from that position. He now says there is no consensus among economists about the Fed’s role. I regard that as progress.
TOO MUCH EASING?
Epstein: On another important component of Fed policy, what do you think of quantitative easing?
Taylor: I opposed the large-scale asset purchases of QE1 and QE2, and believe it’s unfortunate that the central bank is still publicly considering a QE3. They were ineffective and potentially harmful.
These massive purchases of mortgages and medium-term Treasuries were aimed at lifting the value of these fixed-income securities, and thereby bringing down the relevant interest rates. At best, that was the short-term effect. But how long can such an effect last? What basically determines these interest rates are expectations about future interest rates, which in turn are partly determined by inflationary expectations.
Epstein: You don’t believe purchases by the Fed have any long-lasting influence?
Taylor: Let’s suppose something even more misguided: for QE3, the Fed decides to buy the stock of publicly traded companies in order to lift stock prices. Equity prices would rise. But how long could that last, unless the earnings of these companies rise proportionately? Price-earnings ratios would become unsustainably high, and the market would soon correct for the Fed’s aberrational influence. The same dynamics work for the bond market.
Epstein: Apart from not helping, quantitative easing has also hurt?
Taylor: Speaking of the stock market, notice that stock prices have become sensitive to whether or not the Fed will implement QE3. I see no reliable relationship between equity prices and quantitative easing, but the fact that many people believe in it is at least a source of concern. Another important concern is just how the Fed will manage to sell off all the mortgage-backed securities and medium-term government bonds it now holds on its balance sheet without causing disruptions in the market.
But my real worry is that quantitative easing may become a pillar of Fed policy. If the economy speeds up, you do less of QE; if it slows down, you do more. Quantitative easing may become not just the wave of the present but of the future, which could be very damaging.
Epstein: What’s the basic lesson for both fiscal and monetary policy?
The “Do Something” Fallacy
Hoover senior fellow John B. Taylor was awarded the Manhattan Institute’s 2012 Hayek Prize for his new book, First Principles: Five Keys to Restoring America’s Prosperity (Norton 2012). An excerpt from his Hayek Prize lecture:
The great economist and philosopher Friedrich A. Hayek argued that the case for rules-based policy goes beyond economics and should appeal to all those concerned about assaults on freedom. He wrote in his classic 1944 book, The Road to Serfdom, that “nothing distinguishes more clearly conditions in a free country from those in a country under arbitrary government than the observance in the former of the great principles known as the Rule of Law.”
Rules-based policies make the economy work better by providing a predictable policy framework within which consumers and businesses make decisions. But they also protect freedom, a concept Hayek developed in his 1960 book, The Constitution of Liberty. A rules-based system in fact has a dual purpose: freedom and prosperity.
Skeptics ask how a system of policy rules can work when politicians and government officials want to “do something” to help the economy or feel public pressure to do so. A rules-based system with less discretion sounds good in theory, they say, but rules mean you do nothing, and that is impossible in today’s charged political climate and twenty-four-hour news cycle.
Hayek had an answer. In The Road to Serfdom he wrote that it was wrong to say that the “characteristic attitude [of a rules-based system] is inaction of the state.” In our day and age, having rules for monetary policy does not mean that the central bank does not change the instruments of policy (interest rates or the money supply) in response to events, or provide loans in the case of a bank run. Rather it means such actions are taken in a predictable manner.
In the years before the 2008 financial crisis, monetary policy deviated from the more predictable rules-based policy that had worked in the 1980s and ’90s—that is, the Federal Reserve held rates too low for too long. Moreover, government regulators failed to enforce rules on risk-taking at banks and other financial institutions, including Fannie Mae and Freddie Mac. Then came the discretionary stimulus packages and exploding debt, the regulatory unpredictability associated with ObamaCare and Dodd-Frank, and the unprecedented quantitative easing through which the Federal Reserve bought 77 percent of new federal debt in 2011. The U.S. tax code has become particularly unpredictable, and now the epitome of unpredictable policy is upon us: a self-inflicted “fiscal cliff” where virtually the entire tax code will be up for grabs.
Some will claim, of course, that crises force policy makers to deviate from predictable rules. One can argue that bailouts and other discretionary interventions were needed during the panic of the fall of 2008, and perhaps they prevented a more serious panic. But that is like saying that the person who set fire to your house should be exonerated because he helped put out the fire and saved a few rooms.
Taylor: Both policies should be predictable, not discretionary and unpredictable. It is always possible that some brilliant policy maker can do something unpredictable that will yield a better outcome. But history has shown us repeatedly that this usually yields worse outcomes. My new book, First Principles, names “policy predictability” as one of five key principles.
Epstein: And the four others?
Taylor: Rule of law, strong incentives, reliance on markets, and a clearly limited role for government.
Epstein: Thank you, John.