Real Economist Needed: New Keynesian explanation of unemployment confounds me

I have been going around on this one for some time.  The new-Keynesians’ (NKs) widely regard themselves as being a realistic improvement on their new-classical cousins.  One area in particular is said to be major part of their coup-de-grace.  Namely, the explanation of unemployment by recourse to the efficiency wage conjecture.  In short, firms pay an above marginal product wage to elicit the requisite effort of workers. This above marginal wage, in the aggregate, is the cause of unemployment.

Lets do a mind experiment.

Said worker is hired for $12.50 per hour.  The employer gets X work effort (we) which is less than their marginal product.

Said worker gets a raise of $2.00 per hour.  The employer gets X + we*+we** which equals the worker’s marginal product.

The only way the efficiency wage hypothesis can add anything to the neoclassical story is if workers, when hired, deliver a work effort below their marginal product.  Problem is, this would already explain unemployment.

Similarly, if the extra $2.00 per hour actually induces a work effort on the part of labour such that it equals their marginal product then unemployment can’t be explained by recourse to the efficiency wage premium because the wage equals the marginal product.

If, when workers are hired at the initial wage, the worker delivers a work effort at their marginal product then there is no need for efficiency wages:  efficiency wages are then irrational and the NKs have violated the terms of the peace.  That is, they have teased a sub-optimal outcome by recourse to irrational behavior on the part of employers.

So where is the savvy?


Update: Nick Rowe was generous enough to walk me through it.

My problem was that I was stuck on this paragraph out of the Stiglitz Shaprio (1984) article “Equilibrium Unemployment as a Worker Disciplining Device:”

To induce its workers not to shirk, the firm attempts to pay more than the going wage; then, if a worker is caught shirking and is fired, he will pay a penalty.  If it pays one firm to raise its wage, however, it will pay all firms to raise their wages.  When they all raise their wages, the incentive not to shirk again disappears.  But as all firms raise their wages, their demand for labour decreases, and unemployment results.  With unemployment, even if all firms pay the same wages, a worker has an incentive not to shirk.  For, if he is fired an individual will not immediately obtain another job.  The equilibrium unemployment rate must be sufficiently large that it pays workers to work rather than to take the risk of being caught shirking (p.433).

This is the paragraph I had in mind when I wrote the original message.  Nick did me the favour of transposing it onto the supply-side… and then… the light:

Here is how I would write it:

Suppose we start at zero unemployment, supply=demand for labour. One firm realises that if it raises its wage above other firms’ wages, its workers will not shirk. Other firms realise this too, so each tries to raise its wage above other firms’ wages. They cannot succeed at this of course, but the result is that wages rise. As wages rise, quantity of labour demanded falls, and quantity of labour supplied rises, so there is an excess supply of labour, or involuntary unemployment. Wages stop rising when unemployment gets high enough to deter shirking, without an individual firm needing to raise its wage above other firms’ wages.

This is I think a much clearer presentation.  I have several residual doubts about the logical consistency of the explanation.  It seems to me that the EW conjecture relies on workers comparative evaluation of their own wage relative to the prevailing sectoral wage.  I get the Mdisutil side of the story (see comments section) that is, I get the supply side description but I still don’t get or don’t buy the demand side part of the argument.  Actually what I find fuzzy is the dynamics  responsible for firms demanding more labour than mp wages would warrant.  Alas, this will have to wait for another day.

As to the question of whether or not this NK conjecture is Keynesian?  I will just sketch my initial thoughts.  Part of the problem is that Keynes retained the neoclassical labour supply curve and the marginal disutility conjecture.  As Spencer* (2006, p. 467) points out, Keynes’ deployment of effective demand does not account for the origins of unemployment, but, rather, explains its persistence and does not necessarily point in the direction of a certain policy intervention because Keynes retained the crucial neoclassical equality between the real wage and the marginal disutility of labour.

It may be that the NKs are Keynesian but only because they retain the labour supply curve and fail to account for the origins of unemployment (as Nick might say (?): as we understand it when we ask the average unemployed person why they are unemployed).   But in this sense, the strict new-classical conjecture has the same problem: there is a theory of the volume of employment but not much else.  I think Spencer is right: if you retain the NC supply curve then not much can be added to the basic NC conjecture on unemployment: wages are too high relative to what their marginal product would warrant.  The NKs achieve the same result but through the argument that EW>MP wage and hence calls forth too much labour supply resulting in involuntary unemployment.   But at the end of the day wages are too high to clear labour markets.  It strikes me that the essence of Keynes GT was an attempt to escape this logic.

*Spencer, David A. “Work for All Those Who Want It? Why the Neoclassical Labour Supply Curve Is an Inappropriate Foundation for the Theory of Employment and Unemployment.” Cambridge Journal of Economics 30 (2006): 459-72.


6 thoughts on “Real Economist Needed: New Keynesian explanation of unemployment confounds me

  1. lol, so its totally not the completely obvious price floor of the minimum wage. Totally not the answer staring their retarded Keynesians asses in the face.

  2. Mick,

    First, most models assume no minimum wage. In the Stiglitz (1984) article they make the argument that employment insurance needs to be set low enough so that unemployment hurts and so that employers are not forced to pay an effort premium. But that does not solve the contradiction. If workers shirk from the outset and wages are higher than the marginal product would warrant then we do not need efficiency wages to explain unemployment because by definition the initial wage is already too high. The simple NC story already said access to non-market based wages (effectively a minimum wage) meant that employers paid above marginal product wages. The efficiency wage conjecture in this sense is superfluous.

    Second, as you might guess from 1, New-Keynesians are Keynesians in name only. Go read Mankiw if you think I am lying.

    Third, “staring their retarded Keynesians asses in the face,” is at the very least a mixed metaphor.

  3. I should add that for Stiglitz (1984) the implied minimum wage meant that employers would have to pay more than they usually would so that the threat of the sack acted as a real deterrent to shirking by workers. So that for Stiglitz the issue was the level of the unemployment rate but not unemployment itself. But if you stare at it long enough you will see such an explanation also suffers from a similar problem.

  4. Hi Travis:

    I can maybe help with this one.

    First, the theory of efficiency wages in a shirking model was first developed by two Marxians (neo-Marxians?) Herb Gintis and Sam Bowles (It’s a great paper). The new keynesians only picked up on it later.

    In an efficiency wage equilibrium, the wage is still equal to the MP of labour, but it’s above the marginal disutility of labour, that’s what differentiates it from the neoclassical model (and creates the involuntary unemployment).

    Compare Keynes General Theory, where he lists 3 “Classical” propositions. One was Say’s Law, a second was W=MP, and the third was W=Marginal disutility of labour (I may have the order wrong). Keys accepted W=MP, but rejected the other two. Efficiency wage theory also accepts W=MP, rejects W=Mdisutility, (and is silent on Says Law). So efficiency wage theory is compatible with Keynes GT.

  5. Thanks Nick for that and the off-line conversation helped. Will amend the post in the next couple of days.

    Still leaves the question of the compatibility with Keynes’ GT. Another post.

  6. You’re welcome Travis!

    Yep, I don’t think the “involuntary unemployment” in efficiency wage theory is compatible with “involuntary Unemployment” as defined in Keynes’ GT. But then Keynes’ definition was more peculiar/particular, and only made sense within the context of the GT (though it did make sense there).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s