I’m concerned re the lack of nominal acceleration of the blue-collar wage.

April 14th, 2017 at 2:28 pm

I said I’d get back to my tweet from this AM showing the deceleration in real wage growth, particularly for blue-collar workers. This is not exactly the stuff of 140 characters.

One can decompose the change in real weekly earnings for private-sector workers into three parts: changes in nominal hourly earnings, inflation, and average weekly hours worked. As the table below shows, the growth in real weekly earnings equals nominal wage growth – inflation + hours growth. See data note for details, particularly for how to compute the white-collar wage, which, unlike the rest of these data, is not an official BLS series, but a backed-out residual.

Source: BLS, our calculations.

Of course, someone could look at these numbers and declare that the problem is price growth, and that the Fed must continue to tap, if not smack, the brakes. I think that’s wrong. What’s happened on the price side is that, as energy prices have normalized, inflation has also climbed back to more normal levels. Pursuing deflation to boost real earnings won’t work, as it will undercut the demand needed to give middle- and low-wage workers the bargaining clout they otherwise lack.

I don’t want to make too much out of this as these data are noisy and the patterns can flip. Also, a lot of my work shows tight labor markets deliver more bargaining clout to less-advantaged workers, so I tend to think that, as we close in on full employment, the nominal pay of lower-wage workers should accelerate. Still, we know that wage inequality is still embedded in our labor market and need to keep a close eye out for the type of divergent trends you see in the table.

In this regard, the key unsettling number in the table is the one in bold showing no acceleration in the blue-collar wage. That bears close watching.

[h/t: Ben Spielberg]

Data note: The wage series for “white-collar” workers is derived from the following relationship:

Emp_a * W_a = Emp_bc * W_bc + Emp_wc * W_wc

Emp represents private employment and W the average wage (either hourly or weekly), with the subscripts a, bc, and wc representing all workers, “blue-collar” workers (production and nonsupervisory workers), and “white-collar” workers (those who are in the private sector but excluded from the production and nonsupervisory series), respectively.  That is, the total hourly wage bill for all private sector workers is equivalent to the total hourly wage bill for “blue-collar” workers plus the total hourly wage bill for “white-collar” workers.  Performing simple algebra and noting that Emp_wc = Emp_a – Emp_bc, we can solve for the average wage for white-collar workers.

Note that “blue-collar” and “white-collar” workers are not perfect descriptions, as which workers are counted in the production and nonsupervisory series varies by industry; a worker in sales in manufacturing might fall in our “white-collar” series, for example, while a worker in sales in a different industry might fall into the “blue-collar” series.

Print Friendly, PDF & Email

One comment in reply to "I’m concerned re the lack of nominal acceleration of the blue-collar wage."

  1. Smith says:

    Data? Number of blue collar? White collar?
    Your numbers for white collar are pretty skewed away from less rosy reality because you use average wage, pulled up by the top 1% and 10%. Median had problems too, so rough distribution numbers would be better.
    Also using exempt status is not very relevant, especially after Obama’s last minute halfway out the door after eight years of squandering his attention on non labor issues, dooming his legacy, and allowing Trump to sneak in with just enough populist votes, as a measure of class.
    Why not just use income level instead? Which is what Obama’s rule would have effected, so why follow Trump’s classification? Your statistics are not subject to a court order.

    Finally, for seemingly the millionth time, part of the problem for blue collar workers is that white collar workers are taking their jobs. College graduates have a 2.5% unemployment rate. Those without a degree are stuck with twice that rate 5%. The story of college grads pushing out blue collar workers from full employment and wage increases is a lose lose story. Not only does it mean blue collar wages decline, it also indicates college wages decline. And sure enough, the evidence is there. White collar wages flat, although the recent uptick of nearly .8% is welcome (again, suspicious too, since it’s an average and includes Bill Gates).


    The 66% of American without a college degree gets crushed, by low wage countries from below, and college graduates from above. When you ignore college graduates struggling, you pay a price at the polls. Trump won white male college graduates, and narrowly lost white female college graduates, 49% to 51% to Clinton. Letting them suffer, except for highly paid coastal elites, also leads to a divide. Why will they want to help those less fortunate when they themselves are struggling? They may be only 33% of the population, but they earn more, they vote more, and they have an outsized influence. It also means more education will just depress college wages further.