A Young Person’s Guide to Economics

I am somewhat uneasy about writing this, as it is about the fourth post in recent weeks having a go at the Volokh guys, and one of quite a few on Tyler Cowen specifically, but I simply could not let this post pass without comment. It’s part one of a “Guide to Macroeconomics in Five Easy Lessons”, on monetary economics. I wholeheartedly support the idea of someone producing such a guide, but the actual statements made about monetary economics seem to me to be horribly confused. So much so that I’ve been reduced to commenting on it line-by-line; I wanted to write a proper response, but grew worried that by concentrating on my main disagreements, I would be implicitly endorsing some of the errors I didn’t single out.

I’ve edited this twice to moderate some of the more temperamental remarks, but the tone is still pretty angry, as I’m genuinely annoyed that this is being fed to laymen. As a result, I have perhaps been excessively inclined to pick nits; that’s how I get when I’m angry. I will accept the judgement of Brad DeLong as definitive on the question of whether I have been unduly harsh and will post an apology here if he thinks I have been. I pre-emptively apologise to Mr Cowen for the lack of civility inherent to the “fisking” genre; as I mention above, I tried and failed to come up with alternatives.

Macroeconomics in Five Easy Lessons, criticised

(plain text is the original, italics are DD coments)

Most of macroeconomics today is monetary economics.

Most of macroeconomics today is growth theory.

And the stock market is obsessed with the Fed. So the money topic is central.

Complete non sequitur. Central to what? Why does what the stock market cares about matter?

I am assuming you already know lesson one, which is printing [sic] lots of money leads to hyperinflation.

Lesson two is that the Fed can never do a good job in the long run, no matter how smart or responsible it may be. The Fed’s core dilemma is this: it can only control a tiny base, yet the broader superstructure is what matters for the economy.

Horrible abuse of the terms “base” and “superstructure” will be noted by anyone with a background in Marx. Also as we’ll see below “control” is being used ambiguously here. You might just as well say that I can control only a small steering wheel, yet the motion of a much bigger car is what matters for my daily commute

Now bear with two paragraphs of arcane terminology.

It’s not actually that arcane, which is just as well, as no attempt is made to explain it

The monetary base is currency plus reserves held at the Fed, the base of the pyramid. The Fed controls this directly and with great accuracy, if you want just think of speeding up the printing presses for more currency.

As long as you ignore the effect of the balance on the capital account on “reserves held at the Fed”!

(Don’t worry about the discount rate, when the Fed lends to banks, that is secondary and I will ignore it.)

It isn’t secondary at all, particularly if you’re operating on a standard of “what the stock market cares about”. I explained at length a while ago how the control of the overnight money rate can be used to exercise a significant degree of control of the entire yield curve.

But currency is not most of the action. Typically the Fed buys more or fewer short-term T-Bills, and deals with banks (“open market operations”).

What is the purpose of open market operations? To maintain the “effective” Federal Funds rate at or near the “target” Federal Funds rate. In other words, this “monetary base control” is entirely subservient to the aim of maintaining the “secondary” discount rate target. Think, man, or at least open a copy of the Wall Street Journal once in a while. When Alan Greenspan walks out of a FOMC meeting, what does he announce? The monetary base target for the month, or Fed Funds?

In contrast, consider what we call M2, a broader monetary aggregate. M2 contains, among other things, demand deposits. Banks lend money by writing extra zeros into the accounts of their borrowers. The Fed can influence this process (more on this below) but cannot control it with any great accuracy.

Cannot because chooses not to; it targets the interest rate, not M2. The Bundesbank controlled monetary aggregates for years, with great accuracy. Banks do not lend money by “writing extra zeros”; they have to borrow the money first, from someone who is prepared to lend it to them. And this is how the Fed (as “lender of last resort” to the money market) gets its traction.

Milton Friedman told us that “lags are long and variable.” That means that when the Fed changes the base, it has a poor idea how M2 (and other aggregates) will respond. The base could go up a little and M2 would go up a lot. Or sometimes M2 goes down. Or whatever.

Milton Friedman might have believed that, but this idea is not expressed by “lags are long and variable”. Long and variable lags is a statement about lags, not about whether the actual effect is variable. Factive use alert on “told”, by the way.

