Catering to Ignorance

by Russ Roberts on August 28, 2006

in Standard of Living

I suppose it’s a dog bites man story—I shouldn’t be upset when the New York Times news division writes an intellectually dishonest story that plays to the biases of its readership base. But today’s front-page above the fold story on wages depresses and surprises me anyway. Maybe it’s because one of the authors, David Leonhardt, is a good reporter with good economic intuition. (I can’t speak for the other author, Steven Greenhouse.) But I suspect the source of my dismay is simply the knowledge that this article, despite its inadequacies will be met with nods of agreement around the breakfast tables of America.

Here’s how the article opens:

With the economy beginning to slow, the current expansion has a
chance to become the first sustained period of economic growth since
World War II that fails to offer a prolonged increase in real wages for
most workers.

That situation is adding to fears among Republicans
that the economy will hurt vulnerable incumbents in this year’s midterm
elections even though overall growth has been healthy for much of the
last five years.

The median hourly wage for American workers
has declined 2 percent since 2003, after factoring in inflation. The
drop has been especially notable, economists say, because productivity
— the amount that an average worker produces in an hour and the basic
wellspring of a nation’s living standards — has risen steadily over the
same period.

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?

3. There are no government series that I know of on median earnings. Where did those data come from?

There’s a chart accompanying the article. It tells the reader that the median hourly pay data are from the Economic Policy Institute. The Economic Policy Institute has a policy agenda. Their main issue is the alleged stagnant or falling standard of living of American workers. They support a higher minimum wage and the strengthening of labor unions.

The Bureau of Labor Statistics does calculate real hourly compensation for the nonfarm business sector, a measure that includes benefits. Let’s see what those numbers look like:

What these numbers show is that 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.

Any of these measures are at odds with the Times’s conclusion. It would have been nice for the Times to include a measure like this as a counterpoint but it’s missing.

The article continues:

As a result, wages and salaries now make up the lowest share of the
nation’s gross domestic product since the government began recording
the data in 1947, while corporate profits have climbed to their highest
share since the 1960’s. UBS, the investment bank, recently described the current period as “the golden era of profitability.”

the last year, stagnating wages were somewhat offset by the rising
value of benefits, especially health insurance, which caused overall
compensation for most Americans to continue increasing. Since last
summer, however, the value of workers’ benefits has also failed to keep
pace with inflation, according to government data.

There are two claims here. The first is that labor is getting an increasinly small share of the economic pie. The second is that taking account of compensation (which the authors have ignored till now and are mentioning here in passing) isn’t enough to keep workers ahead of inflation.

Both of these claims are puzzling. The first claim, about labor’s share of the pie ignores benefits. 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, as this St. Louis Fed report shows:


Yes, there are some ups and downs. Yes, it may be the golden age of profitability. And yes, CEO and upper management compensation may mean this is a distorted picture of how the average Joe is doing. Yes, these data end in—it looks like—2003.

But it’s hard to use these data to support the claim of the Times that labor’s share is at an all-time low relative to 1947.

The second claim about recent years is hard to square with the BLS data on real hourly compensation which shows that real compensation is growing not shrinking. Later on in the article, the authors make a definitive statement about compensation including benefits:

Total employee compensation — wages plus benefits — has fared a little
better. Its share was briefly lower than its current level of 56.1
percent in the mid-1990’s and otherwise has not been so low since 1966.

This percentage, taken from Commerce Department data, is the ratio of compensation to GDP. If this ratio is indeed lower today than it was in 1966, my guess is that it isn’t much lower and that it tends to bounce around very little like the data I show above.

The second point is that my quick look at the Commerce Department numbers shows that total compensation received by employees as a ratio to GDP is HIGHER today than it was in 1966. I’ve emailed the Times. I’ll let you know what they say in reply if it’s relevant.

The article discusses the reasons for the alleged wage stagnation:

Economists offer various reasons for the stagnation of wages. Although the economy continues to add jobs, global trade, immigration, layoffs and technology  — as well as the insecurity caused by them —  appear to have eroded workers’ bargaining power.

unions are much weaker than they once were, while the buying power of
the minimum wage is at a 50-year low. And health care is far more
expensive than it was a decade ago, causing companies to spend more on
benefits at the expense of wages.

I would love to see a measure of workers’ bargaining power. What could that possibly be? Unions have been weaker for fifty years, a time of great increases in our standard of living.

