I’ve commented earlier that many economic forecasts seem to assume reversion to the mean – here, meaning average economic growth over the last two decades. For a great example, go to the Wall Street Journal and admire the GDP growth rates projected for Q3 2009 through Q2 2010, marching happily up and to the right. (The numbers are 0.6%, 1.8%, 2.3%, and 2.8%.) This recession is different, however, and even if there is a mean to revert to after U.S. households decide how much they want to save, there’s no telling how long it will take.
For one perspective, the Carnegie Endowment for International Peace had a session at the end of April featuring a few IMF economists. Marco Terrones (link to PowerPoint at the bottom of the page) looked at the typical duration of a recession and the ensuing recovery. The duration of recovery is measured to the point at which the economy reaches its previous peak output (the output level when the recession began – December 2007 in our case). He looked at 122 recessions since 1960.
Terrones’s key point was that recessions last longer, and take longer to recover from, if they are linked to financial crises or if they are globally synchronized, for reasons that are probably evident to our readers: popping credit bubbles lead to an increase in the savings rate, dampening consumption; and globally synchronized recessions mean that no country can export its way back to growth. On average (slide 5), a recession that is both linked to a financial crisis and globally synchronized will last three quarters longer than normal (seven rather than four), and recovery, measured from the beginning of the recession, will take seven quarters longer than normal (fourteen rather than seven). (There is obviously some correlation there, since the longer the recession, the more ground needs to be made up.) Those are just averages, but projecting onto our experience that means recovery would start in Q4 2009, but we would not reach our peak (December 2007) output level until late 2011. Since the workforce will have grown by a few percent in the interim, unemployment would not reach December 2007 levels – if ever – for another year or two.
Seven quarters of contraction and seven of recovery to previous peak don’t seem that bad. Of course, this is still just an estimate based on data going back to 1960, and there is nothing in that time period like what the world is going through today.
This analysis is covered in greater detail in Chapter 3 of the IMF’s April World Economic Outlook.