Inflation is confusing. The concept makes crazy people crazier. And even worse, it makes otherwise sober people disagree with eachother. Reading through the accounts of QE2 on the internet the past few days have solidified my view that inflation is a thorny enough concept that we should rid it from popular vernacular. Is inflation important? Sure…but what measure of inflation is correct? CPI-U? GDP Deflator? Your crazy uncle’s index? Does inflation help or hurt savers in the current landscape?

If there is anything that gets turned on it’s head when an AD recession hits, it is the concept of inflation. During normal times (full employment and capacity utilization), inflation is harmful as it drives up interest rates, discourages saving, and encourages misallocation of capital. However, none of those things apply to the current situation in which we find ourselves with a large output gap and high unemployment. Thus, we need higher inflation in order to close the output gap (the difference in money expenditures between where we are currently, and the trend rate from the Great Moderation…currently about -13%), but that turns everything that everyone knows about inflation backward. All of a sudden inflation is good for savers, good for the unemployed, and good for economic growth. Well, stable inflation expectations are key…but it’s hard to steer a ship, and it’s hard to get a non-confusing answer out of pundits and other commentators.

In order to square this circle, I propose we forget about inflation. And not just forget about talking about it, but forget about its use in the setting of monetary policy. Instead, we should target nominal expenditure at a steady growth rate (3% a la Woolsey, or 5% a la Sumner, Beckworth, etc.) with level targeting. What advantages does targeting nominal expenditure have? Well…

  • Targeting nominal expenditure (NGDP for short) allows monetary policy to better address recessions which arise from both aggregate supply and aggregate demand shocks. David Beckworth has an excellent discussion of this point.
  • NGDP is a better indicator of monetary shocks than inflation indicators like CPI. Because prices are sticky, and because measures of inflation are so problematic, a fall in NGDP won’t immediately show up in inflation numbers. Also, if there is a large price shock in something like oil, this will raise the money price of all goods and services, causing anyone focusing on inflation to miss the underlying weak economy…and thus potentially set monetary policy to be too contractionary (sound familiar?).
  • NGDP allows us to broaden our focus to aggregates like MZM, asset prices, yields, excess reserves etc. We’ll relinquish our inane focus on interest rates, which are a very problematic indicator of the stance of monetary policy, and have a much better picture of the health of the economy.
  • NGDP sounds better. People have an innate fear of inflation. Inflation destroys savings, after all…and we all know frugal people are virtuous. Well, how about, in the event of a recession, instead of economists clamoring against the crowd that we need inflation, they say that we want aggregate expenditures (and thus nominal income) to be at some level higher than it currently is? Money illusion is a powerful motivator. Who would argue with that?

Targeting nominal expenditure would be a beneficial step from both an economic theory perspective, and a public relations perspective. Lets take the confusing concept of price inflation out of our discourse, so that we can see the world more clearly.

P.S. We are currently 13% below the target path from the Great Moderation, and are where we were at before the crash of Sept/Oct 2008. To make that up by 2011:Q3, the Fed would have to target NGDP at $17.6bn (to continue on a 5% NGDP growth path). However, Bill Woolsey favors a 3% growth path for money expenditures, which means that the Fed would only have to target a 13.8% increase by 2011:Q3 (or $16.4bn), and then continue on with 3% growth, level targeting, from then.

Update: Found the link to Beckworth’s article!