The requisite 30 days have passed since my op/ed on real wage increases appeared in the Wall Street Journal. So here’s the whole piece.
Wages Are Growing Faster Than You Think
A 0.1% annual rise turns into 1% when adjusted for inflation, benefits and the changing labor force.
By David R. Henderson
Sept. 20, 2018 6:07 p.m. ET
Standard wage data show that between the spring of 2017 and the spring of 2018, real wages in the U.S. increased only 0.1%. But there are three major problems with these data. First, they don’t account for fringe benefits, which are an increasing proportion of employee pay. Second, standard wage data use an index that overstates the inflation rate. Third, each year the composition of the workforce changes, as older, higher-paid workers retire and young, lower-paid workers enter the workforce.
A study released this month by the White House Council of Economic Advisers addresses these three biases and concludes that real wages grew by 1% in 2017-18, not the measly 0.1% reported in the wage data.
- Fringe benefits. Because benefits have become an increasing proportion of employee compensation over time, growth in real wages has been understated. The CEA estimates that including benefits would add 0.2 percentage point to the 2017-18 figure.
- Inflation. An ideal measure would cover a very large percent of what workers buy, would account for the tendency to buy less of goods and services whose relative prices have risen and more of goods and services those whose prices have fallen, and would somehow correct for the improvements in quality of almost everything sold in the private market. As Stanford’s Michael Boskin has pointed out, the usual measure of inflation, the consumer-price index, doesn’t do this very well. An alternate measure of inflation, the personal- consumption-expenditures price index, while also imperfect, is a better measure of inflation. Economists at the Federal Reserve prefer the PCEPI to the CPI. Using the PCEPI adds 0.5 percentage point to the 2017-18 growth of real wages.
- Change in the labor force.As baby boomers retire, they are replaced by younger workers. So even though average wages may not rise much, the wages of the majority of people working could rise a lot. The Census Bureau estimates that 3.57 million people turned 65 in 2017, compared with 2.68 million in 2010. Taking account of the decline in older, higher-paid workers and the increase in younger, lower-paid workers, the CEA estimates that this “composition factor” added 0.3 percentage point to real wage growth from 2017-18.
Because of rounding error, these three factors add up to 0.9 point. The net result: When adjusted for benefits, inflation and seniority, real wages actually grew 1% between 2017 and 2018. This is not a partisan point. The same methodology would show that real wages grew more than was reported during much of President Obama’s time in office.
But there is, in this context, one relevant difference between the Trump and Obama administrations: the 2017 tax cut. Real after-tax wages increased 1.4% between 2017 and 2018, according to the CEA study. This overstates the benefits, given the Congressional Budget Office’s estimate that the tax cut will make the 2028 federal debt 7% higher than otherwise. Yet even aside from the tax cut, real wages are growing at a healthy pace. That’s good news for American workers.
Mr. Henderson is a research fellow with Stanford University’s Hoover Institution.
Appeared in the September 21, 2018, print edition.
READER COMMENTS
Mark Barbieri
Oct 23 2018 at 12:47pm
Excellent article, as usual.
The notion of wage stagnation seems to be widely held. I’ve had conversations with several working age people that claim to be worse off than their parents were. I remind them that their cars are vastly superior to what their parents drove. Their ability to communicate (cell phones, texting, video calls) were science fiction compared to their parent’s era of expensive long-distance calls. They fly more for leisure because it is so much cheaper. I could go on and on. In terms of what I get in return for my labor, I seem much better off than a lot of rich people were in my father’s day and better off than virtually anyone in my grandfather’s day.
Alan Goldhammer
Oct 23 2018 at 12:53pm
The CEA report did not accurately reflect the increase in employee paid health care premiums along with increased deductibles and co-pays. It is a charitable on their part to say that fringe benefits represent a net positive increase (but of course this is a political document and I don’t think they could say otherwise).
JFA
Oct 24 2018 at 9:22am
I am also not terribly convinced by the fringe benefit addition to the increase in real wages. Measuring the value to workers of fringe benefits on a dollar cost basis will bias the increase in real compensation upwards. Due to differential tax treatments, the value (to the worker) of a dollar of fringe benefits is less than a dollar in cash.
That being said, the fringe benefit adjustment the CEA performs only accounts for a small part of the overall corrected increase.
A bigger point is this: a before tax 1% rise in real income is paltry.
derek
Oct 25 2018 at 11:24am
Is there a version of the report’s Figure 9 anywhere that does not factor in the Trump tax cuts into the real total compensation? That is the primary time series comparison of interest, in my opinion, but I am not seeing it anywhere.
robc
Oct 23 2018 at 1:05pm
While I might not be the norm, my premiums are going up about $1.50 per pay period, while the company portion is going up a whole lot more. My deductible is staying the same. I don’t have co-pays, as it is an HSA plan, but the amount I can put into my HSA is growing by $100.
