Feed on

A post from Lawrence Mishel of the EPI repeats the very often-cited idea that the typical worker is not doing much better today:

The issue of wage stagnation, however, should focus on what the vast majority of workers have been experiencing for most of the post-1979 period. Hourly wages, inflation-adjusted, grew only 0.2 percent annually from 1979 to 2014 and did not grow at all if we exclude the 1995-2000 period.

I do not actually want to debate or dispute the big picture behind what this data is intending to portray – that in a broad scale we do not seem to see the middle class surging ahead in living standards as rapidly as they had in previous generations. Now, again, that is arguable, but leave that. What I want to know of course is whether the statement above can even be factually confirmed? I will emphasize what I am asking:

for the VAST MAJORITY OF WORKERS … real wages grew only 0.2% per year from 1979 to 2014.

Show me. What is your evidence? He says, “for the vast majority of workers.” This means that (a) it must mean more than 50%, and considerably more. To be conservative, let’s assume he means 2/3 of workers and (b) that we are talking about workers. So, can anyone show me data that supports the idea that for 67% of people who were working in 1979, their real wages increased more slowly than 0.2% per year over the past 35 years? I find it hard to believe that 67% of the people who were employed in 1979 are even working today. If that is not what is meant by the figure, then what is? In reality, it can mean whatever people want it to mean irrespective of your political stripes. But what I am asking is for someone to show evidence on the actual wage experience of workers over that time and now survey sample data of repeated cross-sections of people. I may even settle for a series of repeated cross-sections with an average age increasing by one year each year. But as anyone can easily show you, you can have an aggregate workforce with no change at all in average earnings over time that still has every single employee seeing dramatic improvements in their living condition year after year after year.

To think of a highly stylized example, consider a firm with 30 workers. Worker 1 is in her first year, worker 2 iu her second, and worker 30 in her thirtieth and final year. Each year, the worker moves up one in the pay scale, with the 30th worker retiring and being replaced by a new worker. If the youngest worker earns $10 per hour, and there is a 5% pay bump at every level, the oldest worker will earn $41.16 per hour. The firm’s total salary bill would be $664.39. The median wage would be $20.29 while the average would be $22.15. If you revisit the firm every year (assume these are all already inflation adjusted), the total salary bill would be $664.39 next year and the year after and 28 years after that, and the median and average wage would stagnate. Meanwhile the “typical” worker would see her wage increase by a healthy 5% every single year. There is no way that aggregate data that does not follow the experience of each and every worker can capture this.

Before you use this simple thought experiment as an “aha” remember that it works in the other direction too – so that the real experience of workers can in fact be far worse than the aggregate data indicate. Now, I don’t think that is likely happening, but the point remains.

So, if you want to be scientific in the field of economics, go track down a panel of workers and tell us what actually happened to their real wages over time. Even if you find that they have not changed much, don’t get too excited, that is only one small step of the many required to illustrate just what has happened to their well-being. As I will not tire of saying, my real wage today is lower than it was when I was two years out of college, and that does not in any way reflect how “good” the labor market has been for me or how well off I am.

So, when we encounter people making very strong claims from aggregate data, I encourage us to ask, “how do you know?” It’s a healthy practice and I suggest it should be viewed as shameful to be making strong claims when it is not possible to do so. The more likely outcome is that you will be viewed as hostile, uncooperative or anti-science. Strange.

2 Responses to “A Good Research Project”

  1. Trapper_John says:

    For the general public, science cannot be nuanced beyond a soundbite. “GMOs cause [insert favorite harm here]”, “Fracking is evil because it causes earthquakes, polluted groundwater, etc.”, (and from the ‘other’ side) “The globe stopped warming 18 years ago”. How do you argue the many flaws of “Capital in the Twenty-First Century” without getting into the weeds? To dig further–to reach the nuance–I’d have to start assessing the relative merits of scientific papers, and that is too messy.

    Your point above is easily made with 5 minutes of attentive conversation, but that brings your audience to the point of asking themselves how someone smarter (or more informed) would rebut your claims. What would (sigh) Robert Reich say? But this misses the main point, which is that things are often more complex than soundbites. How can we break people out of their entrenched positions when they are bolstered by the calming thought: “He must be anti-science because someone smart has thought through this–I read it in the New York Times!”

    My thought is that these things change slowly, if at all. Look at governmental dietary recommendations, where fat and salt have been the primary enemies of the last 50+ years. Only through good science that convinces other good scientists (which write good soundbites) can opinion be swayed. And the incentives for an academic deviating from canonical thought are net slim to negative. Given the absence of real, controlled experiments, I’m not sure any such change can be achieved in economics.

  2. Doug M says:

    What we can say about the stylized thought experiment is that the real wage of entry level workers is stagnant.

Leave a Reply