What Do Economists Know?

The title of my talk today was stimulated by a conversation I had forty years ago with Peter Bauer, the distinguished Cambridge University economist. I arrived early at a dinner for him, and before other guests arrived he proclaimed, provocatively, that the number of things that economists knew that were true, important, and not obvious, was no more than the fingers on one hand. I waited to hear what the four or five things were that were true, important, and not obvious, but other guests arrived, the conversation was interrupted, and I was left forever in suspense.

I reflected on the question from time to time. I couldn’t be sure whether he meant there were only five things altogether that together we know, or there are many important things known but no one among us knows more than five of them. So my program was just to take inventory of how many things I knew in economics that were true, important, and not obvious, and to see whether they added up to five.

Of course, to say that I know five things is to imply that they must be true: I can’t be said to know them if they are false. But I could believe them wrongly to be true. Peter Bauer may have meant that there are many more things that economists think they know-they just aren’t true. Believing something makes it true to me but not to somebody of a more skeptical persuasion.

This excursion on truth is not just a play on words: some important things are true in a different way from the way some other important things are true. Similarly some things are obvious in a different way from other things; in fact some things are believed true precisely because they become obvious. I say “become obvious,” because my candidates for Peter Bauer’s collection all have the characteristic that, while at first glance they are paradoxical, once understood they are seen as incapable of being false.

These are what are sometimes called accounting identities. When I was an undergraduate they were often disparaged as “mere identities.” They were unfalsifiable statements and did not count as scientific truths. They were said to be true by definition.

Actually the truth of all scientific propositions depends on careful definition; but the truth of the so-called identities depends only on careful definition. They are not merely definitions; if they were, they would be obvious. But they can be derived from definitions if the definitions are carefully made consistent.

The simplest possible identity-it sounds obvious when I say it-is that in any sales transaction the value of the item sold equals the value of the item purchased. The need for careful definition arises even here: if there is a sales tax we have to treat it as either received by the seller or as a side payment to the tax authorities. That even this simple identity is not always obvious was apparent a few weeks ago when values on the stock exchanges dropped almost ten percent. I heard several intelligent-sounding conversations on national public radio about where all the money was going. People obviously were anxiously liquidating their port-folios and it seemed important to know what they were doing with the proceeds.

What apparently wasn’t quite obvious was that no money was leaving the stock market. No money could leave the stock market. No one can sell a share of stock without a buyer: taxes and brokers’ fees aside, for every dollar liquidated there has to be a dollar invested. Individuals, yes, were taking money out of stocks; but all investors together could not. Where the liquidated money was going might be interesting; but then equally interesting was a question not asked: where is all the money coming from? Together they had to cancel out. We can’t all get rid of our Canadian quarters by passing them along at the first opportunity.

The national-income accounts contain a number of important identities. What they are is quadruple-entry consolidated income statements. Double-entry bookkeeping is for a single individual: but with every sale of goods or services there are two entries for the seller, two for the buyer. These accounts help us to understand the futility of efforts to get consumers to save more by spending less. The United States has a private savings rate lower than it used to be, lower than other industrial states. We citizens don’t accomplish enough saving to promote faster improvements in productivity. It is seriously proposed that if we didn’t use our credit cards to buy so many needless consumer goods we could have a more respectable rate of saving.

What apparently is not obvious is that the only way private citizens can accumulate savings is to consume less than their incomes, that is, to earn more than they spend on consumption. But there is no way they can all earn more than they consume except by producing equivalent goods that are not consumed. Somebody has to buy those goods. They can be domestic investment, exports, or government purchases; if those do not increase, it is impossible for private saving to increase. I can save another $10 by going without a haircut; but my barber’s income goes down by $10, and his savings too, unless he cuts his own spending, in which case he passes along the loss in a chain reaction that offsets my $10 saving.

Because you graduates have learned this truth it may seem obvious; but it wasn’t obvious to you until you learned it, it isn’t obvious to most people out there, and it wasn’t obvious to economists when I entered Berkeley.

