Arnold Kling asks

Glaeser points out that it is young firms that create these new patterns.

look at the period from 1996 to 2008. Every year, the new firms added more than 2.9 million jobs, on net; every year except 2000 and 2006, the other firms, considered as a whole, destroyed jobs, on net. Similarly, the boom of the 1980s was led by job creation from new firms. Even in 2009, at the bottom of the recession, new firms managed to add more than 2.3 million new jobs–though those job gains were overwhelmed by the 7 million jobs lost by older firms. The lesson here: older firms generally shrink, while new firms erupt, hire new workers, and make up for the older firms’ job losses.

Does the AD/AS paradigm predict this pattern?

The short answer is: no.

The medium length answer is: it is consistent with AD/AS but not strictly necessary.

The longer answer is that I think if you take the underlying structure of AD/AS seriously then the economy’s amenability to churn affects the slope of the AS curve.

The more “dynamic” an economy is the flatter its Aggregate Supply curve will be.

The easiest way to think about this would be that excess cash balances are spent on “trying new things.”

When excess cash balances are high new firms have lots of new customers. They can then expand rapidly. When excess cash balances are low new firms have a harder time getting off the ground.

The more job growth is dominated by new firms the more pronounced this effect will be.

This would also provide a straight-forward reason why the tax multiplier seems to be larger than the spending multiplier even though basic Keynesian logic would suggest the opposite. When the government spends money that spending becomes concentrated into relatively few hands. Indeed, many of those hands will people who were previously unemployed.

Those households who were formerly unemployed now either stop radically dissaving if they were liquid or they expand out from sub-optimal consumption if they were illiquid. But, in either case they a relatively small fraction of their their new spending will be immediately directed towards growing firms.

Thus the supply response will be blunted.

On the other hand, however, tax cuts spread income into diffuse hands, many of whom were already fully employed and indeed facing lower than expected prices. These households are much more likely to spend their new income in growth industries and so the supply response will be much larger and faster.

Now to close this I think we have to assume that the response of growing firms to higher revenues is different at the margin than shrinking firms. So, it could be that the forces wash because slower decline boost unemployment just as much as faster growth.

However, here I think credit markets become key. A growing firm is more likely to be seeking to take on new debt. Higher revenues in the current period allow them to service more debt and engage in more rapid expansion. So, the spending effect is compounded by a borrowing effect.

This effect would only happen in old firms if they were attempting roll over short-term debt. However, most non-financial firms are not rolling over that much short debt. Thus, decreases in their ability to service their debt do not cause the amount of debt outstanding to shrink at an accelerating rate.

Though this could be one reason why financial firms respond so differently and potentially catastrophically to tightening money. Since they are rolling debt they can boom downwards through the same mechanism that new firms can boom upwards.