Two years ago Allan Meltzer warned that the Fed’s policies would lead to high inflation. Paul Krugman and I told him he was wrong. In a new article in the WSJ he repeats this warning. But today I’d like to focus on something more disturbing, the end of monetarism as a powerful intellectual movement that addresses our problems. Here’s how Allan Meltzer begins:
Day traders and their acolytes tried to pressure the Federal Reserve to open the money spigots wider this week. They called for QE3, a third round of unprecedented quantitative easing. Fortunately, the Fed said no to QE3, at least for now. But it did vote to continue its super-easy, zero-interest-rate policy until mid-2013, well after the next presidential election.
Alarm bells are already ringing. This isn’t monetarism at all. Milton Friedman reminded us that ultra-low rates meant money had been tight. And it wasn’t just Friedman. In this (undated) paper, Allan Meltzer argued that Japan needed easier money at a time when interest rates were near zero and banks had plenty of reserves. The argument that money is easy due to low rates is a Keynesian and Austrian argument. Meltzer continues:
How can the Fed know now that a zero-rate policy will be required two years from now? It can’t. Yes, economic growth has slowed, and forecasts of future growth decline daily. But the United States does not have the kind of problems that printing more money will cure.
So what kind of problem do we have?
Banks currently hold more than $1.6 trillion of idle reserves at the Fed. Banks can use those idle reserves to create enormous amounts of money. Interest rates on federal funds remain near zero. Longer-term interest rates on Treasurys are at record lows. What reason can there be for adding more excess reserves?
This is odd for two reasons. It doesn’t tell us what sort of problem we have (demand or supply-side.) Given that NGDP has grown 4% over three years, whereas the trend rate would be about 15%, I would think a monetarist would see the problem as demand side. Friedman often pointed to the slowdown in NGDP growth in Japan during the 1990s. I’m also confused by the point of this paragraph. It seems to imply that the US is in a liquidity trap. But monetarists like Friedman and Meltzer denied Japan was in a liquidity trap, despite large levels of bank reserves. This sounds more like Keynesianism. Meltzer continues:
The main effect would be a further devaluation of the dollar against competing currencies and gold, followed by a rise in the price of oil and other imports. Inflation is now at the edge of the Fed’s comfort range, which is below 2%. Money growth (M2) reached 10% for the past six months, presaging more inflation ahead.
Now I’m totally confused. If the US were in a liquidity trap, then adding more reserves would not devalue the dollar. In the monetarist model there is no mechanism where monetary stimulus depreciates the dollar but fails to boost NGDP. None. This is the sort of argument made by Joe Stiglitz. I’m very surprised to see Allan Meltzer make this argument.
I’m even more shocked by the prediction of more inflation ahead. Oil prices have fallen sharply. Tips spreads are low. Stocks are falling. We have 9.1% unemployment. Where exactly is this high inflation going to come from? Gasoline prices? Rapid growth in wages? I just don’t see it. M2 growth was even higher in late 2008, leading to the previous false prediction by Meltzer (of high inflation ahead.) But at least I can’t deny that using M2 is an application of old-style monetarism. It’s also why us quasi-monetarists now focus on other indicators. Skipping ahead a bit:
A large part of our current unemployment problem reflects the unsold stock of housing left from mistaken past housing policies. We cannot quickly convert most carpenters and bricklayers into computer operators. Short-term policy actions will not solve that problem. But population growth, falling housing prices and rising rents will eventually help by stimulating enough new construction to put many in the housing industry back to work.
This is the sort of quasi-Austrian argument that Friedman and Schwartz severely criticized in their Monetary History. It’s also completely false, for many reasons. First, the job losses in housing construction were far too small to significantly impact unemployment. Roughly 70% of the decline occurred between January 2006 and April 2008, and yet unemployment merely edged up from 4.7% to 4.9%. The big job losses occurred in late 2008 and 2009, as the sharpest fall in NGDP since 1938 (something monetarists should be very worried about) led to a decline in commercial construction, manufacturing and services. Eventually even state and local government jobs were hit. Housing construction over the past 10 years is far below previous decades, the empty homes are due to poverty and unemployment caused by falling NGDP, not excess construction. Many young adults are jobless and still living at home. Meltzer continues:
The U.S. also has to make a major transition from a consumption economy to one that exports more and grows consumer spending more slowly. That transition has started, but it will not occur quickly. Those who look to consumption spending to recover its old path are hoping for a past that should not return.
That might be true, but wouldn’t monetarists argue that any transitions would be easier if NGDP hadn’t fallen at the sharpest rate since 1938? Meltzer told us that monetary stimulus would merely depreciate the dollar. That’s not a monetarist argument, but let’s say it’s true. Why would that be bad if we need to transition from consumption to exports? And why did Meltzer argue that Japan should engage in additional monetary stimulus to depreciate its currency at a time when rates were near zero and banks had lots of reserves? Why was currency depreciation a useful tool for Japan, but not the US?
Meltzer goes on to make lots of sensible arguments for structural reforms to boost productivity growth. I certainly agree we have structural problems—indeed I’m increasingly sympathetic to Tyler Cowen’s Great Stagnation hypothesis. But we also have a demand problem, which seems obvious to me when you look at data such as NGDP growth since mid-2008. In an earlier post Kevin Donoghue made a very astute comment:
You might ask yourself why two people who differ as much in their politics as Krugman and Sumner end up sounding so similar. Could it be that empirical evidence played a part?
I think that’s right. We both look at lots of empirical data, including forward-looking market forecasts. These indicators have been consistently warning of too little AD since mid-2008.
It pains me to write this because monetarism has greatly informed my worldview. Meltzer is probably one of the three most distinguished post-war monetarists (along with Brunner and Friedman.) But it seems to me that this type of monetarism has reached a dead end. It needs to be reformulated to incorporate the insights of the rational expectations and EMH revolutions. It needs to focus on targeting the forecast, using market forecasts, not searching for an aggregate with a stable velocity. And it must be symmetrical, with just as much concern for excessively low NGDP growth as excessive high NGDP growth. It needs to offer answers for high unemployment, and advocate them just as passionately as Friedman and Schwartz argued that monetary stimulus could have greatly reduced suffering in the Great Depression. Just as passionately as Friedman and Meltzer argued for monetary stimulus in Japan once rates hit zero. Unemployment is the great tragedy of our time. History will judge the current schools of thought on how seriously they addressed this issue.
HT: Big thanks to Lars Christensen, who supplied both articles, and some ideas.