David Altig has a nice chart comparing labor force growth in various industries. Lets give it a look and see what it is telling us.


So on the Y axis we have post-recession employment growth. On the X-axis we have employment growth over the mid-nils, the previous episode of expansion.

The size of the bubbles represents the number of folks employed in each industry. The bubbles themselves are labeled so you can see which industries we are looking at.

The 45-degree line represents points where growth (or decline) in the previous expansion and growth (or decline) since the recession would be equal. Bubbles below the line are industries which are growing slower (or declining faster) now than before.

Bubbles which are above the line are industries which are growing faster (or declining slower) now than they were before.

A couple of basic facts we can notice.

  1. Most of the mass is below the line. That means that most of the economy is performing worse than in was in the nils in terms of employment.
  2. Information is the only category in the third quandrant, meaning it was declining before and its still declining. This surprises most people. However, information is dominated by publishing and that in turn includes a lot of newspapers. That should make it clear why they are where they are.
  3. If you look at the slope of the bubbles, except for a few outliers they are not that far off 45 degrees. This means that while the rate of growth is slow, the relative growth rates have not altered that much. If you were slow before, you are likely even slower now.
  4. Manufacturing is the big outlier. Manufacturing collapsed in the nils but is stalled now.

Now what can this tell us about the general class of restructuring stories. First, it says that there hasn’t been much. In terms of employment the main restructuring has been from government to manufacturing. That is government is the biggest outlier below the line, and contains a hefty mass. While Manufacturing is the biggest outlier above the line and also contains a hefty mass.

Its true that this is happening. We are losing fewer manufacturing workers but we are laying off teachers. However, its not immediately clear why this represents a rebalancing that would produce recession.

Also, its hard to tell a story where this is the undoing of loose Fed policy. Indeed, it would be easier to tell a story where it was the undoing of excessively tight Fed policy. Loose Fed policy is going to tend to expand manufacturing both by lowering domestic interest rates and by lower the dollar.

The story that could make sense is that this was all the result of a change in China’s policy. China stopped manipulating its exchange rate. This relieved pressure on US manufacturing but also forced the US government to shrink as its main creditor pulled back.

However, that’s not exactly what happened either.

Overall I don’t know if there is much of a structural story to be told. I think the slowdown in manufacturing job loss is in large part because the shift to China played itself out, at least for this round. The decline in government jobs is because of State and Local balanced budget constraints.

Other than that you can see that even construction, everyone’s favorite culprit for misallocation, nearly lines up with education and health in terms of its distance from the 45-degree.

This is because the boom in construction just wasn’t as big as people think it was and because the contraction in construction began well before the recession started.