What keeps my wages high (and yours) is our alternatives. Is there any evidence that workers have fewer alternatives than they once had? If anything, workers are more mobile today than ever. What sets workers wages are the wages of those alternatives. That depends, generally, on our skills, our knowledge and our drive and discipline—human capital and how well we are able to apply it. Workers are better educated than ever. That is why I believe that compensation, properly measured, is higher than it was five or ten or twenty or thirty years ago.

By the way, the New York Times quotes one economist in this entire article. That would be Jared Bernstein of the Economic Policy Institute.

Update: Greg Mankiw takes a calmer look at the whole question of wages and productivity here.

Be Sociable, Share!



Add a Comment    Share Share    Print    Email


Mike August 28, 2006 at 2:17 pm


I'd also like to note that this article and others like it obsess with the difference between average wages (which are rising rapidly) and median wages (which they claim are not rising); but when they talk about productivity they are only talking about average productivity. In fact it is virtually impossible to measure median productivity.

Workers are paid their productivity. If average productivity increases, and if this is driven by increases at the upper tail of the skill distribution (as economic evidence has found), then why would we be surprised that high skilled folks are gaining more than less skilled folks? If in fact that is actually happening?

John Dewey August 28, 2006 at 2:23 pm

"If anything, workers are more mobile today than ever."

Replacing defined benefit retirement plans with defined contribution plans certainly helped worker mobility. So has the vast amounts of company and job information available via the internet.

John Dewey August 28, 2006 at 2:29 pm

If the income of all other workers had stayed the same, the addition of a few million low-skilled immigrants should have driven down the median income the past few years. Many immigants are doing jobs that formerly received no compensation – such as lawn care formerly performed by homeowners.

save the rustbelt August 28, 2006 at 2:46 pm

"…real hourly compensation has actually increased. It's up since 2003 as well. And this year it's up quite dramatically….."

Any charts on the distribution, raw averages tell us very little.

Peter Duray-Bito August 28, 2006 at 4:28 pm

The productivity miracle loses its sparkle when when measured against the need to work longer until retirement.

Chuck August 28, 2006 at 4:49 pm

Mr. Roberts,

If you are looking for the Data on 'Median Wages per Hour' check the BLS Occupational Employment and Wage Estimates. You'll find a nice breakdown of wages by every industry and income percentile. I used this data today, and it only goes back to 2003(which is why the NYtimes authors don’t use 2001 data) If you look at hourly wages of the 10th and 25th percentiles of workers, you will see that annual increases earnings are below core-inflation, and well below total inflation. Your ignorance discredits your arguments. Better luck next time.

Ivan August 28, 2006 at 4:50 pm

"The productivity miracle loses its sparkle when measured against the need to work longer until retirement."

Just compare it to rising life expectancies, and you'll get your sparkle back.

Retirement age was set at life expectancy when Social Security started.

I see no reason why people feel they have the right to not work for the final quarter of their lives. I especially don't get comments like yours, given the time spent in retirement is only increasing.

David Zetland August 28, 2006 at 5:08 pm

This looks like another tit for tat argument discussed in depth in Econospinning (, which I reviewed at Amazon. I recommend it — especially if you are into labor data issues!

David Zetland August 28, 2006 at 5:09 pm

Previous attempt didn't take…

This looks like another tit for tat argument discussed in depth in Econospinning (, which I reviewed at Amazon. I recommend it — especially if you are into labor data issues!

Mr. Econotarian August 28, 2006 at 5:31 pm

"If the income of all other workers had stayed the same, the addition of a few million low-skilled immigrants should have driven down the median income the past few years."

No, it should have driven up the median income. Not the median income of the U.S., but the combined median income of the U.S., Mexico, etc.

Immigrants with low wage jobs in the U.S. came from much, much lower wage jobs outside the U.S. Their median wages went up a lot when they crossed the border.

Aaron Krowne August 28, 2006 at 5:58 pm

Real median compensation is only a rounding error's worth of growth, by the BLS's own numbers.

The trend towards more family members working actually casts doubt on whether the real improvement is upwards.

However, even with this criticism aside, I have severe reservations about whether the CPI is really so honest. Hedonic adjustments don't belong in it at all. Neither do substitution adjustments, but geometric weighting to approximate them was never even justified empirically.

I also wonder what free cash is left over after the typical level of debt service.

I think the reality for median compensation is more bleak than the BLS data would suggest on their face. This jibes with my daily experience.

EclectEcon August 28, 2006 at 6:26 pm

Whatever happened to the media standard of checking with several sources before running a story? I'll bet if you had fed your take on real compensation to NYTimes reporters, they'd have gotten some opposing views from other so-called economists.

