Tuesday, February 28, 2017

Another Clue About Labor Market Softness: The Decline In Single Quarter Jobs

One feature of the JOLTS data that has changed over the last several years are that both hires and separations (leaving a job for any reason) have declined. Together, these mean that workers are staying one place longer.

Some people start and end jobs in the same quarter. New research (here’s the working paper, since they want you to pay for the journal article) suggests that half of that decline in hires and separations is from a decline in those single quarter jobs. But there’s a problem:

[These jobs are] commonly held by younger workers and by those who take jobs at newer firms, but, nevertheless, the aging of the workforce and the decline of business startups explains relatively little of the decline in single quarter jobs. [emphasis added]

So what does cause the decline of single quarter jobs? It turns out they couldn’t find a decent answer.

The decline of single quarter jobs is not accompanied by individuals substituting long duration jobs for short duration jobs, but is merely part of an overall decline in jobs of both short and long durations.

But maybe the decline is a good thing:

We analyze whether single quarter jobs are “stepping stones” that allow individuals to move into longer duration and presumably higher paying jobs … We find some but not much evidence for this stepping stone aspect of single quarter jobs.

Interestingly, single quarter jobs are treated differently in the creation of two stylized facts about the labor market. One is that earnings have stagnated, and it turns out this stylized fact is developed by including single quarter jobs. The other in increasing inequality, and it turns out this stylized fact is developed by excluding single quarter jobs.

For earnings stagnation, it turns out that including single quarter jobs makes the average wage go down (no one is surprised that single quarter jobs are lower paying), but it also reduces the stagnation. This suggests that for workers that come and go — essentially, a spot market — the labor market is working just fine. Therefore, it’s not working just fine for long term employees, and combined with stagnating earnings, suggests that employers and employees are tied into too many poor long-term relationships.

For income inequality, the exclusion of single quarter jobs reduces measured inequality, but leaves it with an upward trend: this suggests that inequality is mostly about some people lucking into long-term jobs that are really good. Including them increases inequality (because single quarter jobs are mostly low end), but turns the trend in inquality downward. That’s what we’d expect if these low end single quarter jobs were going away.

Taken all together, this adds more evidence to the weird feature of the U.S. economy in the 21st century: people need to be more fluid … there really are more people stuck in ruts than before.

Food for Thought (Optional)

New colleague Paul Schneider repeated this example in brown bag seminar on teaching:†

Paul Schneider Brown Bag Capture

Honestly, I wonder how many adults would respond that this question is not answerable.

In doing quantitative work, particularly with undergraduates, I run into something like this. It’s a sure fire sign that a student doesn’t understand the problem their working with if they print out every possible digit produced by their computer or calculator. Rounding, I guess, is an indicator of comfort with your level of understanding.

† I believe the cite for this is Bransford, John D., and Barry S. Stein. "The IDEAL problem solver." (1993).

Technorati Tags: ,

Monday, February 27, 2017

Long Video for a Quick Idea

Cold Spring Shops posted this video from 1944 about how steel is made. It’s about 10 minutes long, and only the last minute or so is required for class. But, if you’re curious why steel production is mostly done in other countries, the first 9 minutes should make it clear why you wouldn’t want this sort of work.

Starting at the 9:07 mark though, the narrator makes a bunch of points that sound like they came out of Monday’s discussion of Ethan’s question in the Quodlibet on Canvas: at that time there seemed like there were few coordination problems, but people could remember when they’d occured recently, and the idea was dawning that those coordination problems were correlated geographically.

BTW: the volume is very low in the source. Not much you can do about that.


What the heck is infrastructure anyway? Maybe all you know is that politicians want to spend a lot of money on it.†

Part of that is or may be need. But part of it is also the Keynesian story that government spending is expansionary only when the money is spent on stuff that the private sector won’t pay for (rather than just sending out checks to people).

The Washington Post has an excellent article about this entitled “Six Maps That Show the Anatomy of America’s Vast Infrastructure”. There’s too much to copy, so you must click through to the source.

