Inflation calculations: backwards or forwards?

The New York Times has a story on hourly wages that states:

The $20 hourly wage, introduced on a huge scale in the middle of the last century, allowed masses of Americans with no more than a high school education to rise to the middle class. It was a marker, of sorts. And it is on its way to extinction….

Hourly workers had come a long way from the days when employers and unions negotiated a way for them to earn the prizes of the middle class — houses, cars, college educations for their children, comfortable retirements. Even now a residual of that golden age remains, notably in the auto industry. But here, too, wages are falling below the $20-an-hour threshold — $41,600 annually — that many experts consider the minimum income necessary to put a family of four into the middle class….

I'm curious to know who those "many experts" are; I'm not familiar with the notion that $41K is the entry level to the middle class, and in hourly wage households, at least AFAICT, it's generally assumed that the wife works.

Belle Waring asks:

Surely $20 an hour in the 70s would be $60 or so an hour now, adjusted for inflation? It makes a big difference to this article and the author has totally failed to explain the issue. ‘Fewer people of this class make even 1/3 as much per hour as they did 30 years ago’ is a very different message from ‘fewer people of this class make this inflation-adjusted wage.’ It seems clear the article implies the former but muddies the waters with the nominal wage, ironically further masking the dramatic decline of the blue-collar middle class.

Er . . . she's got it backwards. They are inflation-adjusting the $20 wage--back to the 1970s. A $20 an hour wage in 1973 would be about a $100 an hour wage in today's dollars. Even if you think that the bygone industrial days were a halcyon era of plenty, this should flunk an elementary gut check. $100 an hour translates into an annual salary of about $200K a year in today's money. Yet per-capita GDP has more than doubled since 1970.

I think the story's more than a little silly--hourly wages just aren't a very good way to organize most non-industrial workplaces, so enhanced manufacturing productivity sort of definitionally means that they'll fall. But it's not quite as silly as Belle seems to think.

Comments

I don't understand - why would hourly wages fall as manufacturing productivity increased? The labor share is roughly constant, so the return to labor tends to go up as productivity increases.

Per Capita GDP is highly misleading, by the way. I have a paper (still under submission) where we review a lot of inequality data, and even 95th percentile (full-time, full-year) wages have rises less since 1973 than per capita GDP or per capita wage and salary earnings (excluding benefits). There are two reasons: the number of hours worked has increased substantially as women enter the labor force, and the wage distribution has really widened an incredible amount. The median worker, despite Cato's Reynolds proclamations to the contrary, has basically not moved since the late 70s in real terms. I'm as libertarian as they come, but there's no question that policy decisions have massive effects on income distribution, given relatively stable inequality in many other advanced economies over the last three decades.

Interesting fact: If median wages had risen at the same rate as the labor share of GDP over the past three decades, the median worker would earn over $80000 today (as opposed a bit less than half that).

Posted by cure | April 20, 2008 11:14 AM

Cavils aside, I think the base concept for the article is correct.

At one point, an American with a half-reasonable work ethic but no ability or interest in learning higher skills could attain a life-style that put him in the middle class.

That (increasingly) is no longer the case.

Given that a significant (close to majority?) proportion of the population is in the category of not capable of (or interested in) pursuing high-value training (i.e. they've had all the schooling they're willing to tolerate by the end of high school), how does society cope with the loss of much of its middle class?

How does Western civilization cope?

A question I've been pondering for the last 25 years. Haven't come up with an answer yet.

Posted by Tom West | April 20, 2008 11:17 AM

Tom, I think the real question is how does Western civilization cope with a loss of rhetorical questions.

Posted by NutellaonToast | April 20, 2008 11:34 AM

Tom West, where is the data that supports your contention that this scoiety is losing "much of its middle class?"

The data I've seen suggest that people at the very top of the income distribution have been seeing their incomes skyrocket, and that people in the middle have seen their wages stagnate though their total compensation has risen over the last few decades. And that stagnating real wage might be misleading, not only because it fails to include benefits but also because inflation tends to be overstated and thus growth in the real wage tends to be understated.

Meanwhile, median household wealth has risen along with asset prices over the last 25 years or so. The recent declies in housing prices notwithstanding, the "typical" family is clearly better off than it was 25 years ago.

There has certainly been a rise in inequality, but there has been no "hollowing out of the middle class" that I can see. What am I missing?