This is 2019 vs 2018 (we are in open enrollment right now), but basically the same can be said for 2017 vs 2018 also.
The problem is that health insurance premiums are going the wrong way. With a free market they would be going down.
nobody.really
Oct 23 2018 at 1:29pm
Good point about labor force composition.
Good point about fringe benefits. 100% of the amount spent on fringe benefits is compensation, even if the amount is punitively spent by the employer. (Put another way, the idea that your employer “shares” the cost of some benefit with you is illusory; you pay for it all via reduced salary. But, to the extent that your employer can obtain benefits at better terms than you can–for example, via a group heath policy paid for with pre-tax dollars–you get a real benefit, even if you’re bearing all the incremental cost.)
Tricky point about inflation. Yes, it’s a fair argument–but has been since forever. Moreover, inflation is relatively low, so I wouldn’t expect this variable to be especially powerful.
Really? Where can I find some basis of comparison–specifically, a basis for comparison that incorporates all the adjustments included in this essay? When Picketty (sp?) suggested that average productivity growth was 1.0-1.5%, was he using a PCEPI deflator? And with the Dow growing by 25% in 2017, I have difficulty wrapping my brain around the idea that a 1% adjusted increase in compensation is “a healthy pace.”
David Henderson
Oct 23 2018 at 7:45pm
Mark Barbieri wrote:
Excellent article, as usual.
Thanks, Mark.
You also wrote:
I’ve had conversations with several working age people that claim to be worse off than their parents were. I remind them that their cars are vastly superior to what their parents drove. Their ability to communicate (cell phones, texting, video calls) were science fiction compared to their parent’s era of expensive long-distance calls. They fly more for leisure because it is so much cheaper. I could go on and on. In terms of what I get in return for my labor, I seem much better off than a lot of rich people were in my father’s day and better off than virtually anyone in my grandfather’s day.
Exactly. Well said.
Alan Goldhammer wrote:
The CEA report did not accurately reflect the increase in employee paid health care premiums along with increased deductibles and co-pays.
I bet you’re right about the deductibles and co-pays. Hard to believe, though, in a document that explicitly discusses the increased role of benefits, that they would misstate the obvious benefit–namely the amount of the health care premium paid by the employer.
nobody.really
Oct 24 2018 at 12:27pm
Are you confusing employee-paid healthcare costs with employer-paid healthcare costs?
Henderson notes that employer-paid healthcare costs are central to this analysis. Goldhammer suggests that the analysis neglected to address employee-related healthcare costs. But is that relevant to the analysis?
As I opine elsewhere, ALL benefits are a form of employee compensation–a substitute for higher wages. Thus, it’s crucial to measure employer-paid benefits. In contrast, employee-paid “benefits” are just another kind of employee spending. It is only the employer’s costs, not the employees’s, that are relevant to this analysis. True, today health care costs eat up more of an employee’s assets than in the past, leaving less “discretionary income.” And in the 1970s, gas prices ate up more of an employee’s assets than in the past, leaving less discretionary income. This leads to a frustration of expectations, but it’s irrelevant to evaluating the size of the employee’s compensation package. Or so it seems to me.
David Henderson
Oct 23 2018 at 7:50pm
nobody.really wrote:
Tricky point about inflation. Yes, it’s a fair argument–but has been since forever. Moreover, inflation is relatively low, so I wouldn’t expect this variable to be especially powerful.
Yes, it has always been a good argument, but that doesn’t make it tricky. And if it’s a good argument, it should be made even if it has always applied.
What matters is not whether inflation is relatively low but whether to what extent it is overstated. In short, what matters is the difference between the two measures, however low or high inflation is.
You wrote:
And with the Dow growing by 25% in 2017, I have difficulty wrapping my brain around the idea that a 1% adjusted increase in compensation is “a healthy pace.”
I don’t, not when the official data were saying 0.1 percent. The 1.0 percent represents a nine-fold increase over the official data.
nobody.really
Oct 23 2018 at 10:31pm
Cool. And you’ve missed my point.
When running for president, Trump complained about how the unemployment rate was reported, claiming that a “true” unemployment rate would be much higher. And, yes, plenty of economists have noted various other ways we might count and report employment statistics.