Query: how long will it be before Paul Krugman publishes a column that ignores the points you made in this posting?

olivier blanchard August 28, 2006 at 9:00 pm

Pow. Way to fact check, baby! Rock on.

alex August 28, 2006 at 9:13 pm

Firstly, although wages compared to GDP is still 70%, more of that is going to CEO's and other decently compensated Joes (in the millions of dollars each). So yes, wage comp is about the same, buts its not being given to your average American. Secondly, on unions; don't want to rain on the parade, but median family income rose by 2.9% in real inflation adjusted terms annually from 1947-1973. From 1973-2000, they rose by less than .5%. Quite a difference, huh? And I am sure it has nothing to do with unions too. Also, on productivity gains, you would do well to read a paper by two North Western econ professors called Where Did the Productivity Gains Go? Maybe you will find out why productivity doesn't explain how people are doing these days. Nice to see that you are still going after the small errors, and totally missing whats actually going on.

Russ Roberts August 28, 2006 at 10:14 pm


I'll check out those data but I spoke to the BLS today and I was told that the BLS does not have any median hourly wage data. Maybe the person I was speaking with assumed I meant going back farther.

Actually, the data you mention to appear to go back to 2001:

But the BLS has a lot of different measures of compensation some with cheerful findings for wage growth and some gloomy.

The Times article did not use those data, it turns out.

I mentioned the date discrepancy to make the point that if you're talking about the behavior of compensation or wages since the end of the last recession, you'd want to start at the end of the last recession.

As for those wage data and the 10th and 25th percentile, you should remember that they are not the same people today as in 2006. An immigrant for example, who was in the 10th percentile in 2003 may be better off today and his place in the 10th percentile could be taken by a different immigrant with lower productivity and skills and even lower wages. I also think the CPI overstates inflation.

So I'm not as sure as you seem to be that people today are worse off than they were in 2003. I'm even less sure of why that would be true or what to do about it if it is true. But I'll write more on that latter point another time.

happyjuggler0 August 29, 2006 at 11:50 am

John Dewey is right, low skill immigration skews the median.

Imagine you have a American compensation pool of 3x 4x 5x 6x 7x 8x 9x, with x being some amount of dollars.

Both the median and the mean average is 6x.

Imagine that ten years later you have absolutely no improvement in compensation for those same people and thus they are 3x 4x 5x 6x 7x 8x 9x. But due to low skill immigration from Mexico you also have two new people, and they both make 3x. Now the pool looks like 3x 3x 3x 4x 5x 6x 7x 8x 9x.

The median income has now fallen to 5x from 6x! The mean income is a bit better than the median at 5 1/3, but it is still significantly worse than it was ten years ago. Yikes, call the NY Times, inequality is increasing *and* the overall wage is going down as well. Looking at the change in mean and median it also become "clear" that the rich are getting a bigger share of the national income. Those damn CEO's and sports stars, there oughta be a law!

But the simple reality in my example is that all the original Americans had neither worse incomes nor better incomes, they were exactly the same. What changed was that we added workers to the bottom of the pool, and they pulled down the averages without harming anyone. Indeed, it turns out in my hypothetical example that these Mexican immigrants used to earn 0.75x back in Mexico and their incomes thus rose 4-fold. Thus, the horrible picture outlined in my hypothetical NY Times lead story is actually a success story for the actual people on the bottom of the ladder (well, for two of them). But the NY Times says it is bad for the little guy and great for those damn rich people.

To make a long story short, if we have had scads more low skill/ low wage immigrants than we used to in 1973 (and we do), then where born-in-the-USA Americans had stagnant wages it would look like a large fall in median wages and also mean wages, along with an increase in inequality skewed towards the rich.

But if the measured median looks stagnant during a time of heavy low skill immigration, then in reality those born-in-the-USA Americans are by definition doing significantly better even though you'd never know it from the stats.

Steven Vickers August 29, 2006 at 1:39 pm

Chuck, if you check the BLS website: , it's seems they have series going back well before 2003.

But Roberts is kinder than I am. The Times statement that Roberts disputes is: "The median hourly wage for American workers has declined 2 percent since 2003, after factoring in inflation." The idea that you can refute a statement about the median by referencing the 25th and 10th percentiles is really, really wrong.