Now keep that all in the back of your mind for a minute or two. Let’s turn to deflation and inflation.

Lesson three is that both falling prices (deflation) and price inflation are usually bad (how is that for an oversimplification, albeit a correct one?).

Bloody terrible, if you want my honest opinion.

Deflation pisses people off, by making them accept lower nominal wages. Funny, but academics will scream bloody murder if you cut their wages by $500 in a year. Those same educated people find it OK if their nominal wage is constant but eroded by price inflation over time. Don’t try to understand these people, or tell them they should be like Silicon Valley, just accept them for what they are. The bottom line is that a shock deflation will put many people out of work and discombobulate the economy.

The effect of nominal wage stickiness is a very small part of what’s bad about deflation. What’s bad about deflation is that most debt contracts are denominated in nominal terms, so their real value increases when there is deflation. This can quite easily lead to a situation in which the debt cannot be serviced because the real burden has grown too great, leading to financial dislocation as productive enterprises are broken up to satisfy their nominal contracts. Furthermore, in a falling price environment, there is an incentive to postpone purchases of capital assets, which reduces investment and therefore reduces demand (the multiplier effect). Real wage effects are small in comparison.

Inflation is bad too, though no one has a good account of why.

Milton Friedman?

Just believe me on this one (and don’t be tempted to send me all this blah-blah-blah on how inflation “distorts” prices, the claim might well be true but no one has ever produced a good model or account of it).

I am indeed not tempted to enter into a discussion with someone who has such powerful defences against the danger of learning anything. For those who are interested, the work of Robert Barro in the mid 1990s is definitive as far as I am concerned; single-digit inflation has no measurable effect on output, but above about 15% the effects on the real economy are both statistically and practically significant.

Maybe the worst thing about inflation is that if you have enough of it, sooner or later you have to end up with some deflation.

In actual fact, the bad thing about inflation is that it leads to higher real interest rates, as lenders need to be compensated for inflation risk as well as inflation. Thus reducing investment and therefore demand, blah blah blah. It is also in general associated with a loss of political control and general anarchy, which has an exhilarating effect on the animal spirits of artists (pet theory alert) but a debilitating effect on those of businessmen. Furthermore (a less well-publicised effect, but one which I regard as important), it increases the real cost of providing the interest free credit (in the form of 30-day payment terms, etc) which keeps the economy going.

Now let’s go back to the Fed. A good Fed will try to prevent both deflation and inflation. When inflation threatens, the Fed will tighten the money supply, hoping to stem inflation. When deflation threatens, the Fed will loosen the money supply, hoping to stop it.

I include by citation all my comments above; the Federal Reserve of the USA simply does not carry out its inflation policy by targeting the monetary base

Sooner or later they will mess up.

Yes, people, the entire economic argument of this piece is contained in the words “sooner or later, they will mess up”. Aren’t you glad you’re getting an economics education?

I have already noted that controlling the base doesn’t give you much leverage over M2.

(an aside for anyone who cares about how monetary base targeting is carried out). In actual fact, if you want to carry out monetary base targeting, you do it either by legally enforcing reserve requirements, in which case, controlling the monetary base gives you precise control over M2, at the expense of ensuring that M2 is no longer the relevant monetary aggregate because people will find ways of making loasn which aren’t counted in M2 (Goodhart’s Law), or you do it by credibly committing to keep raising the discount rate until any bank which is growing its asset book faster than your target will be running an unacceptable risk of being caught short of funds.

Note that the stock market will second-guess the Fed every step of the way.

Duly noted, although what the stock market has to do with the supply of M2, or indeed how this sentence fits into the overall piece, is a bit of a mystery

On average, the Fed cannot make things much better

You would almost believe that “Sooner or later they will mess up”, “I have already noted that controlling the base doesn’t give you much leverage over M2” and “Note that the stock market will second-guess the Fed every step of the way” formed premises of a syllogism which had as its conclusion “On average, the Fed cannot make things much better”, but you would be wrong.

, and is simply hoping to stop things from getting worse, due to its own inevitable mistakes. The power of the so-called Fed is simply power to mess up, or power to avoid messing up too badly.

