Like This? Please link to /politics/economics/2020/05/10/cassandra.html.
(An abbreviated version of this piece appeated in The International Economy.)
Could we suffer a fate as bad as or worse than that of Japan—low inflation, wage growth, productivity? I am afraid it could be a lot worse. I am afraid that we could lose the essential trust our economy needs to function. Like Cassandra, my own prophesies are unlikely to be believed. Unlike Cassandra, I am not blessed with perfect foresight. Unlike Paul Krugman, I don’t pretend to know the future. I hope to be proven wrong—as a zombie in Krugman’s classification.
There is a popular belief that the Fed is omnipotent. However, this is not so. A central bank is (primarily) merely the guardian of the means of exchange and of the trust in it. When working together with the government (and sometimes against it), this trust can extend beyond just currency to the overall economic system. If people were hyperrational, they could undo much of what their central authorities do and thereby diminish their powers. But people are not so rational and this has given the authorities effective power to intervene.
This magic intervention power is principally one of trust. We are indeed fortunate that the world has so much trust in the power of our Fed that it has not been panicking. This has prevented a self-fulfilling prophesy, in which all of us believe in the breakdown of our economy, retreat to preserve ourselves, and as a result suffer exactly such a breakdown.
Trust has a dark side, too. It is easier to lose than to restore. Money, too, is all about a self-fulfilling prophesy of trust. It is not true that it is difficult to fan inflation. It is just difficult to fan inflation with conventional central bank tools. But tinker with trust through capricious economic policies and inflation will surely follow—in the extreme, destroying money as a viable medium of exchange.
My third self-fulfilling prophesy has been tested out many times throughout history. Credit is based on the confidence that a debtor will repay. If creditors believe that they will be repaid on good terms, default premiums can be low and debtors will be able to repay. But if creditors believe that they may not be repaid, the interest rate may rise to the point where debtors can no longer afford to repay.
In short, our economy is built on trust. Trust in the economic system, in the government, in the central bank, in the medium of exchange, in the ability of the government to repay its obligations on the expected and promised terms, in the future. This trust is maintained in an equilibrium. Without it, modern economies collapse.
The central authorities have a second important function, that of redistribution. They can confiscate wealth from some and hand it to others. Redistribution is typically less obvious when executed by central banks rather than through the tax system, and the instruments are more blunt.
What can governments and central banks not do? There is a simple “summing-up constraint.” Collectively, we can only consume what we produce. If we do not make stuff that makes us live better, we cannot live better. Governments rarely invent the kinds of new technologies that raise our standard of living and make our lives better. By and large, they are not good at inventing mobile devices, rockets, electric cars, solar power, fusion power, vaccines, cancer cures, longevity drugs, better crops, and so on. Of course, they can try to redistribute towards activities that are more likely to increase our collective pie—especially when there are public externalities that limit private companies. Unfortunately, real rather than ideal governments tend to be more prone to distribute funds to friends, families, tribes, political donors, and specific voters, than they are to facilitate activities that are likely to make our pie larger. And determining the best redistribution would be very difficult even for the most benevolent unconflicted dictator. Of course, real governments can do good, too. Without any education and research sponsorship, we would be worse off. But the overall government track record is spotty; and even where it does good, it does so far less efficiently than it should. It should not cost $10 million to build one mile of highway. It should not cost $20,000 per student per year in the LAUSD to deliver demonstrably mediocre education.
The Japanese problem is not one of misguided monetary policy or deflation. It is much more basic. Japan has failed to create the great technological and social inventions that would have made its next generation better off than the last.
Our problem is different. We have been pulling the most talented minds from all of over the world to our U.S. universities (although this is now under threat). We have excelled in inventing the future. Inefficient, yes. But better than anyone else has.
I think Our problem is that we are collectively heading for far worse than the Japanese disease. I am worried about loss of trust. The most urgent one concerns the trust of our creditors. We need our creditors to remain comfortable. Right now, they are so confident that they charge us almost no interest. Can we be confident that our creditors will remain confident?
How realistic is it for our creditors to expect to be repaid on the terms they expect? The following is a good steady-state sketch. (The treatment of the future is just an expository simplications. The argument is robust.)
In 2019, our standing debt was about $80k for each American, about the same as annual GDP. The average American earned $56k/y, paid $17k/y in taxes, and borrowed a deficit of $3k/y. By the end of 2020, our debt will likely reach around $100k per person, to be paid for by the steady-state “good-times” income tax receipts of about $17k/y.
Well, no! Assuming that social security, medicare, and debt repayment are untouchable, we only have about $10k/y to repay debt and interest. If defense is untouchable, too, only about $7k/y. If the baby boomers were already retired now (as they are about to be), only about $2-4k/y. These discretionary tax receipts also have to cover education, welfare, research, etc., plus more interest on more debt. In the best of years, the debt-to-GDP ratio increased by about 3% per year, doubling about every 25 years.
How long can this go on? I believe that our creditors’ confidence today implies that they believe that the US will either enjoy long-term GDP growth in excess of 5%/y, and/or raise income taxes by 20%, and/or cut all discretionary government spending—or that the government will not make good on their promise to them (e.g., by inflating their debt away).
Unless we can stabilize our debt-to-GDP growth soon, our creditors can best be described as willing participants in a Ponzi scheme. Trust can maintain a Ponzi scheme for a while—but not forever. This worries me. But what really worries me is that trust is a self-fulfilling prophesy—more likely to flip suddenly than to erode gradually. In fact, it need not flip in 20 years. It could flip tomorrow!
How much time do we have? What could happen if creditors lost trust in repayment? What could happen if investors lost trust in the power of the Fed? What could happen if the dollar lost trust? What could happen if the public lost trust in the economy? And who will we be able to rely on to restore trust?