2012-12-04

A Market Monetarist Thought Experiment

Considering the recent posts at The Money Illusion (here and here), as well as subsequent posts by Bill Woolsey (here and here), and Robert Murphy, I thought I might add some uneducated thoughts here. My goal is not to add clarity to the debate so much as to talk my way through a series of thoughts that are either not addressed by any of the above or not addressed in a way that feels adequately intuitive to me.

The Thought Experiment
I suppose when the Federal Reserve was created, there was no widespread mechanism to inject money into the economy, other than government spending and Open Market Operations, i.e. Fed interaction with the financial sector. And that has been true for many years of fiat currency. A more democratic solution may have been available, but at least the argument could be made that there was no better way to reach the broader economy than OMOs. So the Fed had to deal directly with banks, and that was that.

Even if that was true for a long time, it is obviously no longer true in these days of virtual banking and near-universal internet access. So let me outline a potential replacement for the current monetary policy apparatus.

Every US resident with a valid Social Security Number could register that number online at the Federal Reserve's website. The Fed could then establish online bank accounts with every individual who registers. Working out a deal with the IRS would enable the Fed to evaluate each person's annual gross income from the previous year and make monthly (or even daily, if needed) deposits in each account in accordance with the stated NGDP target.

Therefore, if the NGDP target is, say, 4% then each US resident who registers with the Fed would receive their previous year's income multiplied by 4% in annual deposits, effectively raising their personal income by the exact value of the NGDP growth rate target.

Let us further stipulate that these accounts would by one-way only. That is, the Fed could deposit money, but not withdraw it. Likewise, US citizens could withdraw money but not deposit it.

Now, such an arrangement would certainly meet the needs of an NGDP level targeting policy. When more inflation is required to increase NGDP, deposits to resident accounts could be increased by the Fed to any degree they like. When less inflation is required to meet the 4% target, deposits could be decreased (but never made negative).

In this way, money would be circulated throughout the economy in a way that all economic actors could easily access. The deposits would be regular and consistent, meaning that they would soon be built into society's expectations. The financial industry would experience no special economic impacts that didn't already impact the rest of us. No one group of people could be said to benefit more than any other.

In short, monetary policy would be carried out through ordinary people, in the most "populist" way available.

To make matters even more attractive, independently auditing this system would be painless. Any independent auditor could simply divide the total value of all deposits by the total value of all reported gross annual income in the United States from the previous year. If the audited value differs from the Fed's publicly declared inflation target at all, then the auditor will have detected corruption and will have authority to determine which account or accounts received a disproportionate payment and the money could be reclaimed.

The only challenge would be correctly forecasting NGDP growth.

Why Not?
So why doesn't the Fed conduct monetary policy this way? All of the necessary technology and infrastructure is already in place. Assuming one actually believes in NGDP level targeting as a monetary policy strategy, this seems like a failsafe way to do it.

Milton Friedman, of course, suggested we replace the Federal Reserve with a computer program capable of forecasting economic growth and setting interest rates accordingly. (Cf., for example, this article.) The above approach could certainly be applied by just such a computer program, too. Automatic forecasts and money deposits are not beyond the capability of a society that set foot on the moon over 30 years ago.

It is just this sort of reasoning that leads many people like myself to suspect widespread corruption in the Federal Reserve system. If all this really needs to be done, then why do we have people in charge of conducting this business with a few mega-rich banks? Why not conduct monetary policy directly with the people? Shrouding the whole thing in the secrecy of a special market segment to which most people have no access does not exactly lend the Federal Reserve system a trustworthy veneer.

So when people like Scott Sumner suggest that "it doesn't matter" who gets the Fed's money first, I think it's natural to point to the secrecy surrounding the system and the public's widespread inaccessibility to it and suggest that perhaps what Sumner says is only true in theory.

In practice, I think the track record for corruption is too consistent to ignore. Perhaps I would be more inclined to Market Monetarism were this not the case.