And you then cap off that mind-jarringly incorrect claim by referencing Roberts "ignorance." Brilliant.

neil August 29, 2006 at 2:11 pm

Vickers, it's not really too much different from Roberts using mean income data to refute a claim about median income.

Steven Vickers August 29, 2006 at 2:37 pm


I would agree, except Roberts wasn't refuting the claim about median income, rather the conclusion drawn from it. That's a far cry from a mathematically incorrect statement.

mike August 29, 2006 at 3:09 pm

From today's WSJ article on DuPont's plan to close its defined benefit plan …

"Mr. Porter said the beefed-up 401(k) plan won't fully replace the company pension benefits, but added that if employees get unusually good investment returns in their 401(k) accounts, they could be better off than under the company plan."

Patrick R. Sullivan August 29, 2006 at 3:34 pm

Take into consideration that the health insurance benefit isn't subject to income tax, but wage increases are, and we see that total compensation is being understated.

spencer August 29, 2006 at 4:44 pm

Russ for your information the median data is from Census and it is not difficult to find. Census is the agency that looks at poverty and regularly publishes many series using median rather then average.

Just today they said: Real median earnings of males age 15 and older who worked full-time, year-round declined 1.8 percent between 2004 and 2005, to $41,386. Women with similar work experience saw their earnings decline by 1.3 percent, to $31,858. (The difference between the rates of decrease for men and women was not statistically significant.)

Let me see now, you did not like the NYT article, but rather then use the data and the sstatement they used you went out and found a completely different series to demonstrate they were wrong.

What they said was right. What you said was something different. You can make an argument that what you said is a better measure, but you can not use it to demonstrate what they said was wrong.

ScottM August 29, 2006 at 4:45 pm

The reason I (and a few other non-economists) don't like to count "total compensation", is that the compensation changes seem to include higher contributions and co-payments for our health plans, while benefits are declining.

If our "compensation" is just profit for a different middleman, why should we feel rewarded, or like out wages aren't declining?

ScottM August 29, 2006 at 4:45 pm

The reason I (and a few other non-economists) don't like to count "total compensation", is that the compensation changes seem to include higher contributions and co-payments for our health plans, while benefits are declining.

If our "compensation" is just profit for a different middleman, why should we feel rewarded, or like our wages aren't declining?

ChrisG August 29, 2006 at 5:35 pm

Don't ignore the Employment Cost Index as a way of measuring total compensation–Nonfarm Real Hourly Compensation is certainly not the only statistic out there. In fact, there is much debate among Labor Economists as to which is the more appropriate measure, and no clear cut answer.

The ECI Compensation series, after being deflated by the CPI, shows that real compensation has been falling since early 2005, and currently at about 2003 levels. As you show, the Nonfarm Hourly Comp. statistic shows it to be rising fairly continuously. So which do we believe? The main difference between the series is the benefits they choose to count. The ECI only includes the basics like health insurance, while the more inclusive Hourly Comp. statistic adds in stock options and the like. As a result, the CEOs are going to carry more weight in the Nonfarm Hourly Comp. measure, and the ECI index is going to be more representative of the average person.

I won't attempt to argue that one is more important than the other for assessing the health of the economy. But no matter which you prefer, the fact that they are moving in the opposite direction does suggest a growing disparity in the different classes of workers–which is more or less the point of the NYT article in the first place.

Fetal Farmer August 29, 2006 at 6:47 pm

High wages suck. Unions are communistic. Wal-Mart rules!

Teri Pittman August 29, 2006 at 7:33 pm

You've chosen to ignore the decline in benefits as well. Most folks now pay more for their health insurance. Pensions are a thing of the past. We get fewer and fewer paid holidays off. I don't get any paid holidays until I've worked for a year at my current job. Many workers have to work a year before they get any paid vacation time. Many also don't get paid sick leave. And when you do get these benefits, you are often expected to put in mandatory overtime.

Bruce Hall August 29, 2006 at 8:19 pm

Median or mean income is an interesting concept, but the dispersion of the data or skewness of the curve are a little more difficult to come by and maybe a little more telling.

I think another statistic from BLS comes under the category of employment. For example, here in Michigan, both employment and unemployment charts reveal a lesser than rosy picture.

Compare this with Florida and you begin to see the meaninglessness of many national statistics… much less median wages.

Kind of like describing the "average" fish in the ocean.

python August 29, 2006 at 8:40 pm

When I read or hear arguments about how bad the economy or working conditions are in the US, I always ask 2 questions: 1) Where is it currently better than the US? 2) When in history was the US better than it is now?

