Bryan Caplan  

Labor Hoarding and Labor Demand Elasticity

Salinas Event on Immigration... EconTalk and Chalk: Implementa...
Keynesian economists have long argued that firms practice labor hoarding during recessions.  In laymen's terms, employers often refrain from laying off workers even when - due to low demand - there is little or nothing for those workers to do. 

Why would firms willingly keep paying superfluous workers?  The standard story is that firms are planning ahead.  If they keep their workforce intact, they won't have to reassemble it after the recession ends.  But you could just as easily appeal to firing aversion: For purely psychological reasons, many employers are only willing to terminate workers as a last resort.

At the same time, Keynesian economists have also occasionally argued that labor costs are irrelevant during recessions.  Why?  Because "the problem is demand."  Or as Krugman puts it:
Um, we have a problem with demand, not supply; time to reread Keynes on wages.
If firms already produce more than they can sell, why would lower wages entice employers to hire additional workers?  This is basically a variant on the "Zero Marginal Productivity" theory of unemployment - but the problem is not that workers aren't physically productive, but that firms can't find buyers for workers' physical products.

Now suppose we take Krugman's story for granted.  As long as you accept the reality of labor hoarding, he's still wrong!  When firms weigh the burden of temporarily useless workers against the cost of rehiring after the recession is over, wages clearly affect the size of the burden.  Why wouldn't it?  Indeed, the tradeoff between retention costs and hiring costs plausibly leads to high labor demand elasticity even during severe recessions.  When cash flow is low, keeping a superfluous worker on the payroll at 80% of his normal wage sounds a lot better than keeping him on at full price.

I've repeatedly argued that free-market economists should take Keynesian macro more seriously.  But Keynesian economists need to lead by example.  Before they confidently declare that key elasticities are zero, they should consider the many subtle margins of the labor market.

Comments and Sharing

COMMENTS (6 to date)
Mike Mathea writes:

Bryan you are missing one component that actually came into play every time this issue presented itself.
Each recession I lived through in the corporate sector sooner or later someone asked what would happen to the firms State Unemployment tax Rate if we employees were laid off. Each state is different but in each case the issue is the same.
If you really the low demand is temporary why lay off workers and permanently increase your state tax bill.
Different answers depending on the situation but always the same question and analysis.

Eric Ulm writes:

Along another margin, it doesn't matter what businesses "want" to do. Holding superfluous employees will cause businesses at the margin to go bankrupt and all of its employees will become involuntarily unemployed. This will become even more pronounced during recessions. This may not affect labor elasticity by company but it does affect labor elasticity in the full economy, right?

Matthias Görgens writes:

Mike, that tax consideration would require coordination between companies; and would also only aoply to some jurisdictions. Labour hoarding exists outside the US as well.

Matthias Görgens writes:

Bryan, add principal agent problems to the firing aversion: bosses like to have underlings.

Matthias Görgens writes:

And an empirical book on the matter: Why wages don't fall during a recession.

Rob writes:

Could I offer another perspective on 'firing aversion'.

It is rare that all of an employee's duties are no longer required, so when a firm wants to fire an employee they also need to re-arrange their remaining organisation. Reorganisation requires detailed working level knowledge, which therefore requires worker cooperation. To get worker cooperation you need to pretend that there won't be job losses. You can then "change your mind" and decide to reduce headcount.

It is difficult to iterate this cycle many times whilst continuing to pretend that *this* reorganisation won't reduce headcount.


Return to top