Arnold Kling writes

Keynesians think of economic activity as spending in the first place. You measure economic activity as spending by households plus spending by businesses plus spending by government, and then you net out imports and add in exports.

Suppose that we all work at the GDP factory making GDP. Unfortunately, spending is down, so some of us are laid off. The difference between the amount of GDP that the GDP factory could produce using all of us and the amount that is actually demanded is called the Output Gap.

The Output Gap looks like a $20 bill that is being left on the sidewalk. If somebody would just increase spending, it would help close the Output Gap, raising GDP and lowering unemployment. Stimulus certainly must work.

I am not comfortable with the GDP factory picture. We are not all interchangeable inputs producing the identical output. To me, economic activity is patterns of sustainable specialization and trade (PSST). Economic activity takes place only because we produce different things.

There are a lot of smart and clever things that can be said about patterns of sustainable specialization and trade. Arnold being clever and smart says many of them. However, the important question, the only question at the end of the day, is whether or not those smart things accord with cold physical reality.

Using MS Paint, I am going to horribly abuse this chart from Catherine Rampbell  to make a quick and dirty point.


It shows what small businesses answer as their single most important problem. As we can see that taxes and government regulation are perennial favorites. However, I want to isolate Poor Sales. I apologize in advance for the quality.


Lets match this against unemployment.


Unemployment rises when small businesses think sales are the biggest problem and fall when they think everything else is a big problem.

That is, of all the detriments to developing patterns of sustainable specialization and trade – and hence driving down unemployment – the ease at which businesses can sell stuff eats up most of the variance.

There are lots of clever ideas but the one that seems to accord with the data is pretty simple: business hire more workers when sales are good. They don’t when sales are bad.