All those that dispute Dr. Roberts comments, or side with the NYTimes article, feel free to answer.

Bruce Hall August 29, 2006 at 9:32 pm

Python, while not many countries may have more per capita wealth than the U.S., there are, arguably, those with higher overall qualitative environments.

Regardless, one could make the case that 5 years ago, before the Federal Reserve decided to create a recession, that the general attitude and "feeling" of prosperity was higher than it is now. Certainly, the "dot com" bubble had something to do with that… just like the "housing bubble" has something to do with the apparent well-being of our economy.

I suggest that you hold your argument for about three months when more home sales/price data is reported. Consumer confidence is diving and the "all is well" mantra may be getting a little strained.

Python August 29, 2006 at 10:17 pm

Bruce Hall,

I certainly don't have a mantra that "all is well", rather, "show me what you got that's better". Because if you have something better I'm all for it.

But, from my experience, critics seem to look back at history or other countries and try to borrow piecemeal from them. For example, a common criticism is that life used to be more simple and now it's too competitive/busy/complex, etc. But when you clearly explain to them what life was like, say, in the 1950s when very few people could afford to travel for leisure, when there was no Internet, no color TVs, no microwaves, and you basically worked until you died, etc., you will find few people who wish this was the 1950s.

Or if you want to say Luxembourg is currently better and that the year 2000 was better, than say it. But you didn't say it, and I ask you and others to think about it and answer without equivocation.

Bruce Hall August 29, 2006 at 11:10 pm

"if you want to say Luxembourg is currently better and that the year 2000 was better, than say it. But you didn't say it, and I ask you and others to think about it and answer without equivocation."

The U.S. is the wealthiest nation in the world… better.

The U.S. has the greatest number of jailed citizens in the world… worst.

The U.S. has the most immigrants in the world… better.

The U.S. has marginal health care and housing for a quarter of its population… worse.

The attitude and economy for most citizens was better in 2000 than it is today; there is a greater economic separation between those that control companies (not necessarily own them) than those who are the "average" worker (white/blue collar) in those companies.

Luxembourg is better… but that is meaningless because it is statistically irrelevant. Norway has a higher per capita income, but they also have lousy winters.

Those are unequivocal answers, but so what? Your comparison of 1950 technology to today is irrelevant because you can make the same argument in 25 years or 50 years if you are still around.

A society that is only measured by consumption is historically irrelevant. But if consumption is the key to your argument, then the U.S. is indeed the world leader. Complexity is a good measure of capability and the U.S. is still the most complex and capable nation in the world.

But there are some disturbing undercurrents that are developing. The cost of education is increasing significantly faster than incomes meaning that those with lower incomes will be less able to get the education necessary to move up the economic ladder. And the higher paying industrial jobs that used to allow families to without higher education to live reasonably well are disappearing. So, more people are going to moved to the economic sidelines as the ability to earn money is reduced while the opportunity for the education they need to get the jobs that will earn higher incomes is priced out of their abilities to pay.

Is that unequivocal enough?

Come to Michigan if you think things really are better. You can have your choice of thousands of homes for sale that have been on the market for more than a year or two as people who have lost their jobs because their former companies are going out of business attempt to leave. Nope, things are not better.

Is that unequivocal enough?

But, gee, the price of poorly made items from China is still low. So I guess that makes us better off. No, I feel some equivocation about that.

The neat thing about some schools of economic theory is that they can remain "theoretical"… far enough in outer space that the dirt doesn't show up.

JohnDewey August 30, 2006 at 8:52 am

Bruce Hall: "Come to Michigan if you think things really are better."

No question that some regions have lost while others have gained. Nissan didn't put its truck plant in Tennessee because of the strength of the local unions. BMW didn't choose Spartenburg, SC, because of all the union expertise there.

If you come to Charlotte or Nashville or Dallas or Phoenix or a dozen other Sun Belt cities, you'll find that things are definitely better.

The median household income in Dallas-Fort Worth has dropped 10% in real dollars since 1999. So how can we say things are better? As Univ of North Texas economist Bernard Weinstein has explained: "The lion's share of growth in this region is going to be among immigrants and low- to- moderate income minorities." Brad Shanklin, of the Plano Chamber of Commerce, explains that high-wage jobs migrated hear in the 90's, and the need to expand the service sector followed shortly after.

Dallas professionals, other than some in telecom and airlines, have realized wage increases and more opportunities than ever before.