Posted by rwe | April 20, 2008 11:34 AM

I would just note that the problem Tom West identifies, the collapse of the living wage for high school graduates, has major implications for our educational debates as well. As people have noted, the income gap between those with a college degree and those without has widened and widened, but the number of people attending college (as a percentage) has not grown significantly. The standard response has been "get more people into college," which has noble intentions, but I don't think that's necessarily in anyone's best interest. Some people simply are not qualified to or interested in attending college, and not just as a function of intelligence; not everyone thrives in that environment. (I certainly think that anyone who wants to go to college should have the ability to, one way or the other, though.)

My solution wouldn't be to send everyone to college, and insert them into the "information economy" or whatever you want to call it presently. It would be to reintroduce the uneducated worker's living wage, like the Times writer is talking about here. Among other things, I think that would go a long way towards alleviating the family-breakdown problems that conservatives have become so obsessed about.

(How we would go about doing this, though, I haven't a clue.)

Posted by Freddie | April 20, 2008 11:34 AM

There are two reasons: the number of hours worked has increased substantially as women enter the labor force, and the wage distribution has really widened an incredible amount.

cure, you mentioned "excluding benefits." How large is the effect of excluding benefits? Certainly benefits making up a greater portion of total compensation would cause wages and salaries alone to fall behind per capita GDP growth. (You say wages alone, but I assume you mean both-- otherwise Megan's point about high-end wage jobs being broadly replaced by salaried white collar jobs holds.)

Another effect that works to increase inequality is education. Education, especially postgraduate education acts to decrease earnings in one's early to mid 20s in return for greater income later on. Since at any one time different people are at different stages in their life, this also increases inequality.

There's also the immigration effect on inequality. Immigration can provide an example of the statistical "paradox" where both people already here and immigrants get better off, but the number of immigrants arriving makes the national numbers look worse. (The classic example in statistics is the famous hospital with the worse overall death rate despite being better at care because it all the hard cases get sent there.) If we gave a green card to all the starving Haitians, it would make our national inequality rate worse off even if they were all considerably better off than they were before and the effect everyone else was negligible.

The median worker, despite Cato's Reynolds proclamations to the contrary, has basically not moved since the late 70s in real terms.

Well, except in terms of the statements quoted in the Times article itself, "the prizes of the middle class — houses, cars, college educations for their children, comfortable retirements." Especially in the number of cars, the median worker's household has made a very large increase in the number of cars. The size of houses and area per household member has also definitely increased. (That's another way of saying that the types of inequality does at some point matter; the difference between eating and starving is more important than the difference between eating McDonalds and at Chili's, which is more important than the difference between eating at Chili's and a fine restaurant.)

It's certainly reasonable that the increase in women working would overstate increase in GDP per capita. Homemakers provide services worth money, just not monetized into GDP. Women working outside the home is economically positive, but not to the extent that simple bean-counting would imply. If a family pays a housekeeping service $X to do services formerly performed by a member of the household when that member starts working, then the net extra value of working should take that $X cost into account.

Posted by John Thacker | April 20, 2008 11:36 AM

Tom,

Such people may form a bare majority today, but what about 40 years ago? Basically, the question being asked is why have those, who have pursued higher value training, left those, who don't, behind?

In 40 years, do you think such people will even form a majority?

Posted by Yancey Ward | April 20, 2008 11:41 AM

Meanwhile, median household wealth has risen along with asset prices over the last 25 years or so. The recent declies in housing prices notwithstanding, the "typical" family is clearly better off than it was 25 years ago.

This is a paragraph that undoes itself. In contemporary America, a middle class family's wealth is extraordinarily tied up in the value of their home. As the housing bubble has collapsed the net worth of these people has been plummeting.

Also I'd like to see the data suggesting that increased benefits have offset wage growth, or indeed that benefits have been increasing at all. Anecdotally it certainly seems to me that the death of pensions and the decline in health coverage (or increased premiums/copays) have hurt the benefits of the average worker.

Posted by Freddie | April 20, 2008 11:43 AM

"I think the story's more than a little silly"--MM

I'll second that, all this talk of 'inflation-adjusting', yet no mention of the Source of, said, mysterious Inflation?

It's a super-natural force, is it?

Posted by MEH | April 20, 2008 11:48 AM

From Freddie:

decline in health coverage (or increased premiums/copays) have hurt the benefits of the average worker

.

