From Arnold Kling writing at The American:

Pump-priming and stimulus policies are a good fit for a manufacturing economy with homogeneous labor affected by temporary layoffs. They are not such a good fit for a post-industrial economy with an educated labor force facing permanent structural changes.

I’m adding this to my list of critiques of the Keynesian Multiplier.

Another problem with the multiplier is that we treat it as a fixed number. Even if we assume that it starts out greater than 1, no one talks about when it falls below 1. When it is above 1, every dollar we spend gets us more than a dollar in economic activity. If we can *actually* do that (and I have doubts in all but extreme cases), we should continue to spend until the multiplier falls to 1. When was the last time you heard about the marginal benefit of the multiplier?

My last problem with the multiplier is that we assume that deficit spending is automatically used for projects that will be effective. That just isn’t the case, partially because politicians are in charge of allocating spending and partially because it takes so long for projects to actually get going. For the latter, stimulus spending is sort of like wonder drugs that cure colds – you’ve got to wonder if the cold would have gone away even if you had just done nothing.

The next time you talk about the multiplier, ask the three following questions:

1) Are we still confident that spending is effective in a post-industrial economy, compared to manufacturing-based economies of half-centuries ago?

2) How much spending can we do before the multiplier falls below 1?

3) How confident are we that spending will actually be effective, meaning both targeted and timely?