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.

« The Non-Mystery Of Labor Compensation (Bob Hall Explains It All, Take Three) | Main | Not The Most Bullish Day For Economic News »

August 30, 2005

More On The Labor Compensation Issue

In yesterday's Wall Street Journal (page A9 in the print edition), Stephen Moore made an observation similar to the one emphasized in my previous post: While wage and growth has been pretty anemic  over the past several years, it is difficult to make the same case with respect to total labor compensation, which adds benefit payments to wages.

In his brand spanking new blog, Daniel Gross takes exception to one of Moore's comments (duly noted by Brad DeLong):

Stephen Moore, a member of the Journal's editorial board, writes in today's paper:

The explosion of benefits paid to workers is in large part an artifact of the federal tax code, which allows employers to deduct from taxes pensions, health care, child care, and the like, but not wages.

Read it twice. Stephen Moore apparently thinks companies can't deduct wages paid to their workers from their taxable income the way they can deduct pension, health care, and child care costs. And apparently nobody at the Journal's op-ed page knew enough, or thought enough, to correct him.

Maybe Dan is mainly upset, one newspaper guy to another, about sloppiness of any kind.  But the economics of the central claim -- that the tax code favors payment in non-wage form -- does not seem wrong to me.  While it is true that employers can deduct wage expense just as they can other employee costs, it is not true that wage payments and benefits are the same when they get into the hands of employees.  Wage payments are taxed, benefits are not.  When the tax system is taken as a whole, a pre-tax dollar delivered to employees in the form of benefits yields a higher net payout to workers than a dollar delivered in the form of explicit wages.  To me, that sounds like the type of tax distortion Moore was trying to describe.

On a (sort of) related point, in the comment section of the previous post Angry Bear's pgl takes me to task (probably with some justification) for not addressing an argument that he has made before:

In my 1st RBC post, I noted Kash's argument that the rise in real compensation is substantially due to more costly health insurance. Not better, just more costly. Why did I mention it? It's a supply side. And yet you don't note that this is the reason for the divergence between real wage growth v. real compensation. Huh?

The reason that I did not take note of it is that I'm not convinced the observation is relevant.  Health insurance is the largest single component of employee benefit expenditures.  For a given total amount of compensation, increasing payments in the form of more insurance expenditure means less in other forms -- including wages.  Whether we get more or less for those higher insurance expenditures is an interesting and important question, as is the question of whether the tax code is introducing welfare-degrading means of compensating workers.  But I don't think it has much to do with the question I was trying to address, which was whether or not the return to working has been growing at an abnormally low rate.

UPDATE: Angry Bear (the orginal!) notes that the offending passage has been corrected:

THE AUG. 27 feature, "The Wages of Prosperity," by Stephen Moore mistakenly reported that wages are not tax deductible to employers. The relevant sentence should have said, "Fringe benefits have exploded in recent years because benefits are tax free to employees, but wages are taxed."

AB is still not happy.

The corrected wording doesn't pass muster as an explanation for flat wage growth, either. Benefits have been tax free to employees, and wages have not, for as long as I can remember. So the ongoing taxability (to the employee) of wages but not benefits simply cannot explain why wage growth has been flat in the last 4 years.

The best I can do here is to simply repeat that I think it can.


August 30, 2005 in Health Care , Labor Markets | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference More On The Labor Compensation Issue :


I think pgl really makes a point that is also in your chart in the earlier post. Real compensation (he would argue), when properly deflated, has not been rising significantly, whereas overall domestic product (per worker) clearly has been rising, no matter how you deflate it. Something – presumably something about the labor market – is causing workers to get the short end of the stick when it comes to dividing the proceeds from recent economic growth. This same point is made by noting that the compensation line in your chart has been consistently (and increasingly) below the productivity line in recent years.

I believe, however, that we are barking up the wrong tree by talking about (ex post) real compensation. To my mind, the whole basis for talking about labor market slack originates from the finding of very strong empirical links between “slack variables” (most typically the unemployment rate) and changes in the inflation rate. I’ll leave you and pgl to argue about the implications of the division of output between labor and capital, and whether policymakers can or should do anything about that, but I think we can all agree that more employment is generally a good thing if it doesn’t put upward pressure on the overall inflation rate. Therefore, the sense in which I would mean the word “slack” is that there is room for more employment without exacerbating inflation.

Thus, when I cite low help wanted advertising as evidence of slack, it is specifically with the knowledge that, historically, there is a strong correlation between help wanted advertising and changes in the inflation rate. (Indeed, the fact that help wanted advertising reached its all-time high in the late 1970s might suggest that oil prices are less important than one might think, both as a source of inflation and as a depressant to labor demand.) Of course, the correlation isn’t perfect, and it exhibits the same sort of instabilities as the inflation-unemployment correlation. However, when we take into account a variety of slack variables (e.g. average duration of unemployment, growth rate of payroll employment, etc.), all of which have strong empirical correlations with inflation changes, the picture that emerges (at least to me) is one of substantial slack.

Posted by: knzn | August 31, 2005 at 10:51 AM

I think at this point, we just agree to disagree. I interpret the instability in the inflation/slack relationships -- which I guess I think are more substantial than you do -- as a sign that the slack concept is on shaky ground. I understand that I am in the minority on this. I do indeed believe that employment could grow much faster without generating inflationary pressures. But I also beleive that this will happen when the real environment changes to make it so.

Posted by: Dave Altig | August 31, 2005 at 05:17 PM

I’m willing to acknowledge that the glass of Keynesian economics is only half full, but you and Bob Hall seem to be saying that, because it’s half empty, we need to go back to the refrigerator. Maybe it is a disagreement about the severity of the instability, as you suggest. But I could imagine a “composite slack indicator” that would have a much more stable relationship to inflation than its individual components.

I’m not sure what you mean when you say that “employment could grow much faster without generating inflationary pressures.” Almost anyone would have to agree there are some circumstances (e.g., a change in labor/leisure preferences) where that could happen. Are you making a substantive statement? What changes in the real environment do you think are necessary, and why?

Posted by: knzn | August 31, 2005 at 06:23 PM

knzn -- Productivity growth -- in the fundeamental, exogenous sense -- is the obvious example. I cannot answer definitively what the circumstances would be that would lead to a pickup in employment growth. That would suggest I know how to explain the pattern of labor market dynamics over the past five years, which I don't. I don't think anyone else does either. There is a pattern that represents a slowdown relative to the fire-breathing pace of the latter 1990s, for sure. I am not providing an answer, but rather objecting to the assumption that deficient demand, in the traditional Keynesian sense, is the explanation.

By the way, when it comes to an explanation of unemployment, I'm not saying the Keynesian glass is half empty. I'm saying that, for all practical purposes, it is completely empty. This should be read the way Hall offers it: Standard neoclassical synthesis models are just not suited to thinking about unemployment. That is not to say, however, that it is never the case that monetary policy mistakes create unemployment, or that policy is impotent to affect the unemployment rate more generally. I think there is a lot of confusion in all of this that comes from not separating the methodological case made by Hall with the interpretation of what types of shocks are actually driving outcomes today. I hope I have not contributed to that confusion.

Posted by: Dave Altig | September 01, 2005 at 01:17 PM

Health insurance is a great benefit and I think all employers should provide it.

Posted by: California Health Insurance | November 04, 2005 at 06:20 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