Breaking (not so) Bad: Why monthly job gains below 200,000 are OK.

July 7th, 2016 at 6:00 am

This Friday morning at 8:30, the Bureau of Labor Statistics will release the jobs report for June. Those of us who scrutinize such things will be watching closely to see if there’s a bounce-back from last month’s dismal count of 38,000 jobs. The consensus expectation is for 180,000, but last month, the consensus was for around 160,000, so you’ve got to take these monthly guesses with a big grain of sodium.

If you take an average of the monthly job gains from 2012 forward, you come up with about 200,000 jobs per month, and many job watchers got used to that number, such that months that came in under 200,000 were considered disappointing. But while we’re not yet at full employment – and obviously, the more jobs the better* – payroll numbers below 200K, say in the 100K-150K range, still represent a solid job market.

To see where I’m coming from, we need to talk about the “breakeven level” (BL) of job growth. The BL is the number of payroll jobs consistent with a steady unemployment rate. It’s a pretty straightforward calculation but it’s dependent on a few assumptions. For example, every BL is specific to a given unemployment rate and labor force participation rate (LFPR, or the share of the adult population that’s either working or looking for work).

These assumptions are necessary because having more people in the labor force implies we’ll need more jobs to employ the new entrants; similarly, a lower unemployment rate means the BL must account for new job holders.

The box below shows the monthly average BLs under four different scenarios for the LFPR and unemployment rate by late 2018: 1) both variables hold constant at today’s values, 2) unemployment falls gradually to 4 percent and the LFPR holds constant, 3) unemployment holds constant and the LFPR gradually rebounds by a percentage point, and 4) the unemployment rate falls and the LFPR grows by the amounts assumed in the earlier scenarios.


Before we get to the BLs themselves, let’s noodle over those assumptions for a minute. Most economists still think an unemployment rate of around 5 percent is the lowest rate consistent with stable inflation. But at this point, that’s more force of habit than rigorous analysis. After all, unemployment’s fallen sharply over the last few years while inflation’s hardly budged. So I can easily defend shooting for 4 percent unemployment while keeping a close eye on inflation.

The LFPR is trickier. It’s down more than three percentage points from its pre-recession peak, but part of that decline is due to retiring baby boomers. It’s hard to say how much of those three points we could get back in a truly full employment economy, but I’d argue at least one point is up for grabs.

Now, let’s look at the breakeven box. It really reduces to a simple point: given the progress we’ve made in the job market and the growth in our working-age population, monthly job gains of roughly 100,000 aren’t cause for alarm. While much lower numbers, like May’s initial count (we’ll see if it gets revised on Friday), are unacceptably low, we can still make with payroll gains below 200,000.

That doesn’t mean, of course, that we—and by “we” I mean the Fed—should be constrained by these BLs. The underemployment rate, at 9.7 percent, is still at least a point too high, elevated by millions of part-timers who seek full-time work. And though the overall LFPR has fallen in part because of retirement, it’s depressed for “prime-age” workers (25-54 year olds) as well. That’s partially a long-term trend, too, so it would be hard to achieve past peaks on that measure, but there’s definitely room for growth there. Faster job growth will help tighten things up sooner rather than later.

With that background, tune back in on Friday and we’ll see how we’re doing relative to these benchmarks.


*That’s not obvious if you’re an inflation hawk, but the correlation between the tightening job market and price pressures has been persistently very low.

Bond yields are just so damn low…what is that telling us?

July 6th, 2016 at 8:51 am

I pour the morning cup of mud, schlep out to the stoop to get my paper, and open my WSJ to learn that the yield curve is awfully flat (i.e., the difference between the interest rates of bonds of different maturities is low). In fact, the yield on the 10-yr–1.367% yesterday–is the lowest on record. And the difference between the 10 and 2-yr Treasury yields, at about 0.8, is the same as it was in November, 2007.

Source: WSJ

Source: WSJ

Not good, and very much worth noticing. There are those who will tell you the yield curve is a revealing recession indicator and that a narrow 10/2 year spread is flashing red. But now isn’t late 2007. Most importantly, of course, the housing bubble was imploding back then, whacking bank portfolios, wiping out trillions in wealth, and generating both a credit freeze and a demand-killing negative wealth effect. Also, the Fed funds rate has been at or near zero since around 2009, so that’s also very much in the mix here, holding down Treas yields.