Each of those methods might have strengths and weaknesses. But whatever the merits of any given measure of unemployment, it would provide little value unless we also had evidence of how that measure changed over time, thereby providing a frame of reference. Traditional measures of unemployment, whatever their strengths and weaknesses, have had CONSISTENT strengths and weaknesses over time, and thus provide some basis for comparison.
You propose some new measure of the compensation growth rate. And you suggest that it’s a “health pace” because it’s larger than some different, non-comparable measure. Huh?
A mere nine-fold increase? That’s nothing. I’ll sell you 100 Japanese Yen for each US dollar. That’s a 99-fold increase, so it MUST be a good deal, right?
Well … no. We’d need some benchmark, such as the number of yen it takes to buy a Big Mac, before we could judge the merits of the exchange.
Likewise, you’ve told us that a 1% adjusted rate of grown in compensation is a “healthy pace,” but you’ve provided no benchmarks. Does it represent a higher adjusted rate of wage growth than average over the past 5 years? Past 10 years? Past 50 years? You don’t say.
I give the quality of this analysis a score of 622. It must be a good score; after all, it’s a nine-fold increase over 66.2. What more could you want?
David Henderson
Oct 24 2018 at 10:10am
You write:
You make a good point, but you overstate it. It would be ideal, of course, to have the same measure over time. But that doesn’t at all mean that coming up with a better measure is valueless.
You write:
Huh? You’re confusing levels and changes.
You write:
I don’t say because I don’t know.
nobody.really
Oct 24 2018 at 11:13am
So, as far as you know, your 1.0% rate of growth might be the lowest its been in 50 years–yet you’d still characterize it as “healthy”? What would it take for you to characterize the growth rate as unhealthy?
Here’s my take-away: You’ve identified methodological improvements in how we should calculate compensation growth rates; good on you. But that still provides no basis for concluding that the current rate of growth is healthy or not.
You can claim that your new speedometer is so accurate, it provides an iron-clad basis to judge whether you’re speeding. I claim that no matter how accurate your speedometer is, you can’t evaluate whether you’re speeding without some frame of reference, such as knowing the speed limit. And on this point, I sense we’re at an impasse.
Mark Z
Oct 24 2018 at 3:34am
nobody.really,
Comparing the growth rate of compensation to the Dow Jones is, imo, rather meaningless. Why would you expect wages to grow as fast as the stock market? For one thing, the stock market is more volatile than wages, so shareholders get a risk premium. Secondly, we have to remember that the economy is global, and publicly traded companies don’t just employ American worker; why compare the profit margin of a company that produces most of its goods in, say, Brazil, with American wage growth. Lastly and most importantly, there’s no reason why wage growth should automatically be expected to be comparable to return on investment. If recent economic growth is primarily attributable to increased productivity of capital, then we should expect returns on investment to exceed wage growth, and that’s not a problem or an injustice.
Instead, we should be looking at how wages compare with productivity. And it turns out, despite claims oft made to the contrary, if you more properly account for inflation, compensation has continued to track productivity pretty well.
https://www.wsj.com/articles/donald-boudreaux-and-liya-palagashvili-the-myth-of-the-great-wages-decoupling-1394151793?tesla=y
http://cep.lse.ac.uk/pubs/download/dp1246.pdf
David Henderson
Oct 24 2018 at 10:06am
Thanks, Mark Z. I’ll add my own response to nobody in a minute.
nobody.really
Oct 24 2018 at 10:41am
I don’t know if its meaningless. But I agree that I’d expect returns to capital to differ from return to labor, and generally be higher.
That’s why my initial comparison was not to the stock market, but to Piketty’s assessment that productivity tends to increase by 1.0-1.5% per year. But since I don’t know whether Piketty calculated his figures using the PCEPI deflator, or adjust for labor force composition, or take any account for how much of compensation is paid in the form of fringe benefits, I don’t know if Piketty’s 1.0-1.5% is comparable to Henderson’s 1.0%. (Ok, I expect that Piketty’s analysis did NOT consider labor force composition, and it’s unclear to me that the form of compensation to labor would have any bearing on his calculation. But inflation adjustment might.)
In any event, I commend you for your cogent argument: We should ground our expectations in some kind of relevant frame of reference. This is also my argument. I don’t know how to judge whether 1.0% growth is healthy without some basis of comparison.
Thaomas
Oct 24 2018 at 11:48am
Looks fine, but one period groth is not too informative. The concern is the slowdown (with the exception of the late 90’s) in median wage growth since the ’70’s.
derek
Oct 25 2018 at 11:24am
Is there a version of the report’s Figure 9 anywhere that does not factor in the Trump tax cuts into the real total compensation? That is the primary time series comparison of interest, in my opinion, but I am not seeing it anywhere.
Comments are closed.