This would be a powerful condemnation of the Fed, except that we have a missing alternative. The Fed “cannot make things much better” … than what? Than the Great Depression? Yes it could and should have done. There is decent evidence that post-war (more to the point, post-General Theory) recessions have been less frequent than they were before the era of active central banking; the economy has by no means been perfectly smoothed, but central banks have certainly accomplished more than just “not messing up too badly”.

That is a good chunk of what you need to know. Oh, yes, some of the time the central bank will “goose up” the money supply to reelect the incumbent. That is usually bad.

Specifics of this extremely serious accusation would have been interesting, if there were any, which there aren’t

It is sometimes said that “the Fed controls interest rates.” True or false?

The Fed most emphatically does not set short-term rates directly in the literal sense.

This is the literal sense in which I do not set the speed of my car; I merely control the throttle of the engine. In other words, a literal sense of no relevance at all.

The Fed can push around the short-term bank lending rate, by increasing or decreasing the monetary base, by more or less trading money for T-Bills. More money usually makes the nominal short-term rate fall, as there is suddenly more liquidity. Less money makes it rise. This it can be with real accuracy, it simply keeps on trading until it gets the short-term rate it wants.

That would be the short term bank borrowing rate, Federal Funds. In actual fact, of course, this sort of Economics 101 supply and demand analysis has little real relevance to open market operations. The market knows that the Fed can force the rate to Fed Funds target, so the market rate jumps there without the actual trading having to take place. Fluctuations around Fed Funds target usually reflect the day-to-day rebalancing of cash inventories.

The Fed has very little control over longer-term interest rates. It can move them lots only by wrecking things (remember the late 1970s?).

Or by credibly committing to an anti-inflationary policy (remember the 1990s?). Again, I include by citation my discussion of the Vasciek model and Rubinomics, linked above.

Now just a little more on how the monetary base is linked to M2. The Fed can push around the monetary base, hoping to change short-term interest rates enough to affect the lending practices of banks in a predictable way. This is tough to pull off for several reasons, one being that banks may care more about the long-term rate than about the short-term rate. Another is that the bank may not care what the rate of interest is, if it feels it will never get its money back. So Greenspan really has a tough job.

There is the kernel of a sensible argument here, but it is being swamped with all this rubbish about M2. Once more, however, I beg to look at the empirical evidence on effectiveness of monetary policy, and include by citation the voluminous Federal Reserve working papers on the subject. It’s a tough job, but it’s no tougher than, say, open heart surgery.

You might wonder why the Fed doesn’t simply do nothing, and freeze the monetary base (George Selgin once pushed this idea). I take this proposal seriously, but it has two problems. First, we would all have to get used to regular deflation. Second, short-term interest rates jump around a lot. Banks would scream if the Fed didn’t smooth out some of those movements.

But what? Now freezing the monetary base is meant to simultaneously freeze all other monetary aggregates? This is not even consistent with the line of argument taken above!

Milton Friedman used to think that a steady rate of money growth was a good idea (he has since moved away from this position, as has almost everyone else). But it is not. If you control the growth rate of the monetary base, M2 still moves around. Controlling the growth rate of M2 is much harder, plus it leads to wrenching swings in short-term interest rates.

It is by this point very difficult to tell what is being asserted. Why are “wrenching swings in short term interest rates” bad if measured, controlled movements in short term interest rates have no effect?

So we’re back to the Fed goofing, sooner or later, no matter what.

As I said above, this is the entire argument of this piece, and it appears to me that no baseline has been set for measuring what counts as a “goof” by the Fed.

The worst thing they can do is to engineer a sudden deflation. At least they have learned not to do that, so we are pretty lucky.

My apologies to the specialists, and please note I am using M2 as a proxy for all broader measures of money.

I have no real scale for guessing whether my comments represent the quibbling of a specialist or not, but I have to say that I think that as a primer in monetary economics, this is dangerously inaccurate on a number of topics. There is a significant danger of an educated layman understanding less about the workings of monetary policy after having read it than he did before (thankfully, it does not actually discuss the subject of monetary economics at all, despite the title). I recommend as an alternative this Powerpoint presentation.