Photo: Wikimedia Commons
Here’s a puzzle. We’d like your input on it.In a speech today (.pdf), St. Louis Fed President James Bullard argues against the notion of an “output gap.”
Specifically he slams the idea that the economy could or should be operating at anywhere near the level it was in 2007.
This chunk of his speech is a few paragraphs, but it’s worth reading closely, because on the surface he makes a very logical argument.
The key to the large output gap story is the use of the fourth quarter of 2007 as a benchmark for where we expect the economy to be today. The idea is to take that level of real output, assume the real GDP growth rate that prevailed in the years prior to 2007, and project out where the “potential” output of the U.S. should be. By that type of calculation, we are indeed stunningly far below where we should be, perhaps 5.5 per cent below, using data through the fourth quarter of 2011.3 What is more, we have made little progress in closing the gap defined in this way, because real GDP has only grown at modest rates since the recession ended in the summer of 2009. And furthermore, using current GDP forecasts from, say, the Blue Chip consensus, we have little prospect for closing the gap any time soon.
Is this really the right way to think about where the U.S. economy should be? I do not think it is a defensible point of view. Let me give you some of my perspectives.
Most analysts seem to agree that the middle part of the 2000s was characterised by a “bubble” in the housing sector. Housing prices were high and rising fast compared to nominal GDP. It is not prudent to extrapolate a bubble into the indefinite future and claim that such a calculation provides a good benchmark. Yet, that is what we are doing when we extrapolate fourth quarter 2007 real GDP.
Furthermore, we normally have the good sense not to do this in other economic situations.
Think about the tech bubble of the late 1990s and early 2000s. The NASDAQ index peaked at about 4800 on a monthly average basis in March 2000. In the month just passed, January 2012, it averaged about 2744, which is 57 per cent of its peak value. In fact, with the type of calculation usually done for real GDP, which extrapolates growth rates, the NASDAQ would be miles below its “potential.” Of course, most observers and market participants do not say that the NASDAQ is far below its potential value today. Instead, most say that there was clearly a bubble in the NASDAQ during the late 1990s and early 2000s, and that today’s valuations are more sensible than the ones that were in vogue during the bubble period.
When I put the case this starkly, I think many would agree that establishing a benchmark by extrapolating from the previous peak of a variable, when that variable was clearly influenced by a bubble, is a mistake. It gives a distorted view of the situation.’
Again, this sounds right: It doesn’t make sense to extrapolate from the peak of a bubble.
There’s just one problem: There doesn’t appear to have been much of a bubble at all in the mid-2000s.
Here’s a look at annual GDP growth going back a ways. Notice anything? The growth throughout the 2000-era bubble was really mediocre by historical standards. Certainly it was far worse than the ’90s. It doesn’t look on the surface like this was some go-go bubble time.
But on the other hand, the housing bubble does seem to have been pretty real.
Here’s GDP vs. annual house price growth. House prices did grow very rapidly in the 2000s.
Now here’s one more chart to muddy the picture a little bit.
It shows GDP vs. house prices vs. annual growth in retail sales.
As you can see, retail sales—the green line—did grow faster than GDP during the 2000s, but it wasn’t growing as fast as home prices were, and it actually grew slower than during the ’90s. So it’s actually hard to argue the conventional wisdom, that homeowners, were going crazy on credit cards and their household ATMs buying flat-screen TVs and everything else under the sun. Basically, nothing looks abnormal on either spending or GDP growth during that time. If anything, it was pretty subdued.
The bottom line: Bullard’s argument that we don’t want to peg our expectations to 2007, because 2007 was such a craaaazy time does not seem based in fact.
But one thing we’re curious about is: OK, why wasn’t there a bubble? Why didn’t GDP and spending grow alongside the booming home prices?
Because it’s not like the home price boom didn’t have an effect.
Check out this chart of growth in household balance sheets.
As you can see, household balance sheets DID grow like crazy during the 2000s.
It’s just that, again household balance sheet growth was way faster than actual GDP growth.
So two takeaways:
- Growth in the mid-2000s doesn’t seem so wild as to believe we shouldn’t aim to re-achieve that trendline.
- The connection between bubbles (even big ones) and growth isn’t all that strong. We’re not clear why.
Any guesses on the latter?