Jobs report: Another strong month as payrolls settle into solid trend, but wage growth still underwhelms

December 8th, 2017 at 9:31 am

The US labor market continues to add jobs at a strong clip, as robust consumer demand is generating a virtuous cycle of job growth, increased weekly hours, wage growth (though see ongoing caveats below), and hiring. Payrolls were up by 228,000 last month, above expectations for about 190,000. The unemployment rate held steady at 4.1 percent, a 17-year low.

Our official monthly jobs smoother filters out some of the noise in the payroll data by taking monthly averages over the last 3, 6, and 12 months. The fact that the bars are all about the same height reveals the underlying trend in job growth to be around 170,000 per month, a healthy pace for this stage of the recovery. As labor markets close in on full employment, job growth slows a bit as workers become more scarce.

However, given the movement of other key variables, namely, wages and prices, the full-employment, full-resource-utilization case cannot yet be made, as I show below.

Once again, hourly wage growth is a sore point. The two figures below show yearly wage growth for all private-sector workers and for the 80% of the workforce that’s blue-collar in goods production or non-managerial in services. In both cases, the 6-mos moving average reveal a flattening trend. Basically, in 2015-16, the tightening job market drove wage growth up from 2 to 2.5 percent, around where it has been stuck since. In fact, that’s precisely last month’s yearly growth rate for all private workers (2.5%). Given the consumer inflation has been running at around 2%, that’s but a small hourly wage gain in real terms, and certainly less than I’d expect given a 17-year low in the jobless rate.


That said, other series show stronger wage growth, as Dean Baker and I point out here, especially for middle-wage and minority workers, so the jury’s not in on the mystery of the missing wage growth. All told, in a series that mashes together 5 different wage series (which I’ll post later) we see a mild acceleration, but again, less than would be expected.

Also, hours worked per week ticked up last month and weekly earnings were up 3.1%, the strongest weekly growth rate since early 2011.

One theory is that the job market is pulling less skilled and experienced workers in from the sidelines, and this is holding down wage growth. But the Atlanta Fed tracker, which tries to control for this, doesn’t show much acceleration either.

All of this creates a challenging dynamic for the Federal Reserve. The next figure shows how the current unemployment rate of 4.1% is below the rate the Fed’s “natural rate,” i.e., the lowest jobless rate consistent with stable prices. But neither prices (core PCE, the Fed’s preferred gauge), which recently went through a bout of DEcelleration, nor wages, are responding much at all to low unemployment.

Are there, then, measures of labor demand that do not show a fully tight labor market? In fact, labor economists consider the employment-to-population ratio of prime-age (25-54) workers is a quick proxy for this question. This rate fell from about 80% before the last recession to 75% at its trough. It’s now at 79% so it has clawed back about three-quarters of its losses. In other words, it’s possible that the labor force still has some room-to-run and that labor demand, as strong as it is, still hasn’t exhausted labor supply.

Turning to sectors, manufacturing had a strong month, picking up 31,000 factory jobs last month. So far this year, manufacturing employment is up an average of 16,000 per month, compared to about zero last year. That’s partly due to stronger growth abroad pulling in more of our exports, but the dollar has gained strength of late, and that could dent these gains going forward, as the stronger dollar makes our exports more expensive relative to imports.

Retail trade stores added about 19,000, which is the biggest gain for the sector in a year. The strong job market could help the sector as the holiday season gets underway, but brick and mortar stores are of course in heated competition with online sellers. (Note: these data are adjusted for seasonal hiring effects, so any gains reflect jobs added above the usual seasonal pattern.)

Surely, some politician will say something foolish about how the solid November report reflects the tax cut that’s working its way through the Congress. If I could, I’d issue fines for such statements ($1.5 trillion sounds about right). As the smoother graph shows, we’re right on a trend that’s been ongoing for years now. As noted, there’s real momentum in the economy, with job gains, if not much in real hourly wage gains, boosting household incomes.

The challenge for policy makers, especially at the Fed, is to not get spooked by the strong quantity numbers (jobs) when the “price” measures–wages and inflation–are not flashing red at all.