JohnDewey August 30, 2006 at 9:01 am

Bruce Hall: "But, gee, the price of poorly made items from China is still low."

What items from China are poorly made? The golf shirts I buy on sale at Kohl's for $8.00 fit just as well as those made in the U.S. 30 years ago. Those shirts seem to last just as long.

This HP Notebook PC I'm typing on was shipped from Shanghai last year, so I assume it was built there. I haven't had any problems with it at all.

What Chinese items have you had problems with, Bruce?

Morgan August 30, 2006 at 10:12 am

Bruce Hall and Save_the_Rustbelt have both requested more detailed data regarding the distribution of income. While the census is probably the best source, it only comes around every decade.

So, in the meantime, check out the IRS website:

IRS releases data regarding the number of returns filed with adjusted gross income falling within various ranges. AGI may not be exactly what you're looking for, but it's a reasonably good proxy. For the most part, the most recently released data is from 2003 (returns filed in 2004). There is at least one table that includes preliminary data for 2004:

Bruce Hall August 30, 2006 at 1:03 pm

Geez, John. Latch on to one minor remark and miss the point….

But to your point:
* pooly made tools with inferior metal
* poorly made kitchen appliance with substandard performance; e.g., toasters that heat unevenly
* poorly made furniture that looks good, but constructed with bad joints and materials that don't last
* poorly made electronics that look good, but deliver sub-standard performance and longevity.

But I'm glad your shirts last well.

What you deem affordable, I call cheap… not the same as inexpensive.

My brother-in-law runs a construction related business in Florida. He visited last month and said that business is off significantly this summer.

I've spoken to family and friends in New York and Washington D.C. where the housing markets have plummeted.

Good for a few remaining spots where retirees are still fueling growth. But even retirees are thinking twice about moving when they can't sell their existing property.

Regarding earnings:

We seem to be selling our futures for cheap goods. Since affordable higher education and well-paying lower-skilled jobs are relics of the 1950s, what magic economic solutions are left for those without the fortune to be born into well-off families?

John Dewey August 30, 2006 at 1:59 pm

Bruce Hall,

My toaster was made in China. It works just as well as any I've ever owned.

If Americans are unsatisfied with the quality of items, they'll stop buying them. But, like me with my Kohl's golf shirts, American consumers are satisfied with their WalMart purchases. They keep going back for more.

Bruce Hall: "Good for a few remaining spots where retirees are still fueling growth."

It's not just a few spots and it's not fueled by retirees. The Commerce Dept reported today that U.S. GDP grew at 2.9% in 2Q06. It had grown by 5.6% in 1Q06, and by 3.2% annually in 2003-2005.

GDP is a production measure, so it does not include retirees' expenditures of savings or government transfers.

Bruce Hall: "We seem to be selling our futures for cheap goods."

How so? The link you provided includes another link to the Census Bureau's Current Population Report. It shows the 2005 median household income, in constant 2005 $, was $11K above 1965 and $4K above 1985. GDP and median incomes would not be growing if our nation were in decline.

Bruce Hall: "My brother-in-law runs a construction related business … business is off significantly this summer. …New York and Washington D.C. where the housing markets have plummeted."

Housing has been a boom and bust industry for all my lifetime. This coming bust will seem larger than normal only because the boom was so big. But it won't change the long-term strength of the U.S. economy, which keeps generating more and more jobs.

Python August 30, 2006 at 2:09 pm

Based on this data: linked by Morgan: in 2004 compared to 2002 there were 1.1 million FEWER people making less than 50k, and 3.4 million MORE people making more than 50k.

I ran some quick numbers, and it doesn't seem that inflation alone could account for that increase. Either way, it's too close to be able to make some federal case about how poor our economy is (although there may be other numbers that make a stronger case). And anyway, their dates are not relevant as Dr. Roberts has already suggested.

And someone mentioned housing. Housing in almost all areas of the country is much, much, much more valuable than it was 5 years ago. ; 57% on average higher since 2001 in the US as of June 2006, if you believe their data.

Bruce Hall, please read point #6 on page 2 of that last link.

David Leonhardt August 30, 2006 at 3:15 pm


I wanted to send you a reply to your blog entry that you should feel free to post.

First, thanks for your kind words about me, even in the midst of criticism. As I think you know, I'm generally a fan of your work, and your opinion means a good deal to me.

But I do disagree with you here. We stand by all the numbers in the story, and your posting contained a few errors. I get into detail below.

