Grog,

I’ll try to address your question after you address mine. Back in #70, you asked (and I think this was meant as rhetorical, not as an honest question, because I suspect you already think you have an answer to it),

Why do you sup­pose he [Krugman] used actual change in employ­ment, but used rate of change in wages?

In #74, I said,

I don’t know. Why do you sup­pose he did?

You still haven’t answered. Well, you gave a portion of a deflection (as I pointed out in #87), but you didn’t really address what “number of jobs losses” would mean, nor did you say anything at all about wage changes (up, down, actual, percent, or rate). In #92, you asked,

Why did Krug­man decide to use actual num­ber of job losses per sec­tor rather than a per­cent of change in each sec­tor?

… which is a different question again, unrelated to any of the questions we’ve asked each other so far. As I said, I’ll try to address your question after I understand what you think Krugman was trying to show by using the metrics he did.

Thanks.

Why did Krug­man decide to use actual num­ber of job losses per sec­tor rather than a per­cent of change in each sec­tor?

Because in rebalancing, you want to look at where the individual jobs are going. If you look at percentages, it tends to skew the relative job movements. That is, a loss of 1,000 jobs in a sector with 84% of all jobs would look effectively like zero, while that same 1,000 moving into the sector with 5% would be a significant improvement. And vice versa, of course.

Does that make sense?

Why did Krug­man decide to use actual num­ber of job losses per sec­tor rather than a per­cent of change in each sec­tor?  Espe­cially when one sec­tor employs 84% of all jobs and another employs 5%?  Why is actual losses more mean­ing­ful than per­cent of jobs in each sector?

Thanks.

Thanks.

“The dan­ger already exists that the math­e­mati­cians have made a
covenant with the devil to darken the spirit and to con­fine man in the
bonds of hell.”

Well, I certainly believe that. (I’ve always thought there was something devilishly fascinating about shortchain… )

  —  Saint Augustine

I’ll back up DC’s observations with these suggestions (and my take on them) for GROG, as a means of answering his question and educating the rest of us:

1. Did Krugman claim that employment would be down equally in all sectors?  (I say: no, he didn’t, just that all would be down.)
2. When there is not enough work for workers in some sector, over the short run (a couple of years), do companies reduce the pay of people in that sector, or do they reduce employment?  (I say employment, mostly.)
3. When there is a demand for people in some sector, do companies immediately go out and hire people, or do they instead post hiring notices and start competing for workers by raising the wages they offer?  (In most sectors you cannot simply go out and round up more RF engineers, more nurses, more heavy equipment operators — there’s an inevitable lag before people can be made available.)

IMO, this is why the “signature” of a structural imbalance in employment is the way Krugman says it is, and why you sometimes have to look at the derivative of a metric, not simply the increment.

But hey, what do I know, I’m just a mathematician/computer engineer with delusions of competence in my own field, not a trained economist.  If someone will explain to me why I am utterly wrong, I would listen quietly, if not graciously.

By the way, you only quoted a portion of Krugman’s position (or rather, you quoted one partial expression of it). Here is the full point:

OK, so what should be the signature of that story? Surely it is that job
losses should be concentrated in the bloated sectors, that employment
should if anything be rising elsewhere — and wages should be rising in
the unbloated sectors more rapidly than in the bloated ones.

Do we see employment rising in the under-allocated sectors? Do we see a significant difference in the rate of change in wages between the sectors you think are “over-allocated” vs. “under-allocated”?

Grog,

Thanks for responding. So, I get it, that you think we need fewer people who are reasonably paid, with benefits, building America’s infrastructure — but more burger flippers at minimum wage with no benefits. I get that.

And again, I thank you for responding. But you didn’t answer my question, and you didn’t address the point you made before. Let’s try again.

You had asked,

Why do you sup­pose he used actual change in employ­ment, but used rate of change in wages?

You wanted to make your own point, and you used percent of change in employment in those industry sectors (which is not the same as “rate of change”). But then you did not, in any way, address change in wages, whether actual, or percent, or rate, or any other related metric.

So, once more, to address the question you asked: What point do you think Krugman made by using actual change in employment numbers? What point did he make with rate of change in wages? And why would he use those different metrics?

I’m not asking what metrics you would have used to make the point you want to make. I’m asking why he used those metrics, and what point you think he was making (since that’s what you asked).

Then let’s go farther, and address the point you want to make. If indeed there is a case of “misallocation” going on, what would you expect to happen to rate of change in wages? What actually did happen to the rate of change in wages?

I don’t know. Why do you sup­pose he did?

Because he was trying to make the data appear to be something they are not. 

Krugman argues that the current unemployment problem we face today has nothing to do with a misallocation of jobs that is being corrected by the market.  He believes in the Keynesian theory the recession is caused by a slump in demand.

He correctly esplains that the the structural theory implies that if “we had too many workers in the wrong industries,  we have to expect a depressed level of overall employment as workers are moved out of these ‘bloated’ sectors.”

Krugman claims the change in employment has been about the same in construction, manufacturing, and services between 2007 and 2011. He concludes that the signature of a structural problem just isn’t there” and asks us to trust him as “a closer look doesn’t do much better.”

The problem is, if someone does take a closer look, they will see that the construction sector employed 5% of the American workforce.  Manufacturing employed 10% and services employed 84% of all jobs. If you do the math, jobs in the service sector was reduced by 2%, manufacturing fell by 15% and construction by 27%.  This is a bit different than what Krugman wants his readers to believe.

Not to mention that manufacturing has been losing jobs since the 1960’s.

Austrians would say that loose monetary policy and affordable housing policy caused the housing bubble and in turn a misallocation of resources in the construction industry. According to the BLS figures that Krugman used, construction saw a 27% decline in employment between 2007 and 2011 compared to 2% in the service sector and 4.5% overall.

This is what happens when we take the closer look that Krugman warned us about.