One must be steely-eyed in these circs and avoid unnecessary bedwetting. On the negative side, along with the flat yield curve, we have:

–the strong dollar: it hurts our trade competitiveness and puts downward pressure on inflation and thus props up the real interest rate;

–volatile markets/uncertainty: surely a bigger problem for the UK than us/US, but watch our equity markets for wealth effects; still, credit flows here seem largely untouched;

–possible weakening job growth: that’s a big one of which I’ll have more to say later this week.

On the plus side:

–low unemployment, a bit of wage growth, and low inflation, while a problem re propping up real interest rates, is helping paychecks go further;

–growth: kind of the bottom line here, and while it’s been bumpy, we’re basically posting growth rates of around 2%, yr/yr, which ain’t great (and reflects a nasty downshift in productivity growth) but is actually the envy of most other advanced economies right now.

Two other things. I suspect the probability of a near-term recession remains relatively low here, well below 50%, though who knows? But what I’m absolutely certain of, as I wrote here, is that we’re not ready for it, either re monetary or fiscal responses.

Finally, I couldn’t be more flummoxed by this next point: these historically low Treasury rates amidst sluggish growth and an engine of job growth that could be downshifting are absolutely SCREAMING for policy makers to borrow and invest in public infrastructure. That would help on many levels now, from labor demand to productivity.

I won’t make free lunch arguments, but given these yields and the potential benefits of the investments, this is lunch with a very large discount. I realize I’m screaming into the gridlocked, political void, but scream I will. And you should too, IMHO.


Productivity mismeasurement: it goes both ways

July 4th, 2016 at 10:59 am

Many economists, myself included, argue that among our greatest concerns is the slowdown in productivity growth, as the growth of output per hour (that’s how productivity is measured) is a key determinant of living standards (yes, the distribution of productivity growth is another key determinant; in our age of inequality, growth is necessary but not sufficient to raise middle class living standards and lower poverty; but it sure is necessary!).

Some critics of the slowdown hypothesis argue that we are undercounting output, and thus systematically undercounting output per hour. I and many others find this hypothesis wanting, but an article I read in the NYT this AM got me thinking about a corner of this debate that tends to be overlooked: virtually every mismeasurement argument focuses on how technology is making us better off that are not counted in the national accounts, but some technologies push hard in the other direction.

The article focuses on tech support, which is often not only unbearable and enraging for users trying to find out how to restore their files after their cat erased them, but can be deliberately set up to be so, in order to save costs and discourage their use. One could say the same thing about phone menus in general. They are typically one of the many ways technology is used to externalize labor functions that were formally internal, which is a cost shift onto those of us endlessly pushing buttons in hopes that maybe we’ll find a person, and maybe that person will deign to help us.

If we were accurately measuring the output of tech service industry, such inconveniences would score as a negative.

There are other ways in which we fail to capture quality deterioration in our national accounts. Air travel is often raised as a poster child. Infrastructure deterioration is another. If you try to commute into DC, where parts of the Metro are shut down making our already terrible rush hour traffic even worse, you see an electronic sign over the highway that “helpfully” says, “Rethink your commute.” Perhaps I lack imagination, but I’ve rethought it, and all I can come up with is driving or taking the Metro.

End of the day, technology probably provides us with more mismeasured good stuff than bad stuff. To be clear, and this is very important, to make the case that productivity growth is faster than we think, you have to show that mismeasurement has worsened, and there’s little evidence to support that claim. In fact, there’s some to the contrary–we’re actually doing a little better in capturing tech’s benefits, which sadly implies the slowdown in productivity growth might be even a little worse than we thought.

But anyone seriously considering this mismeasurement hypothesis must consider both sides of the equation. There’s no point in denying the existence of the often hellish worlds of tech support and phone menus.

We’re not ready for the next recession

June 30th, 2016 at 1:07 pm

I know of two policy channels through which to push back on recessions: monetary and fiscal. As I discuss in today’s WaPo, though who knows when the next downturn hits, we’re not ready.

Monetary policy–ie, the Fed’s main tool: its Fed funds rate (ffr)–is very unlikely to have time to reload. I think this is a pretty striking picture of the extent of ffr reductions in past cycles of monetary stimulus.

Source: Federal Reserve

Source: Federal Reserve

But there’s still fiscal policy, right?? See, Congress, US, under subheads: a) dysfunction/gridlock, b) denial of efficacy of fiscal stimulus, c) embrace of fiscal austerity.