Here are the big 6 components:

  • The electrical grid is all the high voltage lines that connect power plants to your utility.
  • Pipelines connect our oil and gas wells with our refineries. The success of anti-pipeline protesters is hugely problematic, because the opportunity cost of pipelines is not “no pipelines”, but rather “more tanker trucks” and “more tanker rail cars” that can explode in accidents (and the article does a disservice to readers by only noting a rarer, and typically smaller, pipeline accident).
  • Railroads, and particularly the second map of railroad traffic.
  • Airports and flightpaths (that are monitored by air traffic control).
  • Ports. The article does not make a comparison, but what’s probably new to a bunch of students from Utah is the importance of barge traffic in the Mississippi River valley. You can get data on port volume here, and you don’t have to go very far down the list to find big ports that are not on the ocean. Foe example, Cincinnati is busier than Norfolk (the home of most of the Navy’s Atlantic fleet).
  • Bridges are included in the article, but I think this is a charismatic megafauna effect: when bridges collapse it’s big news, but it just doesn’t happen that often. Imagine the “charismatic megainfrastructure effect” if the Golden Gate Bridge collapsed in an earthquake.

† A position usually associated with economist Steven Landsburgh is that in democracies, the parties are supposed to disagree, so we should beware of things that the parties actually agree about.

Wednesday, February 22, 2017

Ken Arrow, R.I.P.

Kenneth Arrow, who I mentioned in my answer to Colin’s question in the Quodlibet, passed away yesterday. Here is the obituary from The New York Times.

Friday, February 17, 2017

Augmented Capital

What if capital owned the labor instead of the other way around?15-12-05; Dilbert; Capital Owning Labor

In growth theory, it’s called augmented labor, when we invent technology that allows a worker to control/use more capital at the same time. The robot above is capital, and it needs augmenting.

In this comic strip, that might be something like a way for a bunch of humans to nap with a minimal number of robots watching over them. Gee … sort of like a more benevolent vision of The Matrix.

This also illustrates growth and level effects, of technology. If the robots invented a technology that allowed them to control more humans, that would be a level effect that’s good for the robots. But having to deal with all those extra humans needing their nap would be the negative growth effect.

From the December 5, 2015 issue of Dilbert.

Thursday, February 16, 2017

Hans Rosling, R.I.P.

Hans Rosling died last week. He was a Swedish statistician known for his TED talks. I usually include at least one of them in this course, but usually later in the semester.

His stock-in-trade was debunking gloomy stereotypes about poor countries and economic development. There were five surprising facts, for instance, that he loved to hammer home: population growth is slowing rapidly; the divide between the global rich and poor is blurring; humans are living much longer than 50 years ago; many more girls are getting an education; and the number of people in extreme poverty fell by a billion between 1980 and 2013.

That’s from the obituary The Economist. Everyone should know those facts by heart (here’s a more detailed version, with a short video on the first point). It is deeply disturbing the extent to which otherwise normal people think life on this planet is getting worse. The broadest sustained improvement in the human condition has occurred during your lifetime. Denying that is a twisted pasttime that is all too popular. Fight it.

(I’ve been putting together a video answer to Pascal’s question for the class Quodlibet. It uses Lego bricks. I swear I was working on it before I saw this video from the obituary in The Guardian):

(In all honesty, lots of people use Lego bricks to make points about statistics, including my son in a 1st grade science project back in 2006).†

From the obituary in The New York Times, here’s Rosling on the magic that is all around us:

“My mother explained the magic with this machine the very, very first day,” he recalled. “She said: ‘Now Hans, we have loaded the laundry. The machine will make the work. And now we can go to the library.’ Because this is the magic: You load the laundry, and what do you get out of the machine? You get books out of the machines, children’s books. And Mother got time to read to me.”

“Thank you, industrialization,” Dr. Rosling said. “Thank you, steel mill. And thank you, chemical processing industry that gave us time to read books.”

This is Rosling’s most famous video (and also the one I usually require students to watch at home in April) about improved well-being around the world:

There’s also a shorter version of this one that I show in class.

