Thursday, March 20, 2014

They updated Potential GDP last month

Real Potential Gross Domestic Product (GDPPOT) is now given in 2009 dollars. Before the February 4, 2014 revision it was given in 2005 dollars.

A lot of other series, including Real Gross Domestic Product (GDPC1), changed to 2009 dollars back in July of last year. Potential GDP has finally caught up, so now Real GDP and Potential GDP are directly comparable at FRED again.

Graph #1: Real (blue) and Potential (red) GDP, in 2009 Dollars
Here's how the two series looked, using the older data expressed in 2005 dollars:

Graph #2: Real (blue) and Potential (red) GDP, in 2005 Dollars
Not a lot of difference. The output gap there after 2008 closes a little more with the newer data. But not as much as I expected in a previous post. Why?

The "comprehensive revision" that changed Real GDP after the 2013-06-26 release didn't only change the base year from 2005 to 2009. It also increased the output numbers. It made Real GDP bigger than it was before. (Nice trick, huh?) So then, Potential GDP also had to be revised upward, to match the change in Real GDP. I didn't account for this change in Potential GDP. So the output gap doesn't close as much as I said it would, in that earlier post.

In order to compare the two versions of the output gap, I subtracted Real GDP from Potential GDP using the 2005-dollar data, and again using the 2009-dollar data, and put the results together on a new graph. Over the full period, the two "difference" lines run quite close together, as you might expect. So I zoomed in on a detail -- just the years since 2007:

Graph #3: The Output Gap using Old (blue) and New (red) data
The main thing you can see on this graph is that since 2009, the output gap gets smaller faster for the new (red) data than for the older (blue) numbers. That is what you have to expect, given CBO policy that reduces any output gap to zero in ten years:

CBO: A Summary of Alternative Methods for Estimating Potential GDP (PDF)

The surprising thing you can see on the graph is that, since 2007 or before, the new numbers show a bigger output gap than the old numbers. Not a lot bigger, but bigger. That's surprising, because Jim Bullard's argument was that Potential Output was over-estimated in the 2000s. Bullard's argument was that potential output was lower than people thought, that GDP was above potential, and that the shocking fall in Real GDP was really just a correction.

Now, it seems, they get to have it both ways: The output gap was bigger than people thought, but it's closing faster anyway.

1 comment:

The Arthurian said...

On second thought, on Graph #3 the red line starts out higher than the blue, probably because it uses 2009 dollars rather than 2005.

That explains why it starts out higher. Doesn't explain why it falls faster.