I used to be absolutely certain that international trade was the area of economics that journalists covered worst. Recently, it’s become increasingly clear that reporting on American wage trends, and specifically the issue of whether they’re finally starting to rise at acceptable rates, is giving trade reporting a run for its money.
The latest and possibly most egregious examples come from reports on the November jobs figures released yesterday by the Labor Department. According to The New York Times‘ Neil Irwin, “[W]hat was warming the hearts of many jaded economy-watchers on Twitter Friday morning were the numbers on wages.” In the words of the Washington Post’s Chico Harlan, “the latest numbers came with nascent signs of wage growth, the result of qualified individuals reentering the workforce, putting pressure on companies to retain their best employees or bid for new applicants.”
Financial Times correspondents told us, “Wages rose by 0.4 per cent over the previous month — the equivalent of 5 per cent at an annualised pace. That was much faster than the pace seen so far in the recovery, and is enough to drive inflation higher if sustained.” Bloomberg’s story featured the headline, “Broadest Job Gains Since 1998 Mean Faster Wage Growth is Coming.”
It’s like no one knows how to count, or bothers to look at the historical tables posted on the Bureau of Labor Statistics website. For this data shows unmistakably that there is absolutely zero evidence of any pick up in wages. More specifically, the type of gain registered in November is anything but unprecedented during the current economic recovery, and has added up to bupkus over the longer haul.
Let’s assume that the November numbers, which are still preliminary, hold up – or even improve. First, the monthly increase was 0.37 percent (before adjusting for inflation), which is only 0.40 if you round up. Much more important, this gain came after two months of data that were bad even by recent wage standards – a 0.12 percent advance in October (also still preliminary), and no change at all in September. In fact, here are all the monthly changes so far for 2014:
January +0.21 percent
February +0.21 percent
March +0.21 percent
April +0.04 percent
May +0.21 percent
June +0.29 percent
July +0.04 percent
August +0.33 percent
September 0 percent
October +0.12 percent
November +0.37 percent
Hats off to you if you can describe this as a wage pickup with a straight face.
But what about Harlan’s observation that November’s wage “spike” was “the sharpest month-to-month change in more than a year”? The only reasonable answer is “big deal.” Anyone familiar with the data knows that such single-month improvements have had no predictive value whatever.
For example, Wages rose even faster in June, 2013 – by 0.38 percent. Here’s how they fared over the next six months: -0.04 percent, +0.25 percent, +0.12 percent, 0.12 percent, 0.25 percent, and 0.08 percent.
Wages also rose by 0.38 percent in November, 2012. Over the next six months, they increased by 0.30 percent, 0.17 percent , 0.17 percent, 0.08 percent, 0.21 percent, and 0.38 percent.
In October, 2011, wage rose even faster – by 0.43 percent. Over the next six months, the rate of change was -0.09 percent, 0.09 percent, 0.13 percent, 0.22 percent, 0.30 percent, and 0.13 percent.
So there’s at least a heavy burden of proof on those who would attach any significance to November’s wage increase.
But these claims look even more fanciful if you examine the year-on-year changes. As widely noted, November’s 2.11 percent increase was still unimpressive historically. (It, too, will be revised.) But in fact, the November gain was unimpressive even by 2014’s measly standards. You’ll see what I mean by looking over this year’s 2013-2014 changes by month so far:
January 1.98 percent
February 2.10 percent
March 2.11 percent
April 1.97 percent
May 2.05 percent
June 1.96 percent
July 2.04 percent
August 2.12 percent
September 2.00 percent
October 1.99 percent
November 2.11 percent
These figures show that the November year-on-year change was only the second highest of 2014, and that it’s been equaled once before. Where’s the momentum?
But perhaps the most convincing rejoinder to the wage pickup claims comes from comparing 2014’s year-on-year changes in wages to 2013’s. Here they are with the 2013 figures in parentheses:
January 1.98 percent (2.11 percent)
February 2.10 percent (2.10 percent)
March 2.11 percent (1.88 percent)
April 1.97 percent (1.97 percent)
May 2.05 percent (2.05 percent)
June 1.96 percent (2.17 percent)
July 2.04 percent (1.91 percent)
August 2.12 percent (2.21 percent)
September 2.00 percent (2.04 percent)
October 1.99 percent (2.25 percent)
November 2.11 percent (2.16 percent)
The message couldn’t be clearer: Wages have been rising more slowly on an annual basis this year than they were last year, not faster. The year-on-year 2013 increases have been greater than their 2014 counterparts for the last four months, and for six of eleven months’ worth of data so far. In three months, the rates of change were identical.
You’ve heard the expression making mountains out of molehills? These reporters have made mountains out of absolutely nothing – at best. One conspicuous exception: The Wall Street Journal’s Ben Leubsdorf did note that “Similar monthly jumps in earnings in recent years have quickly faded.” What are the odds that this instance of sensibility has more predictive value for the Mainstream Media than the November wage gains seem likely to have for American workers?
RealityChek 