The larger question, however, is whether an objective analysis of the economy can show the last few years to have been a period of healthy growth in compensation for most workers. I think the answer is no. Earnings for the median-earning worker — indeed, as we noted, even for the worker at the 90th percentile — have trailed inflation over the last three years, according to the B.L.S series on Usual Weekly Earnings of Wage and Salary Workers. Benefits outpaced inflation between 2003 and 2005, but they've trailed inflation over the last year, according to the Employment Cost Index. Flat real compensation for most workers is not what today's incredible American living standards were built on.

I also don't think it's an especially controversial notion that much of the workforce has less bargaining power than it once did. To generalize, perhaps overly, globalization and technological innovation have raised the value of the work done by the roughly one-quarter of adults who have bachelor's degrees and reduced the value of the work done by the three-quarters who don't. As you note, this is about alternatives. Employers have many more alternatives than they once did to paying an American $20 an hour for blue-collar or basic service-industry work. Alan Greenspan and Ben Bernanke have both addressed this trend in some detail. Goldman Sachs and UBS have written reports about the connection between it and the boom in profits.

As for the longer-term, our numbers don't disagree, though our interpretations do. Both your St. Louis Fed numbers and our Commerce Department numbers show wages and salaries to make up the lowest share of the economy on record. [Thursday’s G.D.P. report now shows a rise, though] They also show the total-compensation share to be near a 40-year low. You consider the wage-and-salary number irrelevant; I consider it to be one part of the story. You consider the recent decline in total compensation to be trivial; I think a few percentage points matter when each represents more than $100 billion. But I understand your arguments here.

Now for those details:

Your posting implied, incorrectly, that the story had said that compensation's share of the economy was lower now than it was in 1966. In fact, we said that except for a short period in the 1990's, the share "has not been so low since 1966." We agree: the share was lower in 1966 than now.

You asked for detailed sourcing on our share numbers. The total-comp and wage-and-salary numbers come from Table 1.10 of the B.E.A's NIPA tables (lines 2 and 3, respectively). They make up the numerators of our share numbers. The GDP number, which makes up the denominator, comes from line 1 of Table 1.5.5.

You wrote, "There are no government series that I know of on median earnings." The BLS series I mentioned earlier in this email is exactly such a series. It's available here: You also said you you didn't trust the Economic Policy Institute's analysis of the B.L.S.'s data. If the E.P.I. data on median wages is flawed, I am quite open to hearing what those flaws are. But we can't just rely on averages, as opposed to medians. Averages don't paint a detailed enough picture.

Finally, you write that we quote "one economist in this entire article. That would be Jared Bernstein of the Economic Policy Institute." That's wrong. The piece also quoted an economist named Ben Bernanke. And it closed by quoting Treasury Secretary Henry Paulson, who's not an economist but is a pretty smart guy.

Enough of this, though. Let me close on a note on which I know we agree: why oh why can't our beloved Red Sox get some better pitching?


John Dewey August 30, 2006 at 4:33 pm

David Leonhardt: "To generalize, perhaps overly, globalization and technological innovation have raised the value of the work done by the roughly one-quarter of adults who have bachelor's degrees and reduced the value of the work done by the three-quarters who don't."

I think this is an over-generalization.

Globalization hasn't impacted the work of millions of non-degreed workers: health care workers, cooks, firefighters, and real-estate agents, for example. Globalization has increased the demand for transportation workers of all sorts.

Technological innovation eliminated some blue-collar jobs, but it enhanced others and created even more. Once trained in computer control methods, plant operators become more difficult to replace. Competent computer repair and network technicians are not going to work for peanuts.

I'd have to see much more hard evidence before accepting Mr. Leonhardt's conclusion about the value of non-degreed workers.

Geoffrey Transom August 31, 2006 at 2:02 am

Typical focus on the income statement – you lot would make good "street" stock analysts.

BALANCE SHEETS, gentlemen.

Consuming one's seed corn (and borrowing more seed corn and consuming that too) can provide a SEMBLANCE of prosperity, but unless you're willing to be a debt defaulter, the piper must be paid eventually… either through massive inflation (which is the usual way governments attempt to defraud their debtors) or through a lower standard of living for the young'uns.

And besides – getting back to flows – all of the gains (aggregate dollars in the pocket) of the entire workforce have gone to the top 0.1% – since 1973. EVERYONE ELSE – in aggregate – has gone backwards. That is not an indication of a sustainable model, and so it should be no surprise that the US is flailing about trying to garner control over resources through (debt-financed) force. that is what dying quasi-Empires DO. That's because politicians have never understood that once your enemy has anything more advanced than pointy sticks, the yield of resource extraction through military force becomes negative – and does so rapidly. All the low-hanging fruit of Empire were harvested before the 1500s, and from there it was all downhill.