There are other important accounting truths. Most bankers understand their own balance sheets well; what most bankers don’t know is that whenever they make a loan they increase the money supply. They don’t know it because, as they see it, they only lend money that already exists, money in the bank’s reserves. What they don’t see is that the reserves still exist-they have simply migrated to other banks -while the customer’s money also still exists; it, too, has migrated.

Most of the non-obvious accounting propositions are true only in the aggregate, not true for the individual. They do not correspond to the experience of the banker who lends only money that he has in reserve, or the consumer who foregoes the haircut. But they hold in the same way that the laws of conservation do—conservation of energy, mass, or momentum. If we launch a squash court into orbit, the momentum of its center of gravity will be undisturbed by the game being played inside. In the physical sciences many of these accounting identities are dignified with the title, “law:” and like the identities in economics they were not obvious without an intellectual struggle. The conservation of momentum—that a moving object will continue moving at the same speed and in the same direction forever until resisted, pushed, deflected, or subjected to friction—has become obvious, but because it couldn’t be observed anywhere in the universe until space flight became possible, its becoming obvious required laboratory experiment and intellectual struggle.

It is sometimes said, in textbooks and in learned volumes, that these accounting statements, being unfalsifiable, do not count as science. I don’t care. The question is whether they tell you something important you didn’t know. The history of our discipline demonstrates that they are not obvious. Disparaging them as “mere identities” at least testifies to their truth. They are the foundation of any macroeconomics. They have their counterparts in physics, chemistry, biology, genetics, and our sister discipline, demography. They are sometimes known as “budgets;” there is the earth’s energy budget, its carbon budget, its water budget, even budgets for non-degradable substances like DDT.

There are many more than five important accounting identities in economics that are not obvious; I could have mentioned the balance of payments, or the input-output matrix. As to their importance, where I would draw the line-at five or ten-I don’t know. Some of them were surely unknown to Peter Bauer; the carbon budget was unrecognized when I had dinner with him.

So if I am allowed five candidates, they are all accounting identities. I cannot hope to have persuaded you in twenty minutes, but I hope I have sensitized you to be on watch for them, to respect and appreciate them, and not to be afraid of counting them among the things you know that are true and important and were not obvious before you studied economics, even though they do not count as empirically grounded, potentially falsifiable, scientific hypotheses.

I am tempted to close these remarks with a small investigation into what it is that economists know that are important, not obvious, and not true. Actually, what I have in mind is propositions or principles that are true but that unintentionally, the way they are formulated, appear to deny something that is also true, raising the question: which deserves the more emphasis, the truth contained in the proposition, or the truth the proposition appears to deny?

I have a candidate. It has become fashionable in the last two decades, not only among economists but among those who like to quote economics, to advert to an incontestable, absolute truth colloquially expressed as: there is no free lunch.

The truth that I think this assertion is intended to communicate is that resources are always scarce, there are competing ends and competing beneficiaries, redistributing in someone’s favor is at someone’s expense and there is no alchemy in economics: you can recycle, but it is hard to find the equivalent of a nuclear breeder reactor that produces, in burning fuel, more fuel than it bums.

Maybe it’s because of where I’ve been in economics, but I prefer the alternative truth, that there are free lunches all over just waiting to be discovered or created. What I have in mind is what we technically call Pareto improvements, or the gains from trade. There are non-zero sum games that permeate the economy that have settled into, or have been forced into, inefficient equilibria.

There are not just free lunches but banquets awaiting the former socialist countries that can institute enforceable contract, copyrights, and patents, or eliminate rent-free housing and energy subsidies. How the lunches get distributed matters; but the lunches are there.

Those of you who move into the economics profession, in government, academia, or in business, will spend much of your time exploring for opportunities to eliminate constraints on mutually beneficial trading, to overcome market failures and to create markets where they are needed, to identify removable deadweight losses, and to promote integrative bargaining.

This is what we do in economics. Technological innovation can push out the production frontier; it is economists who help to find where we are deep inside that frontier, diagnose what keeps us from the frontier, and propose institutional changes to bring us closer to the frontier. To those of you who become professional economists I urge you: get out there and help find those free lunches.