The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« Crabbing About Core | Main | Progress? »

August 29, 2006

More Confusion About The Return To Labor

Regular readers of macroblog know that one of my pet peeves is the insistence of some to measure labor compensation using narrow measures of hourly wages.  Yesterday, The New York Times continued to insist in an article titled "Real Wages Fail to Match a Rise in Productivity." 

Not to fret, though.  Russell Roberts administers the smackdown:

Let me repeat the key sentence:

The median hourly wage for American workers has declined 2 percent since 2003, after factoring in inflation.

That's a very strange sentence for many reasons:

1. Why would you use a measure of compensation that ignores benefits, an increasingly important form of compensation?

2. Why would you use 2003 as your starting point when the recession ended in November of 2001?

... for every year since the recession of 2001, real hourly compensation has actually increased. It's up since 2003 as well. And this year it's up quite dramatically...

As I have mentioned here before--the standard claims you hear about labor's share declining come from using wages without other forms of compensation. When you include benefits, labor's share is virtually a constant at 70% of national income and has been steady since the end of World War II,

I might just stop there, but on this topic I really can't stop complaining. In reading the Times piece, The Capital Spectator instead focuses on what would appear to be the flip side of the claim that labor share is falling: A rising share of income accruing to capital.  Again from the Times article:

In another recent report on the boom in profits, economists at Goldman Sachs wrote, “The most important contributor to higher profit margins over the past five years has been a decline in labor’s share of national income.”

I'm not sure how that conclusion is reached, but my colleagues Paul Gomme and Peter Rupert make this observation:

.. for labor’s share as computed by the Bureau of Labor Statistics, a fall in labor’s share does not necessarily imply a rise in capital’s share; indirect taxes and subsidies constitute a wedge between these two series. Consequently, a fall in labor’s share could be associated with a rise in capital’s share, but it could also be due to a rise in the share of indirect taxes less subsidies... Further, the terms “capital’s share” and “profit share” are often used interchangeably, ignoring the fact that capital income derives from more sources than just (corporate) profits.

What I think is more important in the Gomme-Rupert analysis is this, from one of my previous discussions on this topic:

... the “historic lows” in labor’s share are observed only in the nonfarm business sector series produced by the Bureau of Labor Statistics. Other measures of labor’s share—for example, for the nonfinancial corporate business sector or the macroeconomy more broadly—are currently near their averages over the last several decades.

Those alternative measures are ones that avoid, for example, the problems associated with allocating rental income and proprietor's income between labor and capital.  (In other words, they avoid imputing things that we don't observe.)

In what I suppose is a related topic, there has been plenty of blogging in the last several days on the injunction against a strike by Northwest Airlines flight attendants, union-busting, and wage inequality -- from Dean Baker, from Brad DeLong, and from Mark Thoma (among others, no doubt). 

Just about a month ago the raging debate was about why it might be that changes in income inequality have been concentrated among the top 1% of income earners -- excellent examples of that debate being found here and here. Once you have identified the real issue as being about the top 1% of the income distribution, I'd think you have pretty much ruled out unionization (or the lack of it) as a dominant driver of the major trends in who gets what.  In any event, on the question of unionization and wage inequality, I'm with Greg Mankiw.

UPDATE: Max Sawicky takes me to task for omitting this from the Gomme-Rupert quotation above:

... we find that the share of indirect taxes less subsidies does not vary much.."

Well, it certainly wasn't my intention to deceive, but Max has a good point. I'll consider my knuckles duly rapped.

In a related vein, Dean Baker notes that depreciation can also cloud the picture on capital sharel.  knzn thanks Dr. Baker for us all, but comes to the "indirect taxes" conclusion: There's not enough action there to eliminate the decline in standard labor share calculations.

Greg Mankiw covers the theory of productivity and wages, and provides several reasons the data might not cooperate with theory (including Dean Baker's observations that different price indexes are in play).  If you would like to see the math behind Professor Mankiw's discussion of labor shares and the Cobb Douglas production function, Economic Investigations has just what you're looking for.

If that's not enough, there's more at The Big Picture and at ElectEcon.

August 29, 2006 in Labor Markets | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference More Confusion About The Return To Labor :


I wouldn't want to spoil the party, but I do not see the problem of inequality as being a top 1% problem. In fact, I would say that at least the top 5%, and perhaps the top 10% have gotten more than their share of productivity gains over the last quarter century.

According to the State of Working America, the real wage for workers at the 90th percentile rose by 27.2 percent from 1979 to 2003. Wages for workers at the 95th percentile rose by 31.1 percent.

As you rightly note, there has been a shift to non-wage compensation over this period. This shift was almost certainly strongest among those in the top wage brackets (the tax code gives high-wage earners more incentive to get compensation in tax exempt or deferred non-wage forms). We also know that high-wage earners are far more likely to have health care and pension coverage than low-wage earners.

For these reasons, it reasonable to believe that the compensation of workers in these income brackets did increase more than average productivity growth over the last quarter century. While these folks may not have pockets nearly as much as the CEOs and Wall Street honchos, they are nonetheless accomplices in the great upward redistribution of the period, and can be counted on to provide an important political base to the new status quo.