Do you mean a decline since the 1970s? If so, then you are committing a serious mistake in logic.

Posted by Yancey Ward | April 20, 2008 11:48 AM

Median household income, all household, for Current Population Survey 2007 is $48,000. Maybe that's where some of the tagging to a number of that general size comes from.

Posted by mike | April 20, 2008 11:55 AM

Crap. I have a PhD in engineering and earn

Posted by Matt B | April 20, 2008 12:05 PM

does the current resale value of your home even matter that much?

Only if your plan was to continue to refinance your house in order to afford it. But the question was regarding family wealth, not income, and so it's relevant to point out net worth.

Posted by Freddie | April 20, 2008 12:08 PM

Someone has asked above for evidence that benefits have grown over the last few decades as a share of total compnesation. Anyone interested should look here.

As economist Terry Fitzgerald explains:

"...benefits have become an increasingly important part of employee compensation over the past 30 years. The BLS estimates that benefits currently account for about 30 percent of employer costs for employee compensation. While the BLS does not provide similar estimates for 1975, other sources suggest that the benefit share of total compensation has risen substantially. For example, the Economic Benefits Research Institute estimates that health care as a share of total compensation rose from 3.3 percent in 1975 to 8.5 percent in 2005."

Actually, a more important finding in the article is this:

"When the data are adjusted so that they more closely measure the same conceptual object, the disparity between the microeconomic and macroeconomic statistics largely evaporates, and I find that labor income per hour for middle America has not stagnated."

So "the disappearing middle class" is really a myth. The middle class has done reasonably well in recent decades, though admittedly not as well as those at the top of the income distribution. But it is a big mistake to confuse absolute changes with relative ones. By any good absolute measure (like median wealth or median income including benefits) the middle class is significantly better off than it was 25 years ago.

Creative destruction is working for the benfit of society in general, and not just for those at the very top.

Posted by rwe | April 20, 2008 12:21 PM

And that stagnating real wage might be misleading, not only because it fails to include benefits

This is probably the most irritating dodge in the inequality-denial playbook. The median worker gets fewer benefits today than he did in 1973, not more.

Megan does seem to be correct on the inflation-adjustment point, as according to this article the average wage at GM in 1970 was $4 an hour.

http://www.time.com/time/magazine/article/0,9171,909655-3,00.html

Posted by brooksfoe | April 20, 2008 12:22 PM

why have those, who have pursued higher value training, left those, who don't, behind?

I think that skills are leveraged far more efficiently than before. Therefore the skilled are far more valuable than before.

In 40 years, do you think such people will even form a majority?

I'd hope not. But if the middle class goes from including 80% of the population to 60%, I think we're looking at massive transformations to our society, and not for the better.

Posted by Tom West | April 20, 2008 12:26 PM

The data I've seen suggest that people at the very top of the income distribution have been seeing their incomes skyrocket, and that people in the middle have seen their wages stagnate though their total compensation has risen over the last few decades. And that stagnating real wage might be misleading, not only because it fails to include benefits but also because inflation tends to be overstated and thus growth in the real wage tends to be understated.

Inflation is _understated_? 'The data you've seen?' And yet, no actual data from the guy that questions my patriotism and then runs off.

You've got some crust, there.

Posted by ScentOfViolets | April 20, 2008 12:34 PM

1) In response to my claim that the typical family is better off than it was 25 years ago in part because of asset appreciation, one commenter replies that this passage "undoes itself" and points to the recent decline in real estate values. It is probably obvious to intelligent readers what a non-sequitur that reply is. It was not my claim that people are generally better off (that is, wealthier) than they were two years ago, but that they are better off than they were 25 years ago. In the last 25 years houses and equities have appreciated enormously.

2) Another commenter, brooksfoe, tells me that benefits have declined as a share of total compensation in recent decades. The link I gave above to research from the Minneapolis Fed shows that he is not correct. Perhaps benefits have declined in the last few years as anecdotal evidence suggests, but they have clearly not fallen over the last few decades, as he contends--indeed they seem to have risen substantially over that longer time frame.

It's crucial to look at the statistics honestly and without prejudice before drawing sweeping inferences about "a disappearing middle class" and the need for radical changes in public policy that could damage growth.

