Why we have to be careful equating CPI growth to inflation

We have previously mentioned how important it is to interpret CPI growth carefully. Inflation is the trend rate of growth in the “general price level”, and as a result there are other factors that get caught up in our attempts to measure inflation.

James Hamilton at Econbrowser pointed to a paper by Reis and Watson which points out just how essential this difference is in a “low inflation environment”. The money quote from the abstract for me:

We find that pure inflation accounts for 15-20% of the variability in inflation while our aggregate relative-price index accounts most of the rest.

The pure inflation mentioned above is the fundamental increase in the general price level we often complain about as economists. “Inflation” in this paper is growth in the private consumption expenditure deflator (which is similar to the CPI – except that the bundle of goods is not fixed).

Now we don’t want to try and prevent relative price shifts – as that is the whole purpose of the price signal. So understanding this distinction is important. Very interesting.

Update: I don’t think I was clear enough on where I think the value is here. The paper seems to indicate to me that movements in the CPI and PCE deflator are relatively poor indicators of the magnitude of any change in inflation. This is a very good point, and I love it that this paper was able to mince the data up and show this.

Update 2: Paul Walker blogs on a pre-release of the paper here. Paul concludes:

They found that once they controlled for relative price changes, the correlation between (pure) inflation and real activity is essentially zero

That is true. But let us be clear here, this implies that there is no money illusion (which economists currently assume), and so all quantity issues stem from nominal rigidities in prices (which we have discussed). As a result, even pure inflation is still costly, as nominal rigidities exist causing a mis-allocation of resources (since relative prices get messed up).

Another interesting conclusion in the paper is that the rigidities in the labour market aren’t as strong as we would expect. If this is shown to be the case over time it would change my implicit view of the economy. Man I wish I could do a study like this for the NZ economy 🙂

Andrew says:

May 12, 2009 at 14:43

I got to see Reis present this paper last week: http://www.columbia.edu/~rr2572/papers.htm#DPI

Highly related.

Matt Nolan says:

May 12, 2009 at 14:59

There are a lot of cool papers there – thanks 🙂

Paul Walker says:

May 12, 2009 at 16:40

I blogged on a pre-publication version of that paper back in Feb-08. What I thought was cool was that they could find a measure of relative price changes, which as you note we don’t want to prevent. If StatsNZ could produce measures of pure inflation then people would have better idea of what changes are relative and thus we could get a better allocation of resources.