And before anyone babbles about migration from the bottom half to the top decile, forget it. It's a myth that happens so vanishingly rarely as to be irrelevant – it's like George Bush inventing a coherent sentence all by his own self. Might happen once in a blue moon, but don't base your life's plans on it.

And no, I'm not a rabid Leftie… I am a capitalist to the core. That's why I get annoyed when the US refers to itself as capitalist… If the Yanks HAD free markets they would understand what a good thing they are. Instead they have – and have had since Teapot Dome – a crony capitalist system reminiscent of Suharto's Indonesia.

If you honestly believe that American working class is, on balance, experiencing INcreasing prosperity, then the coming housing-and-debt led disaster will probably surprise the living hell out of you. But that's a balance-sheet issue, n'est-ce pas?

Academics (or would-be same) love to concentrate on FLOWS (and to hell with how they're measured… let's call my new Dell TEN computers for the purposes of GDP)… but STOCK relationships are what causes the intertemporal budget constraint to bite…

Sure, this screed is only marginally related to the core topic (worker compensation), but really the core isue in an economy is whther the system enables significant increases in STOCKS (wealth-type variables) in the household BALANCE SHEET, not the status of FLOWS (as measured by the Government, whose tendency not to tell any bad stories is renowned).


France (where it is raining today and they don't properly understand TOAST)

JohnDewey August 31, 2006 at 7:14 am

Geoffrey Transom: "And before anyone babbles about migration from the bottom half to the top decile, forget it. It's a myth that happens so vanishingly rarely as to be irrelevant"

Do you have any statistics about this, Geoffrey? Your statement is certainly not consistent with the world I've observed.

Many of my fellow Wharton MBA's grew up in the bottom half, as I did, yet almost all are in the top decile today, two decades later. Many small business owners I've met came from low income families, yet now have considerable net worth.

Here's an excerpt from the 26-Sep-2003 Economic letter of the Federal Reserve Board of San Francisco ("Earnings Inequality and Earnings Mobility in the U.S."):

"Upward mobility remains an attainable goal for the majority of working age individuals. The presence of mobility implies that yearly measures of earnings inequality likely overstate the permanent earnings differences between individuals."

I've read another study showing that movement from the bottom income quartile to the top quartile is common. I can't find it just this morning, but I'm sure I will. Why should I believe that the top decile is any more closed than the top quartile?

JohnDewey August 31, 2006 at 8:14 am

geoffrey transom: "all of the gains (aggregate dollars in the pocket) of the entire workforce have gone to the top 0.1% – since 1973."

Those gains, measured either in flows or stock, did not go to the same people that comprised the top 0.1% in 1973. Tiger Woods wasn't born in 1973. Bill Gates 1973 income and wealth were a tiny fraction of what he enjoys today. Mark Cuban had just entered high school. Fred Smith's Federal Express, in which he had invested everything, was nearly bankrupt in 1973.

Certainly we can find cases where inheritance accounts for the wealth some enjoy in America. But most wealth in this country is still the self-made variety. High incomes – especially for athletes, physicians, and entrepreneurs – most definitely has been earned.

Dan Sullivan August 31, 2006 at 11:05 pm

Mr. Roberts,

I have no idea whether you are still reading comments on this post. But, just in case, I'll point out that you were just a couple of mouse clicks away from a series on labor's share that is (very nearly) at an all time low (or at least a low since 1947).

Specifically, when you were looking up the numbers for real hourly compensation in the nonfarm business sector, you just needed to scroll down to the last entry in the list of available series to get data on labor's share. (I'm assuming you were using the "one screen applet," . Alternatively you could plug PRS85006173 into series report .)

The data on labor's share in the nonfarm business sector, which go back to 1947, hit a series low in the first quarter of this year. They moved up very slightly in the second quarter.

Arguably, these data are more to the point of recent discussions on where productivity gains are going than the NIPA data you focussed on above. Relative to the NIPA data, they exclude the output and employee compensation of goverment, nonprofits, and owner-occupied housing (the sectors of the economy whose output is especially hard to measure). The nonfarm business sector is the part of the economy that most discussion of productivity refer to. It is the series whose growth rates you showed in the table divided by a measure of the total output of the corresponding sector.

Previous post:

Next post: