Libertarian Monetary Policy, Part Deux

So far we have two rules for monetary policy:

1. Don’t screw up the medium of exchange and unit of account functions of money, and

2. Shifts in money demand should be met by equal shifts in money supply so as not to create a disruptive change in the value of money itself.

In this post, I want to further explain why rule 2 is necessary. To do so, I have to explain a bit of monetary theory. It’s not hard, and once you understand it, you’ll be ahead of 90 percent of the people who write about money and the economy in the popular press, including at least one Nobel Prize winner with a column in the New York Times.

Well over 200 years ago, the French economist Jean-Baptiste Say popularized what has become known as Say’s Law, which says, in it’s simplest form, that demand arises from production. Suppose you lived in a simple barter economy. You exchange your labor and/or goods you produce in exchange for goods and services produced by other people. If you don’t produce anything of value to others and your labor is not of value to someone, you have nothing to exchange, and so your demand for the other people’s production is zero. If you do produce stuff, you either consume it yourself or exchange it for something someone else made. This is true for everyone, so there is no possibility that there will ever be insufficient demand for goods in the economy, because everything produced is consumed by somebody, and nobody can consume what hasn’t been produced.

Notice that I haven’t said anything here about the rates at which one good will be exchanged for another. If one person produces a large amount of some good, it’s quite likely that the resulting relative abundance of that good will lessen it’s unit value in exchange. If there’s a bumper crop of apples, an apple grower will probably have to offer more apples to get an orange than he would if apples were relatively scarce. All Say’s Law says is that he will, in fact, trade or consume all of his apples. It doesn’t say anything about how relative prices (the terms of those trades) will behave.

We don’t live in a barter economy, we live in a monetary one. We sell our labor and production for dollars and then exchange those dollars for other people’s production and labor. Let’s start by assuming that the only reason anyone wants money is to exchange it for something else, and consider our apple-grower again. An environmentalist group scares people into thinking that apples are covered with nasty pesticides and suddenly hardly anyone wants to buy apples. What happens? The price of apples drops until it is low enough that somebody will buy them, or if nobody wants them at all, the price goes to zero and the apple grower is effectively out of the economy. The total value of goods produced in the economy drops because the apples are worth less (maybe even worthless), but since the apple growers income has also fallen, the demand for goods falls by the same amount. Say’s Law still holds: total demand is equal to total supply.

However, what happens if the supply of money changes? Back when gold was money, a major gold strike could create a lot of new money in a hurry without increasing the supply of other goods and services. When people tried to exchange all that new money for the same limited supply of goods, there wasn’t enough to go around. Shortages and inflation resulted. Recall from my last post that since money is the unit of account, the only way for it’s relative price to change is for the general price level to change. This takes time, and in the meantime there was both an excess supply of money and an excess demand for goods.

If we relax the assumption that people only want to hold money for exchange purposes, we can get other kinds of trouble as well. One fine day the public discovers that many of the investments they thought were safe are actually managed by Bernie Madoff disciples. Suddenly everyone wants to hold something really safe, and what could be safer than holding money itself? Now at least some of the producers arriving at market find people much less willing to part with their money in exchange for what is on offer. Sales and employment fall. There is an excess demand for money, and it is exactly equal to the excess supply of goods and services.

What economists call aggregate demand is the total demand for goods and services. What we’ve just seen is that a shortfall in aggregate demand is matched by an excess demand for money. If the supply of money were increased by that same amount, the shortfall in aggregate demand would go away.

This, of course, is pretty much the story of the last two years. Rising mortgage delinquencies and falling home prices changed people’s beliefs about how safe their funds were in banks and hedge funds that owned a lot of mortgage-backed securities. The “safe-harbor” demand for money, particularly U.S. dollars, rose drastically. Instead of meeting the higher money demand with higher money supply, the Federal Reserve increased the demand for money even more by starting to pay interest on reserves. The recession was an inevitable result.

My third rule for monetary policy is next.

Libertarian Monetary Policy, Part I

Once I started writing this, I realized that it was too long for a single post. So here is part 1, with part 2 to follow in a day or two. I don’t yet know if I’ll need a part 3.

Libertarians want government to do less. Many seem to criticize every change in monetary policy as just another instance of the Fed interfering with the functioning of the economy. Finance and economics blogs have a lot of self-labeled libertarian commenters calling for the abolition of the Fed. Some would replace it with privately issued competitive moneys, others want to return to the gold standard. I believe the former is highly unlikely to come about for several good reasons, and the inherent flaws of the latter were largely responsible for the Great Depression. So for this post, at least, I’m going to take as given the continuance of our current fiat money regime with a central bank that can create or destroy bank reserves at will.

In today’s system, every time the federal government buys or sells anything, it affects bank reserves, because the Fed is the government’s bank. When the government buys something, even something routine like the labor services of government employees, it typically pays with a check drawn on the Treasury’s account at the Fed. When the check is deposited, the depositor’s bank forwards the check to the Fed in return for a credit to the bank’s reserve account at the Fed. Payments made to the government, like taxes, have the opposite effect, as your check to the government results in the Fed debiting your bank’s reserve account. Since every financial transaction involving the government on either end changes the supply of bank reserves, you could think of every such transaction as a mini-monetary policy.