Posted by rwe | April 20, 2008 12:38 PM

Change the numbers and that 1970 Time could almost have been written today. Perhaps things have not really gotten worse or changed too much at all. BTW, I remember the 70s and the amount of change from there to here continues to amaze me every day.

Posted by DCPI | April 20, 2008 12:40 PM

Well, except in terms of the statements quoted in the Times article itself, "the prizes of the middle class — houses, cars, college educations for their children, comfortable retirements." Especially in the number of cars, the median worker's household has made a very large increase in the number of cars. The size of houses and area per household member has also definitely increased.

Don't you think some data would be nice at this point? This seems to be an argument along the lines that since a consumer can purchase a computer for less than $1,000 that would outperform any machine from the mid-70's that they must be better off. Iow, are the number of cars coming out of the savings from other consumer items? And how are auto purchases being financed? Are the terms markedly different from thirty years ago? Finally, have the numbers of automobiles _per_outside_income_earner gone up, or just the number of people employed in the household.

You've got to provide a little more - heck, let's be honest - a lot more than just a bald, and very vague assertion.

Posted by ScentOfViolets | April 20, 2008 12:45 PM

rwe and Thacker,

I'm familiar with the Minn. Fed article about benefits, but there is substantial evidence that cuts the other way. I don't have my references handy, but in terms of benefits:
1) High-wage workers are much less likely to work night shifts or suffer workplace injuries compared to low-wage workers today, as compared to high and low wage workers in 30 years ago
2) Total compensation has not risen significantly for low-wage workers, but has for high-wage workers. The number of jobs offering pension and health has declined quite a bit, and nearly all of the decline has been among low-wage workers

The data I've seen leaves me unconvinced that benefits change the inequality picture.

Now, John Thacker is certainly right that composition of the labor force will affect inequality: older workers, better educated workers, and immigrant workers have more wage dispersion than the alternative, so some of the increase in inequality can be ascribed to this. This is more an *explanation* for a trend than a *description* of the trend, however.

No matter how you look at the data (CPS, PSID, Social Security, IRS returns, SCF, with benefits, after-tax, etc.), increasing inequality and very small wage growth for the median (and particularly, the median male) worker is completely uncontroversial in the economics literature.

Posted by cure | April 20, 2008 12:46 PM

In response to my claim that the typical family is better off than it was 25 years ago in part because of asset appreciation, one commenter replies that this passage "undoes itself" and points to the recent decline in real estate values. It is probably obvious to intelligent readers what a non-sequitur that reply is.

A non-sequitur is something that is irrelevant to the subject at hand. You specifically cited real-estate prices in defending your point. So, yeah, you're a fool.

Posted by Freddie | April 20, 2008 12:49 PM

Freddie,

He specifically cited real-estate prices over a quarter century in defending his point- a point that is not contradicted by your focus on the very tail end of that trend. Thus, yours is a non-sequitur. I bought my house 10 years ago, and even with the recent declines in one of the worst bubble markets, I am still way up in asset value.

Posted by Yancey Ward | April 20, 2008 12:59 PM

2) Another commenter, brooksfoe, tells me that benefits have declined as a share of total compensation in recent decades. The link I gave above to research from the Minneapolis Fed shows that he is not correct. Perhaps benefits have declined in the last few years as anecdotal evidence suggests, but they have clearly not fallen over the last few decades, as he contends--indeed they seem to have risen substantially over that longer time frame.

There was no data given, fool. There were weasel words like 'suggests', along with the frank admission that _no_ BLS data is available.

Pretty un-American, I'd say.

Posted by ScentOfViolets | April 20, 2008 1:21 PM

No matter how you look at the data (CPS, PSID, Social Security, IRS returns, SCF, with benefits, after-tax, etc.), increasing inequality and very small wage growth for the median (and particularly, the median male) worker is completely uncontroversial in the economics literature.

Right. But - once again these 'conservatives' are trying to turn this into a game where we've got to convince them, rather than vice versa.

Not gonna happen. Though I'll cheerfully continue to point out what they're doing, and their complete lack of effort into trying to assemble a convincing argument.

So much easier to have the other guy do the footwork, and then just say "you're not convinced" - shorthand for "if you can't make me say I'm wrong I win."

Posted by ScentOfViolets | April 20, 2008 1:26 PM

I wonder whether there's any point in commenting here any more. It would be interesting if people wanted an honest and reasonable intellectual exchange, but if what they really want is acrimony and paralogism, then I have better things to do with my time. There are some reasonable people here, like Yancey, but there are too many who seem to want a kind of internet street brawl.

