«Like handing a loaded gun to politicians»

By saying deficits don't matter, Modern Monetary Theory might lead to unwise political decisions, says economist Miles Kimball, professor at University of Colorado Boulder.

Professor Kimball, what lesson can we draw from the pandemic?
Despite being a macroeconomist, I have found the very unusual macroeconomic situation much less interesting than the public health response to the pandemic. I think it would have been beneficial if economists were given a greater role on the public health side. Economists are quite well equipped to understand issues in public health, including epidemiology models. In addition to bringing economists into the policy-making on the public health side of the next pandemic, I’d love to see economists get involved in research on diet and health. As an economist, I have gotten some flak for blogging about diet and health every week. But economists actually have the key skills needed to understand issues in diet and health: economists are extremely well-trained in statistics, and well-trained to think about complicated systems. I haven’t found it hard to read papers about nutrition, whether they are based on experiments or on what different people do in their regular lives.

Are economists so much better than other disciplines?
No, I welcome scientists from other disciplines to do research on economic topics and have a debate with economists. Let me say it this way: Any important scientific or policy question needs to have at least two scientific disciplines looking at it. Never trust an important question to just one scientific discipline. It is a lot safer to get multiple perspectives from different disciplines with different scientific incentives, methods, assumptions, biases, and blind spots. Giving each discipline a monopoly over a set of issues that it owns is asking for trouble.

What would economists have done differently regarding the public health response?
It was clear for economists from the beginning that speed was of the essence. Every month you are in the pandemic costs a fortune. I think a lot of countries regret now that they did not spend billions of dollars on speed—for example, by paying to be in front of the queue for getting the vaccines. A simple cost-benefit analysis would also suggest that in a pandemic we should worry less about side effects. We were underspending on public health, while now governments must spend enormous sums on the economic fallout. You could argue that we almost arrived at the worst of all worlds.

What do you mean?
There were two polar opposite options at the beginning of the pandemic: do a very strict lockdown for three weeks to ensure that there are no infections—the eventual China strategy—or just continue to live as before—by which I mean more or less what was done during the 1918 flu pandemic. The first option would have resulted in limited economic damage and few deaths. Following the second option, we would have saved fewer lives, but we would have had relatively little economic damage. We chose a middle way which probably saved lives relative to the second option, but resulted in a huge economic fallout.

Are there other examples of ideas by economists?
Yes, the Nobel Prize laureate Paul Romer argued early for mass testing to catch Covid-19 before any symptoms arise and isolate those who are infected. Suppose for example that everyone in the country had been tested twice a week. That would have done a lot to take away Covid-19’s invisibility cloak early in an infection. And economists understood quite well, maybe better than representatives of other disciplines, that a cheap, fast, unreliable test is much more useful than an expensive, slow, accurate test.

Economists are often attacked for trying to quantify the value of everything, including lives, which many see as immoral.
If you give something a monetary value, in most cases you make it more important, not less. There is a research paper using cost-benefit principles to assess the value of the damage done worldwide by domestic violence in the tens of trillions of dollars. Numbers like that make people sit up and take notice. Putting a dollar value on a problem makes it seem like a problem practical leaders should worry about, not just bleeding hearts. Also (ALSN 280.00 -0.18%), economists will tend to give people more choice, arguing for gentler approaches. For example, to deal with climate change, most economists favor a carbon tax rather than prohibitions.

This sounds like you do not see any faults in economics.
We could do much better. I see too much pressure on economists to publish in top journals instead of following their intrinsic motivation. The external motivations should be dialed down, the intrinsic motivations such as intellectual curiosity or the desire to have a positive impact on the world dialed up. The result of having external motivations dialed up too high is that economists follow fads and there is too little diversity of approaches.

Such as assuming that everybody acts as a homo economicus, i.e., completely rational?
Part of what I mean by “too little diversity of approaches” is that when simplifications are necessary, economists follow the same list of allowed simplifications rather than exploring what can be learned from different simplifications. Assuming that people are infinitely intelligent is a very useful simplification in economics, because there are a hundred ways for people to mess up, but only one way to do things right. But why do so many economists act as if we always have to use that simplification of infinite intelligence handed down by tradition? I don’t mean to say that makes things easy. An alternative to assuming people are infinitely intelligent is to use agent-based models, which pretend the people are much stupider than they really are instead of the usual assumption in economics that people are much smarter than they really are. Knowing the implications both of assuming people are stupider than reality or assuming they are smarter than reality is better than knowing only what happens if they are smarter, but what we really need are models that better match the reality that human intelligence is somewhere between those extremes. I have written about how hard that is in my paper “Cognitive Economics.”

