Interest Rates Don't Matter as Much as Most People Think

by: Scott Sumner

It’s good to see Matt Rognlie blogging again. I’d like to take issue with the following claim:

You may have heard from Scott Sumner, among others, that interest rates are a terrible indicator of monetary policy. This is actually true if we’re talking about current interest rates: it’s possible for monetary policy to be effectively tight because the Fed’s policy rule is contractionary, even if the current rate is very low. (Maybe it plans to raise interest rates dramatically in a few years, or whenever the economy shows signs of an expansion.) The key is to remember that monetary policy works through the entire expected path of future interest rates, not just the current rate. But interest rates are ultimately the key mechanism through which monetary policy affects the economy.

I’m told many models imply that interest rates are the key mechanism by which monetary policy affects the economy. But I’ve never been convinced. Start with a flexible wage/price model. In that case, a one time increase in the money supply will tend to lead to a one time increase in all nominal prices and aggregates, leaving interest rates unchanged. Ten percent more money leads to 10% more NGDP. A currency reform is a good example. Obviously interest rates play no role in that process.

Now it might be argued that wages and prices are sticky, so my model isn’t at all interesting. Not so fast. Even Keynesians agree that wages and prices are flexible in the long run. So it’s not obvious that interest rates play any role in explaining the long run affect of M on NGDP. They might, but theory says we’d get to the same long run equilibrium, with or without interest rates playing a role in the transmission mechanism. The “hot potato effect” is enough.

The more interesting question is whether, in a sticky wage/price model, interest rates play an important role in the short run non-neutrality of money. They might, but once again that’s not at all obvious. For instance, suppose prices were much more flexible that wages, with commodity prices adjusting immediately to monetary shocks. In that case you can get non-neutrality via sticky wages. I’d guess you could also do so with monopolistic competition and sticky prices, but it’s not my area of expertise.

So theory doesn’t resolve the question, we need to look at which approach is the most useful. And here’s where I am very underwhelmed by the Keynesian interest rate approach. Maybe this isn’t fair, but I can’t help pointing out that I repeatedly see Keynesians make spectacularly incorrect calls by using interest rates as an indicator of monetary policy, rather than NGDP growth expectations.

I see Keynesians argue that tight money didn’t cause the first year of the Great Depression, because rates fell sharply. It had to be an “IS shock.” I see Keynesians argue that tight money didn’t cause the Great Recession, because interest rates were very low, and falling, in 2008. And I’m not just talking about a few crackpots. These sorts of claims are made by many (perhaps most) of the most respected Keynesian macroeconomists in America. (And don’t ask me to document this, if you disagree please produce a long list of famous Keynesians who publicly stated money was tight in 2008–and put the list in the comment section. I think everyone honest with themselves knows I’m right.)

Now of course it’s possible that I’m wrong and the Keynesians are right. Maybe the low interest rates really do mean that money was easy in 2008. Then we can argue that issue.

Matt points out that the more sophisticated Keynesian models talk about the entire future path of interest rates, not just the current setting. But I don’t believe many Keynesians understand just how knife-edge those paths can be. Take the December 2007 Fed announcement, which cut rates less than expected. Rates obviously rose for short term maturities at 2:15 pm, but only out to a few weeks maturity. After that, rates fell for maturities from 3 months to 30 years. Thus a tight money policy reduced almost all interest rates at the point it was adopted. Why? Because it led to slower than expected NGDP growth at a particularly fragile time for the economy. Now tell me whether you think higher interest rates on maturities out to a few weeks is enough to drive important real effects in the economy if longer term rates are going the opposite way. Yes, using real rates makes this flaw in the Keynesian model smaller. But most macroeconomists paid no attention to the fact that real interest rates soared in the second half of 2008. Even worse, tight money can reduce even real interest rates in a forward-looking rate model.

Part 2: While I’m commenting on Matt’s recent posts, I’d also like to discuss an excellent summary he provided of the “commitment problem.” There are good theoretical models explaining why it might be hard for a central bank to credibly promise to inflate. My only problem with this literature is that it’s a set of solutions in search of a problem. No fiat money central bank has ever, ever, had the slightest difficulty in inflating. No fiat money central bank has ever tried and failed to inflate. And I think there is a good reason for that. Central banks are very concerned about their reputation. It would be humiliating to promise, say, a 5% NGDP target, level targeting, and then 4 years later when out of the recession, go “nah, nah, I was just kidding.” Again, this isn’t to belittle the excellent models that have been developed to address the “problem,” but I’m afraid the problem doesn’t exist.

Matt seems to have the same view as I do:

Recent events, however, suggest that elaborate commitment devices aren’t really necessary. The key constraint for the Fed isn’t sticking to its promises. The Fed cares a great deal about its credibility, and has a decent record of keeping promises whenever it has the nerve to make them. Instead, the problem is what kind of commitment to make: how can the Fed make tangible promises about future policy that don’t run the risk of creating larger problems down the road? This is the second issue in the literature, and it’s not trivial.

Paul Krugman on the other hand still doesn’t get this point:

By the way, it’s worth reading Ken Rogoff’s discussion at the end, in part for his confident assertion that raising the rate of growth of the monetary base would be enough to increase inflation. Actually, Japan tried that a few years later, without success; nor has the surge in the US monetary base since 2008 been either inflationary or indeed effective. All of which is exactly as the model predicted.

Actually, the BOJ did not want inflation, which is why they raised rates in 2000, and why they raised rates in 2006, and also reduced the monetary base by 20%. They have the price level right where they want it. Policy didn’t “fail.” The problem is that they don’t want higher prices, not that they can’t get them.

And of course, Bernanke repeatedly says the Fed has lots of extra tools and is far from being out of ammunition. They just don’t think extra stimulus is needed right now. They paid 1% interest on reserves in 2008 to prevent that base expansion from causing inflation. It’s clear from the market response to even trivial Fed rumors that the stock, bond, commodity and forex markets agree with Ben, Matt, and I, and disagree with Paul Krugman. Monetary policy is highly effective at the zero bound. It’s time to use it.

PS. One area I do agree with Krugman is that QE unaccompanied by explicit communication is not the most effective tool, but market reactions to rumors suggest that even that’s better than nothing.

PPS. Another reason I don’t like models with interest rate transmission mechanisms is that some of them imply the Fed can’t create 3% inflation. It’s 2% or 4%, with nothing in between. I suspect that’s because you need at least 4% inflation to get real interest rates low enough to generate enough AD expansion to raise actual prices by 4% (or even 3%), given a fairly flat SRAS and when in a liquidity trap. I’d appreciate if someone would tell me whether my intuition is wrong. I can’t read math very well.