The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« Yet another cut at the recent retail price data | Main | Another side of the administration’s tax plan »

March 06, 2009

Dueling forecasts

When smart people debate, something interesting is bound to come of it, so I have been reading an interchange over the past couple of days in the blogs of Greg Mankiw and Paul Krugman. Krugman's blog provides the necessary background on the source of the debate:

"Greg Mankiw challenges the administration's prediction of relatively fast growth a few years from now on the basis that real GDP (gross domestic product) may have a unit root—that is, there's no tendency for bad years to be offset by good years later.

"I always thought the unit root thing involved a bit of deliberate obtuseness—it involved pretending that you didn't know the difference between, say, low GDP growth due to a productivity slowdown like the one that happened from 1973 to 1995, on one side, and low GDP growth due to a severe recession. For one thing is very clear: variables that measure the use of resources, like unemployment or capacity utilization, do NOT have unit roots: when unemployment is high, it tends to fall."

It is certainly true that when "unemployment is high, it tends to fall," but where it falls to is not always so obvious:


Prior to the 1973–75 recession, the average quarterly unemployment rate was 5 percent. If you had a forecast contemplating a return to "normal" following this particular recession you would have been holding your breath for a couple of decades.

Professor Krugman makes the central point, I believe, when he makes reference to the "difference between, say, low GDP growth due to a productivity slowdown… and low GDP growth due to a severe recession." That statement is, itself, recognition that the economy does periodically experience protracted episodes during which average growth and average unemployment simply do not revert to previous levels—at least not for a long time.

One of the striking things about the economic projections reported by the Reserve Bank presidents and Board's governors in the minutes from the last meeting of the Federal Open Market Committee was the rather large variation in views about GDP growth, even as far out as 2011:


That sense of uncertainty is shared by private forecasters:


What gives? There are lots of reasons for differences of opinions, and I obviously cannot (and should not) try to divine what is anyone else's deepest forecasting thoughts. But for me, "low growth due to severe recession" does not automatically imply a demand-driven downturn from which the economy will quickly spring back.

When I look ahead, I envision the U.S. economy over the next several years in terms of a simultaneous process of recovery and reformation: Recovery in the sense that the actual contraction of GDP will end, but reformation in the sense of structural transformation in financial markets, consumer behavior, and perhaps an adjustment of the global imbalances that are arguably at the root of much of the financial instability that has characterized the past decade.

If we are right, the long run is indeed rosy, but the long run will only arrive after some significant and protracted headwinds abate. And that is not a picture that suggests a rapid bounce back to "normal" growth.

By David Altig, senior vice president and research director at the Atlanta Fed

March 6, 2009 in Economic Growth and Development , Forecasts , Labor Markets | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Dueling forecasts :


That's very nicely, even gently and gentlemanly, put. Thanks for the clarification of something that's been bothering me. We are not only faced with the combination of a major recession inter-acting with a credit crisis (cf. Rogoff and Reinhardt) but the severity is triggering long-delayed structural adjustments in the socionomic system that we've been putting off for far too long. The chickens are coming home to roost and they're big, ugly and mean.

Posted by: dblwyo | March 06, 2009 at 05:58 PM

Most U.S. economic forecasters have plied their trade in the post 1980 world where asset, debt, and consumption growth have outstripped income growth. In the last 10 years non financial debt has increased $18tr
while national income has increased less than $6tr. The process is now running in reverse and has broad implications. The 8tr
in debt paydown/default needed to restore
the longer run debt/income ratio will be a headwind we face for years, not quarters. After-tax corporate profits increased from
a long term avg of 5.5% of gdp to 9% of gdp
during the debt expansion and are now headed
back down, perhaps for good. Business investment plays a key role in most economic
cycles while consumption is more stable. As
consumers repair their net worth through savings and debt repayment, the likely new lower level of consumption will surprise most forecasters. The cycle will steady itself when households have settled in at a lower consumption level that business can deliver at a 5% net margin. That is not in sight yet. The increase in govt spending and the expected improvement in net exports are small in relation to these secular changes.

Posted by: Dave A. | March 07, 2009 at 11:11 AM

"If we are right, the long run is indeed rosy, but the long run will only arrive after some significant and protracted headwinds abate."

I dont see any long run picture being rosy with the derelict energy policy we now have.
Lets not get distracted from that.

Posted by: retracer | March 07, 2009 at 12:08 PM

Interesting. Where does the USG regulatory and tax environment and consequent significant increases in the cost of doing business get factored in? Also, where does the likelihood of increased rigidity in labor markets due to [effectively] government mandated unionization of the U.S. workforce get factored in? The destruction of fossil fuel energy production for the chimera of "alternative" energy production? Etc., etc. The only outcome that seems assured as of today is an exponential increase in government involvement and control of U.S. business and markets. It seems to me that this might influence the dates and strength of "recovery" in the future.

Posted by: boqueronman | March 10, 2009 at 08:14 PM

Thank you for making the point that a reversal in the jobless rate need not mean that there is any particular trend to which the rate reverts. DeLong, in his first (second?) cut at Mankiw's unit root argument, seemed to imply that the jobless rate would return to some trend, so that the unit root question was not a big deal for real GDP. In comments, I made the point that the jobless rate didn't seem to have a stationary trend. No answer from DeLong (who tends not to respond to his comments section in a useful way). It is entirely an ego issue at this point, but I am happy to see I am not alone in seeing this point.

Posted by: kharris | March 11, 2009 at 07:40 AM

If Okun's Law holds, as Krugman seems to accept, the presence of a unit root in GDP says nothing about whether there's a unit root in unemployment. Okun's law says that the rate of unemployment depends on the change in GDP (this works best in log real GDP terms). If log real GDP has a unit root, its change will be stationary, so if Okun's Law holds, the rate of unemployment will be stationary. If we think of this as meaning that unemployment has a dynamically stable equilibrium, it still doesn't prevent the level of equilibrium unemployment from changing occasionally. Then it becomes a matter of distinguishing between those changes in observed unemployment which are due to changes in equilibrium unemployment and those which represent adjustment towards the equilibrium.
I would have expected that someone of Krugman's Keynesian leanings would actually tend to believe that there is a unit root in GDP. If there's no unit root, so GDP is trend stationary, the case for a stimulus program is much weaker.

Posted by: Brian Ferguson | March 11, 2009 at 12:07 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in