Posted by: Dean Baker | August 29, 2006 at 10:36 AM

Dave, What's your point, that leftys are not without blame when it comes to slanting economic argument? It's the arguer's preogative what he includes & within that, what he emphasizes. It's tough for me to find fault with this irksome peeve of yours because, as you are sufferingly aware, I think adoption of the Boskin Commission recommendations was inappropriately frontrun by AG, was illconsidered, & has been disastrous for our economy. (By the way, Dean Baker was one of VERY few Economists who questioned the wisdom of the changes).

Posted by: bailey | August 29, 2006 at 04:39 PM

OK, real compensation is up a little. Real GDP per capita growth appears to exceed real compensation. Put another way - if labor productivity growth exceeds real compensation growth, how can it not be that the profit share has not increased? As I said over at Max Sawicky's place (he's touting this year's upcoming release of EPI's "big book") - are we talking about the same economy?

Posted by: pgl | August 29, 2006 at 06:37 PM

"Consequently, a fall in labor’s share could be associated with a rise in capital’s share, but it could also be due to a rise in the share of indirect taxes less subsidies..."

talk about slanting the debate - the part you misssed, deliberately obfuscated by the ellipsis - surprise, surprise -

.. we find that the share of share of indirect taxes less subsidies does not vary much..

You are just a plain liar

Posted by: comeon | August 29, 2006 at 09:22 PM

Dean -- I agree.

bailey -- I'm not accusing the Times of being either lefty or of slanting things. I just think that the use to which they are putting the wage measure -- for making some statement about the overall return to labor -- is inappropriate. I've probably said it enough, and should stop, but geez, people just keep tempting me.

pgl -- Although it is probably not clear, I don't really feel like I have a stake in whether labor share is historically low or not (other than a bounce back might have some consequences for short-run inflation dynamics and I would dearly love to keep using Cobb Douglas production functions). I'm just not sure that the labor share data is reliable, for the reasons articulated by Gomme and Rupert. But as you point out (differences between profits and total returns to capital -- which includes returns to lenders -- aside) this does mean I am also suspicious of productivity numbers.

comeon -- I suspect that you posted your comment prior to my mea culpa. I do find it distressing that you see fit to call me a liar. I try very hard to be forthright in my writing here, to promote a culture of civility, and keep an open mind about things. I fear that it says something about the nature of much debate in the blogoshpere that your immediate assumption was that I was dissembling, and not that I simply made an honest mistake.

Posted by: Dave Altig | August 30, 2006 at 09:20 PM


I also get annoyed when people focus on a narrow measure of compensation ... when a broad one is available.

But, if you want to talk about distribution, or even just medians like in the NYT article, there isn't a broader measure of compensation that you can turn to. Since non-wage payments are a significant share of total compensation, this is a significant hole in U.S. economic statistics. So in this instance, you should probably direct your annoyance to the BLS rather than the NYT.

BTW, the best evidence that I know of on true ineqaulity of compensation is (BLS economist) Brooks Pierce's paper that was in the QJE a few years ago. Using the data underlying the ECI, he was able to construct reasonable measures of total compensation at the micro level. His conclusions support what Dean Baker said above: (1) Data on wages and salaries significantly understate the degree of inequality. I.e., the upper end of the distribution gets a bigger share of the benefits than they do of the wages. (2) The wage and salary data understate the magnitude of the increase in inequality that occurred in the 80s and 90s.

Posted by: Dan Sullivan | August 31, 2006 at 12:01 AM

Dan -- Thanks a bunch for those comments. I wasn't aware of the Pierce paper (though I should have been -- Sounds like a blog opportunity!) As my response above indicates, I agree with Dean's comments. I hope that none of my own comments are construed as an attempt to deny the fact inequality has risen -- no sensible person would do so. My beef with the nyt article was its insistence on attempting to correlate wages ex benefits etc. with productivity, labor share and so on. That even though the authors seem to know better, as evidenced by their notice of broader compensation numbers. I'm not even claiming that their general conclusion is wrong -- as I said above, I have no horse in that race beyond wanting to understand the facts. I just dearly wish folks who claim to inform would focus on measures that are meaningful for the question at hand.

Posted by: Dave Altig | August 31, 2006 at 08:39 AM

"I just dearly wish folks who claim to inform would focus on measures that are meaningful for the question at hand."
BRAVO, Put it on the masthead!

Posted by: bailey | August 31, 2006 at 10:12 AM

Income inequality is such a loaded term. Is the claim that those who saw a higher rise in real income did not produce an equivalent rise in productivity?

If workers on the lower end of the pay scale produce less of an increase in returns for their labor, why shouldn't their compensation increase at a lower rate than those whose labor produces returns which have grown more substantially?

Posted by: stan | August 31, 2006 at 11:31 AM

Dave -- Your welcome. I think the Pierce paper is an important one that should be better known.

BTW, if what Russell Roberts administered was a smackdown, David Leonhardt just delivered a pretty effective smackback.

Posted by: Dan Sullivan | August 31, 2006 at 11:15 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad