Here’s a new post by John Taylor, which discusses a recent BIS report on monetary policy:
BIS analyses often contain useful warnings, including their prescient warning in the years around 2003-2005 that monetary policy was too easy, which turned out to be largely correct, as the boom and the subsequent bust made so clear. So the Annual Report is always worth reading.
That’s not clear to me; indeed NGDP growth wasn’t unusually high during the 2000s. Taylor continues:
This is especially true of the Annual Report released today because it devotes a whole chapter to serious concerns about the harmful “side effects” of the current highly accommodative monetary policies “in the major advanced economies” where “policy rates remain very low and central bank balance sheets continue to expand.” Of course these are the policies now conducted at the Fed, the ECB, the Bank of Japan, and the Bank of England. The Report points out several side effects:
- First, the policies “may delay the return to a self-sustaining recovery.” In other words, rather than stimulating recovery as intended, the policies may be delaying recovery.
- Second, the policies “may create risks for financial and price stability globally.”
- Third, the policies create “longer-term risks to [central banks’] credibility and operational independence.”
- Fourth, the policies “have blurred the line between monetary and fiscal policies” another threat to central bank independence.
If you simply replaced the word “accommodative” with “contractionary” I would be in complete agreement with Taylor. And that’s actually pretty mind-boggling, when you think about it. It’s not that strange that we might differ on whether current monetary policy is contractionary or accommodative (although non-economists must be appalled that economists can’t even agree on something as basic as that.) What’s really shocking is that we even agree on the four effects from that monetary policy, even though logically one would think that accommodative policies would produce the exact opposite effect from contractionary policies. To make this point clearer, consider the fifth effect quoted by Taylor, which I omitted:
- Fifth, the policies “have been fueling credit and asset price booms in some emerging economies,” thereby raising risks that the unwinding of these booms “would have significant negative repercussions” similar to the preceding crisis, which in turn would feed back to the advanced economies.
Since I think policy has been contractionary, I obviously can’t agree with that fifth effect. But I do agree with the first four!!
How can this be? How can two economists disagree on something so fundamental? It would be like two physicists disagreeing on whether a rising bar of mercury in a thin glass tube indicates rising or falling temperatures. No, it’s even worse. It would be like one scientist saying it indicates rising temps, and arguing that’s why ice is melting, and the other arguing that it indicates falling temps and arguing that’s why ice is melting. I.e. that H2O has the property of melting when things get really cold.
So the next question is; who is the crackpot who thinks a rising bar is colder temperatures, and who also thinks water melts when it gets really cold.
Obviously I don’t think either of us is a crackpot. And I freely acknowledge that I’m in the minority here, so that if there is a crackpot it’s probably me. But nonetheless let me try to explain how (I believe) most of the profession ended up being wrong here.
1. Although I’m in the minority, Milton Friedman argued that ultra-low rates are a sign that money has been tight (in reference to Japan in 1997.)
2. Ben Bernanke said in 2003 that neither interest rates nor the monetary aggregates were good indicators of monetary policy, and that only aggregates such as NGDP and the CPI are reliable indicators of the stance of monetary policy. If you average those out, then you notice that the growth rate of nominal aggregates since mid-2008 has been the slowest since Herbert Hoover was president.
Here’s where I think Taylor went wrong. He has money being very accommodative in 2003, 2004, 2005, 2008, 2009, 2010, 2011, 2012. OK, then tell me how all this accommodative policy brought us the slowest NGDP growth since the early 1930s? Why has headline CPI inflation averaged 1.1% since July 2008?
To go back to my temperature analogy, I’m claiming that NGDP and the CPI are the melting ice. We can disagree about whether a rising bar of mercury means that temps are rising or falling, but surely we can all agree that melting icecaps show it’s getting warmer.
And we can disagree about whether ultra-low interest rates and a bloated monetary base indicate easy or tight money, but surely we can all agree that ultra-slow NGDP and CPI growth are signs of tight money?
So how did Bernanke (in 2003), Friedman, and I end up in the minority? (So much so that even Bernanke’s jumped ship, calling current Fed policy accommodative even though his 2003 criteria suggests it’s ultra-tight.)
Monetary economics is a very strange field. As long as things are fairly normal (as during the Great Moderation) economists of wildly differing stripes see things in roughly the same way. Day-to-day adjustments in monetary policy are done via the liquidity effect, which means an unexpected rise in the Fed’s target rate really is contractionary, relative to the alternative. In addition, velocity is reasonably stable, so countries with higher trend inflation tend to have higher trend rates of growth in the money supply. Both Keynesianism and old-style monetarism have some validity—both the interest rate and the money supply seem at least slightly informative.
Now go to much higher inflation rates, such as the 1970s, or even better the Latin American hyperinflations. Now the Keynesian focus on the liquidity effect looks hopelessly out of touch, as interest rates are no longer very informative. Even worse for the Keynesians, monetarist ideas hold up pretty well. Yes, velocity speeds up, but not so much as to overturn the money/price level correlation—high inflation is associated with faster growth in M, and hence the money supply looks like a better policy indicator when inflation is high. Throughout history the quantity theory always becomes popular when inflation is high. Not just during the 1970s, both Wicksell and Keynes flirted with the QTM during the early 1920s European hyperinflations, despite their previous and subsequent bias towards interest rates as an indicator.)
Now go to much lower inflation rates, and especially a situation where nominal interest rates fall close to zero. Now both the Keynesian and monetarist indicators break down. If tight money created the fall in NGDP which drove rates to zero, then Keynesians will misdiagnose tight money as easy money. Even worse, monetarists will make the same mistake. The reason is complicated, but it has to do with the fact that velocity falls very sharply near zero. So central banks are forced to massively increase the ratio of base money to NGDP, to avoid severe deflation. But since prices and NGDP are fairly stable, the huge rise in the base/NGDP ratio means the nominal monetary base also rises sharply. That means the nominal base is sort of U-shaped, high in both liquidity traps and hyperinflation, and more modest during Great Moderations. So both monetarists and Keynesians tend to misdiagnose monetary policy during periods of ultra-low rates.
During the 1970s, the monetarists were able to show how Keynesians got it wrong, by pointing to the fast growing money supply. Then the vast majority of economists in the middle, who are open to pragmatic arguments, could say to themselves “OK, the monetarists are right, obviously if people need wheelbarrows of money to buy a loaf of bread, then money is not “tight” no matter how high the interest rate is.” But when we get to ultra-low rates and both the interest rate and the money supply signals are giving off false readings, then there is no one to correct the Keynesians. The (old style) monetarists are just as wrong. And we market monetarists are not influential enough (yet) to overcome the near universal view that money has been accommodative in recent years.
I’ve never studied topology, but I wonder if there isn’t an analogy here for the saddle point surface. The Great Moderation is like the sweet spot where things are locally flat, and Cartesian in all four directions. But when you move away from that point then both money supply and interest rate indicators become hopelessly unreliable. And monetary policy breaks down. The high inflation breakdown is less complete, because the money supply indicator still works somewhat. But the low inflation/interest rate breakdown causes mass confusion among policymakers, and may require some solution that involves thinking outside the box. In the 1930s that involved leaving the gold standard and devaluing the dollar—with the price of gold being the alternative indicator/instrument, when interest rate adjustments and/or QE weren’t working. Who will produce the out-of-the-box thinking required for our modern crisis?
I think you know where I’m putting my money.
PS. Take a look at the 5 BIS bullet points quoted by Taylor. Isn’t number one damn close to “low temps cause water to melt.”