Jobs report: There’s still room to run in this job market!

March 9th, 2018 at 9:29 am

Payrolls rose 313,000 last month, well above expectations, and the unemployment rate held at 4.1 percent, as wage growth moderated a bit  from last month’s pace (up 2.6 percent, yr/yr). Though these monthly data are notoriously jumpy, the out-sized job gains were accompanied by a nice pop in labor force participation rate–up 0.3 percent to 63 percent–suggesting that the hot labor market may be pulling new workers in from the sidelines. If so–if this turns out to be more of a trend than a blip–this has important, positive implications for the increased “supply-side” of the economy, implying more room-to-run than many economists believe to be the case.

This was the first over 300K month since July 2016, and the jump in labor force participation comes after the rate was stuck at 62.7 percent since last September. However, both of these values jump around and so our monthly smoother provides a look at the underlying trend in job growth by taking 3, 6, and 12 month averages of the monthly changes.

This month, the smoother tells a surprising story. Typically, as economies close in on full employment, we expect the rate of job gains to slow, as the labor market nears its capacity. But the smoother shows the opposite pattern, that of a (slightly) accelerating trend. For example, over the past three months, the average monthly job gains come to 242,000, but over the last 12 months, they amount to a lower 190,000. We should be careful not to over-interpret even these smoothed numbers, but the punchline is that it’s hard to make a case that employment growth is DEcelerating, as would be the case if the economy’s water glass, if you will, were full to the brim.

This question of how much room-to-run exists out there is leading, as it should, to close scrutiny of wage growth for signs that job market pressures are pushing up paychecks. The figures below show yearly hourly wage growth (along with a smoother trend) of both all private sector workers and the lower paid 80 percent who are blue-collar or non-managers. The wage pop that spooked markets last month (Jan17/Jan18) was revised down slightly, from allegedly scary 2.9 percent to 2.8 percent. As noted, this month’s wage pace slowed a bit to 2.6 percent.

Again, the smooth trend (6-mos average, in this case) in wage growth deserves a close look, and it shows remarkably little acceleration given the persistent tightness of the job market. I’ll discuss why that may be occurring in a moment, but in fact, there’s a bit more there than meets the eye. Taking annualized growth rates of quarterly averages reveals more wage acceleration: 2.9 percent over the past three months compared to 2.5 percent last spring. This suggests perhaps a bit more pressure than the annual growth numbers reveal, which is, of course, what we’d expect at this point.

That said, there’s just no story right now, at least in the actual data (as opposed to expectations), of an overly tight job market leading to inflationary wage gains. The figure below shows some of the key indicators from the Fed’s dashboard, including unemployment, the Fed’s guess at the “natural rate” (the lowest unemployment rate consistent with stable inflation), actual inflation (PCE core, the Fed’s preferred gauge), and the Fed’s inflation target of 2 percent.

As you see, unemployment has been below the Fed’s “natural rate” for about a year (!) and both inflation and wage growth remain subdued. In other words, the links in the chain that go from a tight labor market, to faster wage growth, to faster inflation, remain uniquely weak.

There are surely many reasons for this, not all of which are understood. Productivity growth is lower, which tamps down potential wage growth. But also, worker bargaining power looks weaker than it should be at low unemployment. That’s partly because the unemployment rate isn’t telling the full story. Consider the prime-age (25-54) employment rate. Its peak in 2007 was 80.3 percent and its trough was 74.8 percent. Last month, it popped up three-tenths to 79.3 percent, thus it has recovered 4.5 out of 5.5 lost points. Given the population of these workers, each point amounts to 1.3 million potential workers added to the labor force. In other words, while there’s no question that there’s less slack in the job market, it’s very likely a mistake to conclude there’s no slack left.

A few more details:

–Warmer February weather compared to a harsh January may have played a role in the big jobs pop. The strong construction number (61,000 jobs added) may be evidence of that effect.

–Retailers (brick and mortar stores) got an off-trend-to-the-upside bump of 50,000. I suspect that trails off in coming months.