Are there other instances where this is true?
One example is Quantitative Easing, the purchase of long-term or risky assets by central banks. In the simplest traditional model, QE is neutral: there is no important effect because it is only a swap of two assets that investors frictionlessly accommodate, responding to the new situation. Macroeconomists are definitely interested in writing down models in which QE has an effect. But there are many ways to write down a model in which QE has an effect. There is too little curiosity about which of all the possible mechanisms are really in play. The focus on coming up with some possible mechanism. This matters because we know that QE at the dosages we actually used was not enough—we experienced a prolonged recession after the financial crisis. So what we really need to know is what would happen if we did three or four times as much QE as we did last time. To make a reasonable guess about the side effects that would be caused by such a large dose of QE, we need to know which of all the possible mechanisms that could make QE do something rather than nothing is in play.

One heterodox way of looking on macroeconomics is Modern Monetary Theory, MMT. Their proponents argue that we should not worry about the fiscal deficit. Do you agree?
I think they are right that we should be more agnostic about the effects of deficits—and more fundamentally, the effects of the debt-to-GDP ratio. But overall, I think the true parts of MMT are standard economic theory. What they do is to jump from the scientifically correct statement that we should be more agnostic about just how dangerous more debt is to saying we shouldn’t worry about it until the harm is obvious. There are a lot of things we don’t know for sure, simply because—from a statistical point of view—there isn’t that much data in the macroeconomic time series. But my own guess based on what little evidence we have and from the theory that I share with the MMT folks is very different from their guess that we can spend with abandon and it will all be OK. Even if we don’t know exactly what the limits are, I think there are limits to the safe level of spending, and I am very concerned about using up our safe government spending capacity on anything less than the most important projects.

What MMT proponents often say is that inflation should be regarded as the critical factor.
Yes, if we see inflation going up strongly, we should worry. The problem is that inflation could build up over a long time. Think back to the 1960s when inflation was slowly rising. It took a long time for inflation to gain momentum, but it did. It is like a supertanker that is slow to turn. Inflation rising slowly now doesn’t mean it couldn’t rise faster later if we acted as if there were no limits to spending. Also, expectations matter. Currently, US inflation expectations are still anchored around the Fed’s target of 2 per cent. But they could become unanchored.

So, you would argue for caution?
MMT says it is speculative to say there is a problem before bad consequences are on our door step. That is scientifically true, and it is right to point to our ignorance. But it isn’t a good idea to fill in our ignorance with wishful thinking. Saying to politicians “Don’t worry about debt or deficits” is like handing them a loaded gun. Loaded guns can be used wisely, but often aren’t. I’d rather steer clear of danger.

There is now a lot of discussion regarding Central Bank Digital Currencies, CBDCs. Would you welcome such digital currencies?
In discussions of central bank digital currencies, there is a lot of focus on private households. I also fear that digital currencies will be modeled as a replacement for cash rather than as an account with the central bank. Central banks could take a gradual approach: they could begin by allowing companies to do transactions using a central bank account. In the US, to make sure those accounts aren’t too cash-like, I’d like to see them built as extensions to the Fed’s reverse-repo overnight facility.

What is wrong with having central bank digital currencies by cash-like?
Digital currencies should allow for negative interest rates. Any central bank that looks ahead should not commit the mistake of modeling CBDCs after paper money that does not allow for negative rates. If there is a zero-lower bound, i.e., the interest rate cannot fall far below zero, you might end up in a situation where you could lend easily and safely to the government with zero interest rates – which might in certain economic situations be too high. People would lend to the government instead of investing, doing research and development, or hiring. It could crash the economy.

Do you see negative rates as an option for the Fed?
Fed officials still resist it. Negative rates are not even part of the monetary framework yet. And there might be legal limitations to implementing negative rates on reserve accounts. (I am currently coauthoring a paper on negative interest rate law.) Fortunately, those legal limitations don’t apply to the overnight facility (which is based on Treasury-bill repurchase agreements). As it is optional to use that facility, the Fed is free to implement negative rates there. You could allow regular companies to use the facility for ordinary transactions. Limiting the amount of reserves a bank is allowed to hold at the Fed would then be enough to make the overnight facility the central lynchpin of the whole financial system.