Print Friendly

16 comments in reply to "Jobs report: Another strong month as payrolls settle into solid trend, but wage growth still underwhelms"

  1. Smith says:

    If the prime age labor participation is 1% below pre recession levels (79% vs. 80%) it’s worth noting that 1% represents close to a million people. With 90,000/month new jobs needed for population growth, the current 170,000/month new job creation trend would take another year to absorb them.
    Corporations would rather give up sales than see wages eat profit margins. G.E.’s announcement of an 18% cut in it’s workforce and decision to sell of profitable business units is part of this meme, although special factors motivated them to act now. The larger corporations put pressure on vendors to hold wages down. There is no competitive market for labor. Supply and demand affects wages but not in the conventional ways economists assume.
    I’d also add that post millennial trend of stagnant wages and slack in the upper reaches of job market, white collar and professional jobs, has a different effect than previous distribution where the same level of overall unemployment yielded greater wage pressure. Erstwhile higher income workers displacing lower income workers drags down wages for both college educated where a glut forces them to displace others, and in the labor market of the displaced. Thus college educated maintain the same low rate of unemployment, but missing any wage pressure. They no longer can drag lower incomes up with them either.
    How come you buy into higher wages increasing costs motivating productivity to offset costs, but never mention the route of higher wages creating more demand motivating need for productivity to meet supply. Is there data on this?

  2. cawley says:

    In trying to explain the lack of wage growth, I’m curious why you picked “near 80%” for your prime working age EPOP discussion. That was the number prior to the last recession. However, that was down from a sustained peak of over 81% – reaching nearly 82% – from mid-1997 through early 2001. That was also the last period of sustained real median wage growth.

    I recognize that you should be careful about using peaks and troughs in tracking cyclical measures however EPOP experienced a sustained rise from 1974 through 2000, interrupted only by recessions – in other words lack of supply of jobs, not lack of demand for jobs.

    The period from 2000-2008 was also characterized by sluggish real wage growth despite EPOP above 79% for all but 11 months – during which, it never fell below 78.6%.

    While it may be reasonable to conclude that the trends that drove that decades-long climb may have been exhausted in 2000, is there some reason to think that they’ve reversed? Even if you have an explanation for why full employment EPOP should be less in 2017 than it was in 2000, it hasn’t even reached the 2000-2008 levels that also failed to generate significant real wage growth. Post 2008 EPOP didn’t reach 78.6% until April of 2017 and only hit 79.0% for the first time in November.

    Given the level of employment, along with the decline of labor unions, the proliferation of “Right to Work” laws and prolonged period of job insecurity, the real mystery seems to be why so many economists think the lack of real wage growth is a mystery.

    • Smith says:

      Women in the workforce drive down wages because they get paid less. And that’s when they’re not being attacked. If you want a raise, try getting women equal pay. You can’t have 45% of the workforce paid 5% less than the prevailing wages and expect anything but wage stagnation.

  3. spencer says:

    Interesting, if you look at the year-over-year percent change in jobs it shows that employment growth is slowing significantly. For example, the year over year change in payroll employment is now 1.6% as compared to 1.8% last November. Job growth actually peaked at 2.6% in February, 2015 and has been on a downward sloping trend since.

    • Smith says:

      But the workforce only grows about 1%, as the population increases about .7% but it’s weighted towards working age population, young childless adult immigrants vs demographic of U.S. where 1/3 below 25 years old, and 15% over 65.
      The means new job growth of 170,000 per month can’t continue for more than 6 months unless EPOP increased. Since it does so easily, unemployment is vastly overrated as a metric. 1% of prime age population is 1 million. Also, the number of people working past retirement age is growing, adding maybe 1% to workforce over ten years.

  4. Smith says:

    To be more succinct, thoughts on low unemployment and stagnant wage mystery…
    The prime age labor participation at 1% below pre recession levels, 79% vs. 80%, represents almost a million people. With 90,000/month new jobs needed for population growth, 170,000/month new job creation takes another year to absorb them.
    Large companies sacrifice increased sales to hold down wages and sustain high profit margins. Of like GE, make strategic blunders, then lay off 18% of workforce (12,000) and sell off low margin but profitable businesses (Light Bulbs up for sale). Smaller companies are constrained by monopsony.
    A glut of college educated increasingly displace workers competing for positions with lower salaries.
    Greater control over wages and their stagnation lowers any increased demand, so more likely that low unemployment and stagnant wages can coexist.
    Real hourly wages are lower (by 1%) than they were nearly three years ago (early 2015).
    I’ll throw in women getting paid less is a drag on wages, nothing new, but neither is wage stagnation.

    • Fred Dole says:

      no, no,no. it is the boomer withdrawal. this is the first time since the BLS labor stats came into being, we have had a demographic cohort change like this one. When the Boomers were coming in during the late 60’s, the proceeding retirement cohort generation was vastly smaller meaning when they dropped out, they actually increased yry wages. This continued to build until the mid-00’s as the Boomers began their retreat from the job market.

      I would ignore this and go with ECI. It has been rising at the same rate outside the 2015 surge since 2014. By January 2019 we will be at 3%. That is maximum employment guys. The BLS non-sup due to the retirement boom, can’t see that……..and won’t. My guess .5-7 a month is subtracted from it due to the withdrawal. This also real spending lost and explains the slower trend growth and productivity. By next decade, this begins to wane thankfully!!!!