Here’s a similar one, that’s a bit more about improving technology:

Here is a longer one about why we should not worry about overpopulation entitled “Don’t Panic — The Facts About Population”; you do have to view this one through the website of the enterprise he founded, Gapminder.

P.S. Rosling tweeted in 2010 that he’d noticed how average he was statistically back in 1972. Some of you are probably still idealistic, and certain you will turn out nothing like your parents. You probably would not even move the meter on that one when compared to me at your age. But how do things turn out? I have a 14 year old who thinks I’m sooooo old and out-of-touch because I’m 52. And yet in 1978 I was 14, my dad was 52, and I was certain he was sooooo old and out-of-touch. Turning out like your parents in a world with washing machines or Uber is actually pretty sweet.

† I would not differ from Rosling on what’s going to happen to the number of little black disks representing carbon dioxide, but I would question his implicit assumption that they matter much. I’m not an anthropogenic global warming denier; but I will point out that global temperature is really, really, inelastic with respect to carbon dioxide emissions, which in turn are really, really, inelatic with respect to people’s quality of life. That’s a recipe for worrying a lot more about the quality of life of poor people, and for worrying a lot less about temperature. Rosling is on the record agreeing with that conclusion.

Monday, February 13, 2017

Record Tax Revenue

Federal tax revenue over the first 4 months of the 2017 fiscal year† (October through January) set a record.

Big deal.

There are two things to consider here.

First, if tax rates are more or less constant, and income goes up, tax revenue will go up too. So we should not be surprised if tax revenue sets records when GDP is setting records. It’s just not that hard for a growing series.

Second, this is actually a good thing. No one like paying taxes. But if tax revenue is going down, and your policy hasn’t changed, something’s really wrong.

† Do note that this is a politically conservative site. It’s biased. But they’re just reporting numbers that are public information.

Wednesday, February 8, 2017

Unemployment Rate for January 2017

The unemployment rate went up a tad, from 4.7% to 4.8% in January.

This isn’t an official result, but my feeling has always been that no one can feel a change in the unemployment rate sharper than about 0.5%. So I view this uptick as … nothing at all, really.

Here’s the table of rates from the last 10 years:

17-02-08 Unemployment Rate Table Capture

You can see that we’ve been bobbing, mostly downward, through a 0.5% range for over a year.

I think we’re at full employment, or alternatively, near the natural rate of unemployment. That is not a solid number (it depends on demographics, and people’s self-definition of whether or not they’re looking for work or not). I usually assert that it’s like the mucky bottom of a muddy stream — not really very solid but firming up somewhere down there. For me, that’s about the 4-6% range.

You can see that in the two ends of this chart:

17-02-08 Unemployment Rate Chart Capture

Of course, some people might interpret any increase in the unemployment rate as a bad thing (It’s Trump’s fault! or It’s Obama’s fault!). Don’t take that seriously. If you look up in the table in late 2008 and early 2009 you can see that we were getting movements of 0.5% (the amount you can feel) every month or two. That was serious.

You can also see the asymmetry of the unemployment rate in the chart: it goes up faster than it comes down. This is normal. From the same site, here’s a chart of the unemployment rate over the last 70 years or so:

17-02-08 Unemployment Rate Chart 2 Capture

The asymmetry is fairly obvious across the entire period. There isn’t much we can do about that. But we do need to keep it in mind when we use the unemployment rate to evaluate policy: the rate not coming down fast enough in response to, say, Obama’s post-recession policies in 2009, or Bush’s in 2002, or Clinton’s in 1993, or Reagan’s in 1983 is normal and not their fault.

These charts do not mark NBER business cycle peaks and troughs. But, for reference, they were in December 2007 and July 2009 the last time around. Eyeballing the first chart and the table, the peak was still when we were in that soft 4-6% full employment range (although it had snuck up 0.6% since May of that year), and the economy’s trough was about 3 months before the unemployment rate topped out.

All of these tables and charts are from the Bureau of Labor Statistics (BLS) unemployment rate page. They’re site has a lot of slick tools for analyzing their data.