–Manufacturing, perhaps helped by the weaker dollar, continues to post decent gains, and is up 224,000 over the past year.

–Black unemployment, which spiked last month, fell back down to 6.9 percent, close to its historical low, suggesting the tight labor market is helping to employ minority workers. Still, the black rate remains close to twice the white rate.

This morning I read a pretty typical market take of the current job market. Economists at Barclays Bank wrote: “Looking ahead, we will view solid growth in employment less favorably.”

While I’m sorry in this analytic context for getting my class warrior on, why must Wall St. begrudge Main St. right now? The answer is high pressure job markets threaten wage gains which threaten both profit margins and inflation. Well, we’ve had too few wage gains relative to profits, and not enough inflation. So I, for one, will be looking at more reports like this one “more favorably.”

If we’re so productive in steel, why the persistent deficits?

March 7th, 2018 at 9:03 am

First, let’s get this out of the way: I’m certain the Trump tariffs will do more harm than good. But I’ve been trying to add a bit more nuance to the conversation than “trade war!” and “higher prices!”

It’s been clear forever that team Trump mistakenly views the trade deficit as a scorecard, one that’s not improved on their watch so far. Again, nuance is required. There are times when the overall trade deficit is a clear drag on growth, and times when the capital flows that support it are distortionary. But this is not one of those times, and targeting bilateral trade deficits makes no sense and can be counterproductive, as I describe here (and I’ll have more to say about this question of when trade deficits are problematic in coming days).

Still, Trump’s lousy tariff idea is surely motivated by our persistent deficits in steel and aluminum, a point I thought was missing from this otherwise useful article in the AMs WaPo on US productivity gains in steel.


The piece describes impressive productivity gains in steel production:

Labor productivity has seen a fivefold increase since the early 1980s, going from an average of 10 hours of work for each finished ton to an average of two hours in 2016, according to the American Iron and Steel Institute. Many North American plants were producing a ton of finished steel in less than one person-hour…

But if we’re so damn productive in steel, which should imply competitive pricing, why are we by far the world’s largest importer with persistent net imbalances? Why is so much domestic demand for steel met by imports? Obviously, price—but again, why?

It could be that other countries’ productivity gains in steel production have been greater than ours, or their labor costs are lower, i.e., our unit labor costs are not so competitive. But at least in broad manufacturing, that’s not the case–our ULCs are, on a dollar basis (so factoring in exchange rates), are below that of most of our trading partners, both in levels and growth rates.

Source: BLS

Certainly at the heart of the problem is China’s out-sized contribution to excess global capacity, which neutralizes the productivity gains documented in the WaPo piece (excess capacity is roughly unutilized production). This 2016 Duke University study (sponsored by the Alliance for American Manufacturing) gets right to the point:

The global steel sector is once again in a state of overcapacity. The sector, predominantly fueled by China’s expansion since 2000, has grown to over 2,300 million metric tons (MT) while only needing 1,500 MT to meet global demand. The result is a global steel sector at unviable profit levels and an influx of cheap steel in the global trading system adversely affecting companies, workers, and the global trading regime.

The first figure shows the Duke studies measure of steel capacity and production, along with the difference, which is overcapacity. The table below that shows production by country, wherein you can see the extent to which China is an outlier.

Source: See text.

Source: See text.

The table also shows how clearly Trump’s scattershot tariffs are not the solution; just look at Canada’s production! But the fact that Trump’s tariffs are the wrong solution does not mean there isn’t a problem!

College, wages, inequality, and the implied policy agenda

March 6th, 2018 at 12:20 pm

I stumbled on a number of facts today that put me in mind of the old, venerable debate among labor economists about the extent to which educational wage premiums driven by employers’ increased demands for skills is driving overall wage inequality. It sounds obscure, but it actually has very potent policy implications. Let me try to quickly break it down.

–The gap between high, middle, and low wages has grown a great deal since the late 1970s. No secret there. In real terms, this has meant long periods of wage stagnation for middle- and low-wage workers.