      My point is, the the economy is clearly overheating. The Fed as Fisher said 2 years ago, overstimulated the economy and now we have excess liquidity. It is pouring into stocks right now and will continue until the economy slows enough. Then it will pour into oil as the global economy lags creating a illusion of rising wages/inflation. Guys, we have seen this game before. Though, luckily, it won’t be as bad as last time. Which was a doozy liquidity buildup.

  5. Smith says:

    No, no, no, and no, nothing to do with boomer withdrawl. I’ve looked at the data previously and done some calculations. The boomer retirement has practically no effect on expected measurements of income and compensation, especially changes from one year to the next. As boomers retire, the demographic of the workforce barely changes. That’s because for the most part, they die after they retire. There is a very small effect from a slightly greater number taking early retirement. Keep in mind boomer births didn’t peak till around 1960, so those boomers are well under retirement age. There is a big fall in births from 1963 and 67 years later when they’re mostly retired, in 2030, you might some effect. We’re talking about order of magnitude .25 percent cumulatively spread out over a few years. It also might be offset completely as wages rise in response. I don’t buy into a supply and demand curve determining wages, but there is some effect. To the extent boomers withdrawal creates a shortage of senior experienced workers, that should boost wages, not lower them.

    Data indicates compensation costs are the same as they were in 2015, and yet the economy grew, there was some productivity increase (except for all of 2016), and hourly wages are down 1.1%

    The Fed for sure did not overstimulate the economy. If there is a bubble, it’s due to inadequate regulation, which is the job of the Fed.

    ECI data

    Not to be confused with the completely different use of the term ECI

    • Fred Dole says:

      Wrong on all accounts. The intro and outro have been distorted by so many people leaving the workforce over the last decade.

      Your trying to tell a story rather than looking at cold hard facts. The BLS data here is useless. 2015 sprung up when entry level wages surged back to the trendline after the recession faded. The Boomer retirement surge has seen its highest level yet in 2017.

      ECI does not comfort you and it shows.

      • Smith says:

        For anyone interested in basing their conclusions on actual data, you can check here.
        Mathematically, there can be no boomer effect suppressing wage increases, the difference for each year (100,000 births) is too small in a 155 million workforce, plus the birth cohort which peaks around 1960 and doesn’t fall much until 1965, haven’t retired. Participation at age 55 is around 80 percent, and 70 percent at age 60, even 55 percent at 65. 55/80 * 100,000 = 68,000 still working, or 32,000 retiring per year before 2025 tops, out of 150 million for a .02 percent if their replacements work for zero dollars.

        A further note on prime age participation now at 82 by FRED charting. How historically bad is current prime age workforce participation rate?
        It was still 82.5 at the end of the recession. Obviously a lagging indicator as continuing and increasing unemployment levels caused people to continue to drop out of labor market. The low point was July 2015 at 80.5, was only 81.5 when people voted in Trump last November. In the two terms of George Bush it only dipped below 82.5 once, and was below 83 percent about 1/3 of the time, 36 out 96 months. Before the recession you have to go back to August of 1986 to find participation lower. Thus the 2000s pre recession 83 rate seems to indicate there are still a million workers missing from the labor pool due to poor economic conditions.

        Here are some more projections on demographics of age, see Table 2 on page 6 from a Census Bureau report issued May 2014:

        • Smith says:

          The 32,000 retirements is not the total retirements, but merely the increased retirements from the previous year due to larger number. That figure is very conservative because the 100,000 fall in births is a maximum observed only once 1961 to 1962. Yet cumulatively you could add the .02 figure for ten years and still only reach .2 percent.
          Again, if boomer retirements contributed to a labor shortfall, to the extent supply and demand influences wages, that would boost wages. The story of the senior worker replaced by the entry level worker vs less senior, is dubious and still the .02 percent effect is based on the replacement receiving zero salary.

        • Smith says:

          The effect is actually much smaller than I report because 100,000 is max difference in births, but workforce participation is only 63%. The effect is closer to .01%, not .02% as given above.

        • Pinkybum says:

          “Thus the 2000s pre recession 83 rate seems to indicate there are still a million workers missing from the labor pool due to poor economic conditions.”

          The 1990s labor force rate reached as high as 84.5 so we could expect at least another 2.5 percent of prime working people being enticed into a booming economy – that would be 160*2.5/84.5 = 4.7 million people on the sidelines. Obviously the economy isn’t going so gangbusters so we can expect it to stay that way until more people are entering the workforce. I see the Fed nonsensically raised the FFR another 0.25% today – way to keep the economy suppressed.

Leave a Reply

Your email address will not be published.