I want to explain why I am relatively unconcerned about an outbreak of inflation when it comes time for the Fed to unwind its stimulus efforts, efforts that involved pumping large amounts of new reserves into the banking system, and also say a few words about more immediate worries about deflation.

Inflation and deflation are related to quantity of money supply circulating in the economy, so we can learn about inflationary and deflationary pressures by looking at how the money supply is determined.

The Fed's Control of the Money Supply
The equation that determines the money supply can be written:

Ms = (multiplier)(Monetary Base)
or, more compactly,
Ms = (mult)(MB)
I'll explain this equation as I talk about how it is used by monetary policy authorities (I should note that I am focusing on the broader measure of money, M2, which is much more highly correlated with economic activity than the narrower measure of money, M1).

To control the money supply, the Fed takes the multiplier as given, and then sets the MB at a level that gives it the Ms it desires (the Fed estimates the multiplier daily). The Fed can control the monetary base, MB, fairly well, though not perfectly. This is because the MB consists of two parts, currency and bank reserves. The Fed can control currency precisely enough, and it adjusts the currency supply to meet public needs (e.g. it rises at Christmas and in the summer when the demand for actual cash is high).

Bank reserves are created in two ways, open market operations and discount loans. The Fed only controls one of these directly, and this is the main source of uncertainty in its control of the money supply. Bank reserves created through open market operations can be controlled perfectly by the Fed, but reserves created through discount loans can only be estimated, and these estimates can be wrong. The number of banks asking for a loan on a given day is not known in advance, and this is the source of uncertainty in reserve creation and hence in the MB. But the uncertainty is not very large, and I'll speak as though for all intents and purposes the MB is controlled perfectly.

Bank reserves, and hence the MB, have gone up substantially as the Fed has pumped reserves into the system to combat the financial crisis. However, the money supply has not gone up very much at all, particularly in recent months. How could that be? It's because the multiplier for M2 depends upon the quantity of excess reserves -- reserves held by banks over and above the minimum mandated by regulation. The multiplier falls when excess reserves increase, and the dramatic increase in excess reserves during the crisis has caused the multiplier to fall substantially, enough to offset the increase in MB. The result is that the quantity of money actually circulating in the economy, (mult)(MB), has remained relatively constant.

Why are Some People Worried about an Outbreak of Inflation?
The worry about inflation is that as the economy begins to recover, banks will begin to loan out their excess reserves. As excess reserves fall, the multiplier will go up driving up the money supply. As noted above, the money supply is related to inflation, and as the money supply goes up so does inflationary pressure.

What Can the Fed do to Avoid Inflation?
What can the Fed to avoid the inflationary pressures? There are two potential ways to do this. First, as the multiplier goes up, the MB can be reduced that the money supply, the product (mult)(MB), does not rise at an excessive rate. But to reduce the MB, the Fed must sell assets from its balance sheet, either government bonds or private sector assets it purchased to support the financial system, and the quantity of sales required would potentially disrupt financial markets. So the fed may not be able to reduce the MB fast enough.

But there is a second, important tool the Fed can use. The Fed was granted the ability to pay interest on bank reserves early in the crisis (on both required and excess reserves), and this gives the Fed an ability it didn't have before, the ability to influence the size of the money multiplier. How can it do this? I noted above that as the economy recovers, banks will want to make more loans. If they do, this will decrease excess reserves, increase the multiplier, increase the money supply, and create inflationary pressure. But suppose the Fed raises the amount it pays on reserves to a level slightly above the amount banks could earn by loaning the money out. In that case, banks would have no incentive to part with their excess reserves -- they make more money by holding on to them -- and the inflationary pressure would be avoided.

When the economy begins recovering, the Fed does not want to completely shut down the flow of excess reserves into new loans, just the excessive part that is inflationary. So it wouldn't want to actually raise the amount it paid on reserves above the market rate of interest to deter all loan activity, but it would want to raise the rate it enough to give banks the incentive to hold onto enough excess reserves to keep inflationary pressures in check. The incentive to make new loans depends upon the spread between what banks can earn loaning reserves out, and what they can earn by holding them as excess reserves. As this spread narrows due to an increase in the amount the Fed pays on reserves, the incentive to make new loans falls.

I believe that when the recovery does finally come, the use of traditional open market operations to reduce MB along with changes in the amount paid on reserves that keep the multiplier from rising will be more than sufficient to keep the money supply, (mult)(MB), from rising rapidly and keep inflationary pressures in check. But those worries are down the road. My more immediate worry is actually the opposite right now, that we are headed for deflation, and I am far from alone in worrying about that.

What Can the Fed do about Deflation?
I hope to take up deflation worries in more detail at another time, but let me add a bit here since the equation above is useful for this topic as well. If deflation is the worry, and it should be right now, one alternative is to do the opposite of the policies described above, i.e. reduce the interest paid on reserves or even charge a fee for reserves that are held in banks (so they will increase loans to avoid the penalty). But it's not clear that we want to encourage banks to make loans they believe are overly risky just to avoid a financial penalty, and other means of spurring new loan activity may be more desirable.

The key to avoiding deflationary pressure comes through the changes in the multiplier, not through changes in MB. Traditional open market operations that increase the MB by simply piling reserves up in banks will not offset deflationary pressures. As the Fed increases the MB, the multiplier will fall in proportion as banks hold the newly created reserves in their vaults as excess reserves, and there won't be much impact on the money supply. So manipulating the MB is unlikely to work.

What will work is something that induces banks to lower excess reserves through the extension of new loans. I believe that incentive must come mainly from the demand side. As just noted, increasing the money supply by increasing the MB won't work, and penalizing banks for holding reserves is unlikely to result in the type of loan activity we'd like to see. However, if banks see firms with solid investment opportunities coming through their doors asking for loans, they'll make them, so the key is to generate new loan demand. Additional fiscal policy spending initiatives, investment tax credits, and other traditional demand side mechanisms could be very helpful in bringing this about.

Will the Fed be Successful?
I have little doubt that the Fed has the will and the means to control inflation, and that it will do whatever is necessary to keep inflation under wraps. What is less clear is whether the Fed has the will and the means to combat the deflationary pressures we are now facing. As for the means, it will be difficult to generate new loan demand by working on the supply-side of credit alone. The Fed may have the ability to help a bit by using quantitative easing, or by generating expectations of future inflation to lower long-term real rates. But this relies upon firms being willing to respond to lower interest rates and that requires changes on the demand side, not the supply side, and, in any case, to the extent that the Fed can help, it's not clear that the will is there to do so. The failure of both monetary and fiscal policy makers to do all that they can to offset the deflationary pressures that are becoming more and more evident is worrisome -- and fiscal policy makers have more responsibility here than monetary policy makers since fiscal policy has a better chance of stimulating new demand. But both need to do what they can given the poor outlook for the economy, and the failure to aggressively pursue policies to offset deflationary pressure would be a big mistake.