For anyone who is interested in the subject matter at hand, I would point to this passage from the Minneapolis Fed paper:

"I've established that wages for the median worker went up by 20 percent between 1975 and 2005, while wages plus benefits increased by around 28 percent... My main finding is that the microeconomic and macroeconomic facts are, in fact, compatible and that the microeconomic wage series grew notably once a common price deflator was used and benefits were included."

The author's central contention is that the evidence points to a significant (if somehwhat disappointing) rise in incomes in recent decades. That is consistent with my own reading of the data.

I would also add that we ought to look at median houshold wealth, which has risen significantly as well. Now, I hasten to add that inequality certainly has increased and that there is good reason to be concerned about that. But there is a big difference between saying that the income disaprity betwen the rich and the middle class has been widening and saying that the middle class has been going backward in an absolute sense.

Posted by rwe | April 20, 2008 1:39 PM

And, once again, a person responds to SOV's request for data and citations, and those citations are dismissed because SOV disagrees with them. This is SOV's mode of operation.

SOV, the paper linked to by rwe clearly outlines the method for comparing non-wage benefits in 1975 with the non-wage benefits of today. It does so in a logical and consistent way. If you wish to contend with them, then you should cite some of your own data that contradicts them, or shows them to be in error. However, I know you won't even bother trying to do so since you never respond to such requests yourself.

Posted by Yancey Ward | April 20, 2008 1:42 PM

The middle class of 1970 is not composed of the same individuals as the middle class of 2008.

Since then, America has taken in fifty million people (immigrants plus descendants), most of whom have been dirt broke.

When my parents came here from our Secret Asian Homeland, they were poor students. We didn't have much money when I was growing up. By the time I was in high school I lived in the land of Camrys and four-bedroom houses.

Fifty million immigrants ought to artificially depress the bottom end of the scale.

On the other side, the American population has gotten older and has less children per woman. An older population raising less children ought to have gotten richer, artificially raising the middle-class income.

The question is not "what are the income quintile distributions in 1970 compared to 2008".

The question is: Given a point X on the income distribution, does an equivalent-sized family have more opportunity in 2008 or 1970?

Posted by secret asian man | April 20, 2008 1:45 PM

I wonder whether there's any point in commenting here any more. It would be interesting if people wanted an honest and reasonable intellectual exchange, but if what they really want is acrimony and paralogism, then I have better things to do with my time.

If you wanted an 'honest and reasonable intellectual exchange', you wouldn't go popping off about certain people being unpatriotic[1]. If you wanted an 'honest and reasonable intellectual exchange', you would provide some data, not someone else's massaging of data that happens to agree with you. Not 'estimates', and not 'models'.

If you have data, post it. If you don't have data, don't pass off 'analysis' for data - especially something as slipshod as this.

[1]You know, you could always just say that you never meant to imply that people like me are not patriotic, and you apologize for any misunderstanding you may have caused.

Posted by ScentOfViolets | April 20, 2008 3:18 PM

Project, much, Yancey? 'Estimates' - especially the slipshod way they are done here - are not data. Why don't you actually do some research and post what part of the compensation package insurance, pensions, vacations, etc. was the median in 1975, instead of relying on 'estimates'?

Posted by ScentOfViolets | April 20, 2008 3:22 PM

I know very little about economics, so that data massaging article (provided by other commenter) was a bit of a slog (as in, over my head).

If I understand correctly, the author suggests that average hourly earnings fell 4% and median hour wage rose 12%(75-05), at the same time that national labor income per hour rose 39% over the same period, and thus, he assumes the national labor number is correct and attempts to harmonize the data to the higher number by changing the inflation measure used, and adding benefits. Is that correct? (Though still leaving him a percentage points short in data matching).

My question to those who might be able to explain this better to me is, would an increase in workers (females) into the economy, and at lower wages, better explain the differences in the micro and macro data, or is that largely a non-factor or already accounted for?

Also, if benefits are used to explain the difference between Average Hourly Earnings/Median Hourly wage to that of national labor income per hour (with the latter assumed to be more accurate), could it be argued that many items listed as benefits don't actually make you better off now? (or again is that possibility irrelevant as well).

