An important new macro paper

When you disaggregate the data at the state level, wages don’t look so sticky any more:

…states that experienced larger employment declines between 2007 and 2010 had significantly lower nominal wage growth during the same time period…Our estimates suggest that real wages also vary significantly with local measures of unemployment at the state level…there is a strong relationship between local employment growth and local wage growth at business cycle frequencies.

In other words, the supply and demand model doesn’t do so badly after all.

So why do aggregate wages appear so sticky in the data?

…not because wages are sticky in the aggregate, but because different aggregate shocks have relatively offsetting effects on aggregate wages.

And in conclusion?:

…we find that a combination of both “demand” and “supply” shocks are necessary to account for the joint dynamics of aggregate prices, wages and employment during the 2007-2012 period in the US…

That is a new NBER Working Paper from Martin Beraja, Erik Hurst, and Juan Ospina.  Here are ungated versions.


Of course, we already knew that the great recession affected different geographic regions differently. And it's not surprising that wages in the regions experiencing the worst of the recession would be the least sticky. That was compounded in this recession by the housing crisis, which discouraged labor from moving to more prosperous regions (because houses were unsaleable). There's always Texas. December's data reflects slipping mining and manufacturing sectors but construction holding its own. What will the data reflect in six months if oil continues to slump and if construction follows. Florida's economy is similar in that the largest sector of the economy is . . . . growth. That sector was hit hard by the housing crisis - because people were unable to sell their houses and move to Florida. Recovery of the growth sector is reflected in construction, where jobs are up 7% in December. If you build it in Florida, they will come. The joke about the U.S. economy before the great recession was that Americans prospered by selling houses to each other. That's reflected in the number of additional real estate agents in Florida in 2015: more than 29,000.

Who ever thought wages were sticky upwards?

From Figure 6 in the ungated paper, I count 8 states with falling nominal wages over that period. This suggests they aren't completely sticky downwards.

Nevertheless, the point still stands, why didn't wages fall more in more states?

Is that due to above average income earners losing jobs at an above average rate, or due to wages falling for the same work?

Tim Taylor does the work - breaking down U.S. inflation rates by category (many categories).

"a combination of both “demand” and “supply” shocks are necessary to account for the joint dynamics of aggregate prices"

Wow. Who could have guessed? There is a relationship between demand and supply reflected in prices? Better get started rewriting the textbooks.

As you point out they found that nominal wage growth responded in line with the shocks. Bigger shocks had a more severe impact on wage growth than smaller shocks did. No problem with that.,

But they did not find any examples of actually falling wages.

So the paper seems to support the Keynesian analysis that wage growth is extremely sticky at the zero lower bound.

Am I reading this wrong?

Not sure why it would be implied that the zero lower bound is relevant here.

Honest question:
Would it be reasonable to predict that falls in nominal wages are partially caused by greater layoffs at the higher end of the wage spectrum in 08-09? I'm not actually sure at what level the theory of downward sticky wages is supposed to operate–individual or aggregate.

So what were the supply-side labor shocks?

Does the paper take into account lower overtime? Most papers showing wages aren't sticky really are showing effects of going from 50 hours a week to 40 hours a week. The average wage goes down without time and a half. The core wage is still sticky.

Also, an annual income may not be sticky if employer implement furloughs. INCOME goes down but WAGES stay flat.

Does disaggregated flexibility imply aggregated flexibility? Here are two who disagree:

Comments for this post are closed