What does "Keynesian" mean?

The title of this post is the title of a post by Jonathan Portes  at the Not the Treasury View blog.  One interesting point about his answer is that it all seems to boil down to your view on the effectiveness of fiscal policy - you're defined as a Keynesian if you believe fiscal policy works (to oversimplify his detailed argument to a ridiculous degree).  But isn't there more to it than just that policy argument?  After all, A.C. Pigou joined with Keynes in signing a letter to the Times (linked to here) calling for public works to counter unemployment, and Keynes, in the General Theory, treated Pigou (rather unfairly) as the representative Classical economist.



The next question, then, is: to be a Keynesian, do you have to accept Keynes' analytical apparatus - his basic model?  Because, contrary to what we're often told, Keynes had a very definite set of micro-foundations in mind. (There's a set of posts blogging their way through the General Theory going up at the Cocktail Party Economics blog, and those posts will apparently eventually be collected together at the Blogging Keynes blog.)

The world of the General Theory was a short run Marshallian world, stocked with profit maximizing perfectly competitive firms (Keynes doesn't seem to have been into that imperfect competition stuff which Joan Robinson was developing at about the same time as he was working on the General Theory).  Keynes parted company with Pigou and Marshall to large degree in his analysis of the labour market: in the General Theory, as in the classical model, a firm always operate on its Marginal Revenue Product of Labour curve (that goes with firns being Marshallian profit maximizers, and with the General Theory being set in the Marshallian short run), but while, in a situation of involuntary (Keynesian) unemployment the wage is above its equilibrium level, that is not the cause of the unemployment, rather it is a consequence of unemployment. 

We tend to teach Keynes as having a theory of unemployment based on downward wage stickiness -  demand in the aggregate falls and unemployment happens because the wage does not drop to the new equilibrium level.  That's not how Keynes saw it at all.  A lot of our "New Keynesian" type models rely on lagged adjustment of the wage to generate unemployment, but in them unemployment is a short run, transitory phenomenon, and the long run, dynamically stable equilibrium point is back at full employment.  But if you go back to the General Theory for your definition of Keynesian, those models aren't Keynesian, they're Classical with lagged adjustment.

That doesn't mean that you have to accept the details of Keynes micro-foundations to believe that fiscal policy can reduce unemployment, but it does, I think, mean that, if you reject Keynes' micro-foundations, you need to build and justify your own model - perhaps a classical one with very long adjustment lags and some labour market mechanism which makes fiscal policy effective.  But if you do that, are you still Keynesian or are you actually some flavour of new Classical?  Thinking of the way I've taught macro over the years I've probably been more the latter than the former.

Again, this isn't to argue that Keynes was necessarily right about things like the way the labour market works.  And in some ways perhaps it's a quibble.  But at the very least it makes for the basis of another lecture, one on the difference between Keynes' justification for fiscal policy effectiveness and someone else's, different model which leads to the same policy prescription.