Bryan Caplan  

Sumner's Common Sense

Arnold on the Current State of... David Brooks Repeats Himself...
In a post that officially attacks "common sense," Scott Sumner shows that he possesses a great deal of it.  Don't believe that nominal wage rigidity can explain continuing high unemployment?
[W]age stickiness is a much bigger problem when inflation is low, not high.  In 1982 all employers had to do to restore labor market equilibrium was grant smaller pay increases.  In contrast, consider the following example from George Selgin:
Here at UGA we haven't had any raises for five years running. I know professors elsewhere with similar experiences.
...So each year the UGA  administration sat down and discussed all the possible pay increases they could grant.  How about 2%?  How about 0.5%?  How about negative 1%?  And so on and so on.  Then it just so happens they decided on 0%?  What a coincidence!  The odds must be 10 to 1 against that particular number.  And then the next year it was zero again!  And the next year.  Now you are talking a 1000 to 1 shot.  And the next.  And the next.  Amazing coincidence?  Or money illusion?  I think you know where I stand, just from the title of this blog.  I don't know why people have money illusion, and it's certainly a bizarre thing to put into a model.  But it's there.


Money illusion is one of my few concessions to the behavioral economists.  I wish it weren't there--it makes our models much uglier and it causes enormous human suffering.  But wishing it weren't there doesn't make it go away. "The world is as it is."  Between mid-2008 and mid-2009, NGDP fell about 9% below trend, while wages kept chugging along at roughly trend.  That's a really big problem.  Although unemployment has since fallen from 10% to 8.2%, wages still haven't entirely caught up to that massive demand shock.

Sumner continues:

There's another problem with using common sense.  The real world is very complex, and no one model can explain everything. 

<Chandler voice>Can you get any more commonsensical?</Chandler voice>

Obviously the 40% increase in the minimum wage right before the recession slowed the downward adjustment in wages, especially for the lower classes where unemployment is concentrated.  People typically ridicule the minimum wage hypothesis by saying "Come on, that can't explain all the unemployment."  True, but no one claims it can.  Ditto for the extended UI benefits.  The natural rate of unemployment is estimated to be 5.6%, the actual rate is 2.6% higher.  That extra 2.6% is partly extended UI benefits, partly the big minimum wage increase, and partly sectoral reallocation of labor.  But I believe it is mostly wage stickiness, and the stock market also seems to think more NGDP would help a lot right now.

The W/NGDP ratio shot up in 2009, and unemployment soared.  Since then the ratio has come part way back to trend, and the unemployment rate has come part way back to trend.  We have excellent data showing George Selgin's example is common, wage increases bunch around zero percent.  Until we get a more plausible theory of unemployment, I'm sticking with stickiness.

I can hardly imagine a sharper contrast between Sumner's official message ("the pitfalls of common sense") and his actual common sense case for the power of nominal wage rigidity.  If only he'd titled his post "why nominal rigidity is intuitive" and linked to this!

COMMENTS (4 to date)

My experience with Scott is that his sense is most uncommon.

Doug writes:

Okay here's the thing about money illusion and policy targeted towards it. It's a self-fulfilling prophecy. Why do people have the sense, that in terms of pay, staying still is going backwards? Is that a sense that all humans, or even all sentient beings, in any conceivable culture would have?

No. It's a culturally trained response. Since the US de-pegged from Gold in 1933 the dollar has lost 94% of its value. Even in the "low inflation" period of the 21st century the dollar has lost 25% of its value.

Is it any wonder that people have a simple in-built heuristic that they need to increase their pay by 2% a year just to stay afloat? Now contrast it with an alternate world where fiat money was never tried and the US was still on a hard gold standard.

Essentially money supply is fixed in a gold standard (the rate of gold flows against gold stocks is well less than 0.1% a year). That means that 3.5% GDP growth translates into 3.5% deflation in a typical year.

The man on the street knows that if he can just KEEP his pay at the same level he's doing very well. Someone can comfortably start work at a salary and stay around there and end up quite successful by the end of his career.

If firms need to institute pay cuts it's not viewed with such revulsion. Because workers treat stagnant or increasing wages as a reward, not a right. In hard times workers just expect falling pay, just as workers now expect stagnant pay during recessions.

But if you're saying there's a tendency to a psychological anchoring on zero, this scenario is far better. Zero is far away from expected during hard times, so workers know their nominal salary has to fall. But the best part is in -3.5% equals 2% in real terms with hard money. So while there's a psychological anchoring on zero, there's no anchoring on anything in between -1% - -7%.

Keeping hard money removes zero from the realistic distribution during hard times. The new distribution has no similar anchorable number. This would give labor markers much more flexibility.

D. F. Linton writes:

Yes, people and companies are loath to adjust their prices downward nowadays.

How can it be that the very best solution anyone can think of is to inject large quantities of fiat money into the economy, changing relative prices with unknowable lags and variances, in the hope that this will improve the realism of the overall price structure? Isn't this the economic equivalent of kicking a broken television?

roversaurus writes:

My company had a couple of years with 0% raises ... Except once they suspended 401K matching contributes for over a year. And another time they changed their overtime policy so that the first 5 hours of over time were unpaid.

I suggest the sticky wage guy look harder. I doubt that total compensation (which is what the employer cares about) had the exact same percentage change year after year.

Comments for this entry have been closed
Return to top