Trend Spotting

Go on over to Barry Ritholtz’s Invictus at The Big Picture’s post on GDP and take a look at the first graph. It’s from the invaluable FRED database of the St. Louis Federal Reserve and juxtaposes real gross domestic product from 1990 to the present with a purported potential real gross domestic product.

Leave aside that GDP is a construct (by construct I mean an artifact that may or may not have a real world referent; its relationship to a real world referent is assumed) as is real GDP, derived by adjusting the calculated gross domestic product by an imputed rate of inflation. Potential real GDP is, apparently, GDP if it increased at a fixed annual rate.

IMO there is very good reason to believe that potential real GDP is hooey—something with no real world referent. To believe otherwise is to believe that misallocation of resources has no effect on the economy.

Consider real GDP during two different periods on the graph: 1994 through 1996 and 2000 through 2003. Rather clearly real GDP during those periods grew more slowly than the presumed trend. Each of those periods was succeeded by a bubble, the dot-com bubble and the housing bubble, respectively. The interpretation implicit in the potential real GDP story is that these two bubble were merely a return to trend.

IMO that flies in the face of the facts. Those two bubbles were specific responses to events and policies, in the case of the first bubble a change in the treatment of capital gains under the tax law and in the second Federal Reserve policies. Rather than returns to the previous trend they were artificial and non-sustainable attempts at subverting the new trend.

There’s a reason there’s a $1 trillion hole in the economy: it was never there to begin with. We just thought it was. We’re not as rich as we thought we were.


Take a gander at the updated real GDP graph drawn by regular commenter Andy. Note particularly the update—the green line. I understand why one would prefer to believe in the potential real gross domestic product as calculated. I’m not convinced that it describes anything that’s not an artifact, is realistic, or sustainable.

7 comments… add one
  • Andy Link

    I take it this is what you mean? The green line is a trend I made with nothing but my Mark I Eyeball.

  • Thank you, Andy. I thought of doing that but you’ve spared me the trouble. That’s exactly what I meant.

  • Or even shorter…what really is the true potential GDP. It is an unobservable number and we can only try to infer it via historical data that needs to have various extraneous factors removed. That is nobody can point to either the green or red line and claim that is truly potential GDP.

  • Steve V., my gripe is that whether you’re talking about potential real GDP, real GDP, or aggregate demand it’s denominated in dollars. If the value of a commodity drops sharply (as it will when a bubble collapses), that putative value is simply gone. It’s not replaced with something else. It’s gone.

    The value of housing has dropped (and IMO will continue to drop for several more years). That value is just gone. There is no there there.

  • FWIW, I have no problem with your analysis or your hypothesis; you might very well be correct. I guess time will tell whether we can step on the pedal and close the presumed “gap.” My work (usually) seeks to present the facts at hand, as with the GDP/GDPPOT charts. Whether the downshift we’ve experienced changes everything remains to be seen.

  • That’s the core of my point, Invictus: GDP and, especially, GDPPOT aren’t facts. They’re calculations based on assumptions. If the assumptions are wrong or the calculations are wrong, they have no referent. They’re not just wrong. They’re meaningless.

    Imagine a magic widget-producing machine. It produces widgets, so many an hour, without any inputs. The machine can’t be convert to any other purpose or even sold for scrap. All it can do is produce widgets.

    Assume that people buy these widgets as fast as the widget-producing machine can make them. As long as people want widgets and are prepared to pay for them with money (or with credit as frequent commenter Ben Wolf might remind us), the widget-producing machine is a real money maker. What it adds to GDP is what people pay for the widgets it produces.

    If people suddenly don’t want widgets any more, what does the widget-producing machine add to GDP? Zero. GDP is therefore reduced from the time that people wanted widgets by the amount they were willing to pay for widgets before they stopped buying them.

    That’s true whether the machine continues to make widgets or not.

    In other words in my example there is no such thing as potential GDP.

    In the real world it’s a lot more complicated than that. The GDP-producing machine that is the economy does have inputs and, when people stop buying some or all of a component that once added to GDP, over time some of inputs of that component may be turned to other uses.

    The way to facilitate that process is to reduce the barriers that inhibit the conversion process. Waiting for people to start buying the component again is nuts.

  • Fascinating to see that Paul Krugman just posted on this very topic:

Leave a Comment