According to Knut Wicksell (one of TH’s favorite economists) this is how you conduct monetary policy in an economy that relies purely on banks to generate money in the economy.
He came to that simple conclusion in 1906 after writing a brilliant book on the subject some years before. And although it sounds obvious, it was somewhat “out there”. Why? Because Wicksell was arguing that in a world where the government had no control over the amount of money being produced in society, it could still determine the price level. At the time it was radical and foresighted – money was tied to gold, and national central banks often didn’t even exist.
Anyway, this is (briefly) how the idea works. When you borrow money, you care about the real cost, which is the interest rate minus the inflation rate that you expect over the term of the loan (and a negligible interaction term that we’ll ignore).
where r is the real rate of interest (which is what you care about), i is the interest rate you pay on the loan, and p is the expected inflation rate. If the interest rate is 5% and the inflation rate is 3%, then the real cost of borrowing is only 2%. The lower the real rate of interest, the more borrowing that will take place and the greater the level of aggregate demand and the more expansionary monetary conditions are.
There are two ways to implement monetary policy in this framework. The first is to do what Wicksell suggests. Suppose, just for illustrative purposes, that prices are falling and people expect that to continue so that p is negative. Then, the real cost of borrowing will be above the nominal rate and rising, so firms and households that are borrowing to buy goods and services, will tend to borrow less. To prevent a fall in aggregate demand (which would just add to the process of already falling prices — Wicksell’s cumulative process), the central bank must cut interest rates to get the real cost of funds down. That is fairly straight forward.
But suppose the central bank can’t cut rates because, say, they are already close to zero, what then? Well, in this framework, the only other variable that the central bank can tweak is the expected inflation rate. And they can do this, through what is known as “open mouth operations” (a play on “open market operations”, which is a traditional way that central banks can implement monetary policy, and a term that was coined, TH believes, in a Reserve Bank of New Zealand discussion paper in the mid 1990’s but he can’t find it – though it has since become somewhat common amongst monetary economists).
How do open mouth operations work? Well, the central bank just tells you what you should expect in terms of inflation. That’s it! And so long as you believe them, it should work in pretty much the same way as a cut in interest rates (Torrens knows that this is probably not quite true, but for now he’ll roll with whatever flack he gets).
This raises an interesting question. Why don’t central banks engage in open mouth operations more fully? There have certainly been some efforts to do something similar – central banks pay careful attention to their monetary policy statements because they know that markets pore over them for tid bits that will help them to understand what to expect in the future. The Fed, has stated that it expects to maintain an exceptionally accommodative until mid 2015, while highlighting that if monetary policy wasn’t accommodative, inflation would be too low. But the efforts at conditioning expectations fall short of directly conditioning inflation expectations altogether. See Jim Haley’s musings here.
It’s somewhat understandable when there is a mandated inflation target that binds the hands of a central bank, but that is not the Fed? Couldn’t the Fed simply say (hypothetically – TH doesn’t know what the right numbers are) that it will keep monetary sufficiently loose so that inflation can be expected to rise to say 4% over the next 2 years, after which, if the output gap has closed, it will tighten monetary policy to bring inflation back down to its target? The effect of telling people that once the economy was operating at its potential it would reduce inflation would simply signal that it would tighten policy once conditions were right, so surely no problem there. This would be a bit like price level targeting or nominal GDP targeting, but without the central bank tying the inflation outcome to the growth of potential GDP, something that the central bank has no control over.
I guess one reason might be a time consistency type argument. What happens if in 2 years Inflation is at 4%, and due to some unforeseen circumstances, the economy is still in the hole? Then the central bank has two problems – inflation and unemployment. And, indeed, this could soon be the predicament that the ECB will find itself in. And it is a predicament that no inflation targeting central bank wants — recession and inflation. Euro area inflation is currently 2.5%, which is above its target of 2% (though technically the ECB does not really have an inflation target – it just aims to keep inflation below, but close to 2%). And this leaves the ECB in an awkward position because, since it has a defacto inflation target and inflation is above target, people will start to expect that inflation will fall. This means that the real interest rate will rise if not offset by a further interest rate cut, which could lead to a further fall in aggregate demand. That most certainly will lower inflation, but it will hurt growth, and boost unemployment. So maybe the ECB should just say that it expects that inflation to rise further as the economy comes out of recession and that the rise will be reversed thereafter. That would be better than the alternative of leaving markets unsure of what to expect – after all uncertainty is currently killing growth.