In fact, these routine transactions are sterilized (offset) by the Fed buying or selling short-term Treasury securities, or by short term loans called repos (repurchase agreements) and reverse repos. The Fed intervenes this way in the overnight money markets just about every business day. If it didn’t, the fluctuations in bank reserves due to federal transactions would make short term interest rates bounce around much more than they already do. But notice that the decision to smooth them out is itself a policy decision. Unlike the gold standard days, there’s nothing automatic about it. The idea that there is some kind of “non-interventionist” monetary policy out there that could and should be implemented is wrong.

If “non-intervention” is not the right guide to monetary policy, than what is? Stop for a minute and think about the institution of money itself. Money is one of the greatest inventions of all time, right up there with fire and the wheel. Selling products and services for money is far more efficient than arranging long chains of barter trades. And being able to express values in a common unit is also of great importance. How can anyone make intelligent decisions about resource allocations without being able to numerically compare the alternatives? Indeed, it is hard to imagine how any society could advance much beyond the aboriginal without something serving as a medium of exchange and unit of account.

So the first rule of monetary policy is: Don’t screw it up so badly that money can no longer function as the medium of exchange and unit of account. You might think this is so obvious that it goes without saying, but these historical examples of hyperinflations will disabuse you of that. In many of these instances, the value of money kept changing so fast that it became worthless as a unit of account, and so much of it was needed for even ordinary transactions (trade you a wheelbarrow of currency for that loaf of bread, sir?) that the medium of exchange function was destroyed as well.

To motivate the second rule, consider what happens if the demand for oil increases but the supply of oil does not. Absent any interference from the authorities, the price of oil will rise until it reaches a level at which the quantities supplied and demanded are the same. The same is true for all traded goods and services. When supplies and demands change, the individual prices change to balance supplies and demands.

Well, not quite all. There is one thing we all trade whose price can’t easily adjust because it is the unit of account. That is money itself. The monetary value of a dollar is, by definition, one dollar. There’s no way to change that. But the real value of a dollar is what you can get in exchange for it. Suppose all money prices suddenly doubled. Then the real value of a dollar is cut in half. If prices change by X percent, the value of a dollar is its old value divided by (1 + X/100).

The important point is that, since money is the unit of account, the only way to change it’s value is to change the general level of prices. In the real world, however, that is always disruptive. Prices don’t all change at the same rate and in exact proportion to each other. So you’re going to end up changing the relative prices of some things, and there will be shortages and surpluses in various markets while things adjust to the new overall price level. Financial assets and liabilities are denominated in dollars, so if you’re increasing the price level, debtors will be repaying creditors with dollars that are worth less than the ones they borrowed. If you’re decreasing the price level, you make it harder for borrowers to repay the loans they took out, and some may be ruined. Declining price levels usually come about after long recessions.

Because changing the value of money is so disruptive, the second rule of monetary policy is that shifts in money demand should be accommodated by equal shifts in money supply. Actually, this rule may need a bit of modification as we go along, but for now it’s OK.

That’s all for part 1. Part 2 should be posted in a day or two.

How macro theories become influential

In reply to Scott Sumner’s question

How was it that just a few years later it was almost impossible to get anything published without assuming rational expectations, efficient markets, etc.

Arnold Kling complains that a handful of influential economists fell in love with the mathematical techniques involved, and refused to hire young economists who didn’t jump on the bandwagon.

But Kling doesn’t really say just why it was that these influential guys liked Rational Expectations (Ratex) so much. I was a few years behind Kling in grad school, starting in 1984, and the new way of doing macro (or not doing it, as Kling might argue) was well-established by then. But we did cover some of the older stuff in our classes, and I later read quite a lot of it.

What Kling forgets is the terrible mess macroeconomics was in before Ratex. It featured interminable arguments between Keynesians and Monetarists, with Keynesians holding the academy while recent history favored the Monetarists. There were a few people like Clower and Leijonhufvud looking for ways to come up with reasonable assumptions that would explain patterns seen in real-world data, but their models had an adhoc feel to them.

Ratex was a clean and rigorous approach that bypassed the stale arguments, at least at first. You can argue that the simplifying assumptions needed to make the models tractable are wrong, or that they’re leaving important stuff out, but as Tom Sargent often says, “It takes a model to beat a model” and Ratex is the only game in town.

After all the effort that’s been put into it, however, none of the macroeconomic theories have really convinced most people that they’re right enough to serve as a reliable guide to policy. What policymakers mostly believe in is experience. Friedman and Schwartz’s Monetary History had a huge impact because it presented a lot of empirical evidence, not theory. Keynesian economics purported to explain the Depression, and was later discredited by it’s failure to explain the 1970’s. Especially when it comes to arguments aimed at the general public, theoretical considerations don’t seem to persuade anyone of anything.

Milton Friedman used to say that the job of the theorist is to have a new theory ready to go when the current one fails. By that standard, it seems the quasi-monetarists have earned a shot at the big leagues.