–At the same time, the earnings of those with college educations have grown a lot relative to those with terminal high-school degrees. Economists interpret this as increasing returns to “skill,” or more wonkily, evidence that technological advances, like computerization in the workplace, are increasing complementary to the skills college-educated workers bring to the job.

–The policy conclusion is “get more education!”

–But while I solidly subscribe to that admonition, there’s a ton more going on under the surface. For example, anything that drives down the pay of the non-college educated, who still account for about 2/3 of the workforce, also drives up the college wage premium. So, you must try to parse out declining union power, falling minimum wages, weak macroeconomies (the absence of full employment), and more, from the skill-demand part of the story. And this, of course, has major implications for the policies we pursue to close the gap and reverse the wage stagnation for the majority of the workforce.

–Moreover, and this part is particularly notable, the college wage premium, while as high as ever, hasn’t grown much over almost two decades. The figure below was in the WSJ this AM, in an interesting piece about how, given the rising costs of college relative to incomes, some kids and their parents are taking a closer look at alternatives like technical/vocational programs. The figure makes two important points: the premium is as high as ever (wage signal: go to college!), and it hasn’t risen since 2000.

Source: Census

The next figure relates to that stable trend, and it’s kind of the punchline, in terms of the facts. It’s from EPI economist Elise Gould’s recent comprehensive wage analysis (I’ve got a piece on her minimum wage by state findings coming out soon) and it takes some unpacking. The dark blue bars of the left show a measure of the growth in wage inequality: the yearly growth of the gap between high (95th percentile) and middle (50th percentile) hourly wages. The light blue bars on the right show the growth of the college premium, adjusted for a bunch of stuff including race, age, gender, etc. (see figure note).

Basically, the figure is showing you how much of the growth in the left bar might be explained by the right bar. From 1979-2000, the answer is “all of it.” Putting aside my important caveats that it’s not just skill premiums driving the college-HS wage gap, one can certainly look at those trends in the left two bars and call it an education story.

You can’t tell that same story, however, for the bars on the right. Though wage inequality by this measure has grown at a constant pace since 1979, since 2000, something other than the college wage premium mostly drove it up. I’ve given you some candidates above, and I’d add over this period: especially large trade deficits and the rise of the finance sector.

But the larger point is policy arguments like “it’s all about education” or “if only everyone got a college degree, these wage problems would go away” are simply not supported by the post-2000 data. To be crystal clear about this, higher education remains a critical way up the ladder of opportunity, especially for those groups that are under-represented among college graduates, and robust policies to help them must be a big part of the mix. But we must also pursue policies that strengthen labor standards, worker bargaining power, and full employment. In fact, this two-sided policy mix is the essential combination for pushing back on wage inequality and stagnation.

Trump’s tariffs

March 4th, 2018 at 6:29 pm

To an extent, I join with the conventional wisdom that Trump’s tariffs on steel and aluminum will do more harm than good, but if that’s where your analysis stops, you’re not going nearly far enough: At WaPo, with more coming tomorrow or Tues (with Dean Baker; and here it is).

Probably the most salient concern here is retaliation–ie, trade partners blocking our exports–though could be mitigating factors there as well. With 12% of GDP in exports, we’re less exposed to countervailing tariffs than other advanced economies. Also, to the extent that retaliation generates a GDP drag from larger trade deficits (think about that, Trump), the Fed could raise less quickly–or pause in their “normalization” campaign.

Also, here’s an interesting wrinkle. As I note below, most of our trade partners have good reason to object to the administration’s rationale (national security risk generated by diminished capacity in sensitive industries). But since, unlike team Trump, they’re likely to be more rules oriented, they might decide to take their case to the WTO, which takes at least six months to deal with such cases.

But while the Chinese dump steel below cost on global markets, most others (Canada, Brazil) do not do so, and we buy a lot more from them than we do from China. And there is no scenario I can think of wherein Canadian exports invokes “national security” risk, which was Trump’s rationale for this.

So do not confuse my attempt to see some nuance here with support for Trump’s actions.