For example, I get healthcare in 1975 through to 2008. The costs of that healthcare rises by a huge amount. Let's say my pay wage stays at $12 per hour, but healthcare alone goes from $3 to a $9 benefit. I am getting the same amount of product (healthcare) but it's costing more. Further, and if I a marginally my normal state of healthiness most of the time, is that benefit accruing to me since, after all, it's insurance and vaporizes the minute a point in time (or that job) has passed. And that's assuming that I take advantage of benefits, which, often, people don't at the low end of the wage scale.

Thus, if you use the benefits of either healthcare or say, a 401K, they won't necessarily make you better off, and chances are, simply adding benefits to wage numbers does not factor in the masses of people who totally do NOT utilize the benefits to begin with.

Any thoughts, corrections and clarifications appreciated. Not sure how coherent this.

Posted by Finn | April 20, 2008 4:56 PM

Finn, you're understanding of the article is correct. Entry of females doesn't matter for the article's purpose, however. As for whether benefits make you better off, I don't know how germane that is to the situation. Insurance still gives you an expected value payoff regardless of what result (e.g., sickness or no-sickness) is realized.

Having looked back at the article, though, there are three fairly straightforward problems. First, he adds benefits in by assuming benefits grew at the same rate as wages for all workers. The evidence suggests this is not the case (pardon my lack of citations - you'll just have to trust me on this one); there has been growing inequality in benefits as well since 1975. Second, deflation of wages by the PCE rather than by the CPI drives much of his result, and the footnote explaining his choice seems unsatisfactory. I don't necessarily have a problem with the use of the PCE, but more detail is needed. Third, the fairly obvious answer for why production worker wages grew slower than average wages is that (as Saez and Kopczuk and many others have demonstrated) wages at the very top have driven much of the increase in the average since 1975. I think this trend is clear in other non-topcoded data sources.

Posted by cure | April 20, 2008 5:33 PM

I think the story's more than a little silly--hourly wages just aren't a very good way to organize most non-industrial workplaces, so enhanced manufacturing productivity sort of definitionally means that they'll fall.

What's "silly" is the Atlantic Monthly parading Megan AcArdle as an economics writer when she is so clcearly illiterate in the field.

She writes that hourly wages aren't a good way to organize most workplaces. First, wages are not a means of "organization" no matter how they are paid. Second, most employees, be they manufacturing or not, are paid an hourly wage.

Megan, so back to writing about the topic you're so clearly interested in: yourself.

Posted by Laughing | April 20, 2008 6:01 PM

I think the story's more than a little silly--hourly wages just aren't a very good way to organize most non-industrial workplaces, so enhanced manufacturing productivity sort of definitionally means that they'll fall. But it's not quite as silly as Belle seems to think.

Can you explain the part about organizing workplaces? I'm not sure what you mean...

Posted by Earnest Iconoclast | April 20, 2008 6:59 PM

Cure,

He used the PCE deflator because it is adjusted in an on-going fashion (you don't consume the same basket of goods in 2005 that you did in 1975), whereas the microeconomic data were deflated using CPI data that has not been adjusted as regularly. He could, I suppose, deflated the national labor income by the two CPI measurements, and this would have had the same effect of removing a portion of the difference between the increases in the macro and micro data, but I think his choice is the more valid one for the very reason he cited.

One can note, however, that he clearly outlines that increasing inequality will explain some of the differences between the three sets of data, but not all of it. I don't doubt that benefits also show some skew towards increasing inequality, but this would not mean that proportion of compensation devoted to fringe benefits have not increased at the median over the last 30 years.

Posted by Yancey Ward | April 20, 2008 8:20 PM

SOV,

You are the one claiming the paper's analysis is flawed. Come up with data showing it is wrong. He isn't just making up numbers out of thin air, or did you not read closely enough? He is making estimates based on the data that is actually available. You are free to demonstrate that his estimates are wrong, or the data he is using is wrong. Have at it. I am willing to examine your evidence. What doesn't constitute as evidence are assertions he has no data- he does.

Posted by Yancey Ward | April 20, 2008 8:25 PM

No, you bloody idiot, the article means just what it says. At that time the holy grail was a 20 dollar an hour nominal wage. The UAW was able to negotiate that. That meant that someone with only a high school education, but a good union job could make as much as a middle manager. 20 dollars an hour, in nominal terms, represented that threshold. That's what the frickin' article says.

And now it's not only not possible to get a job for the same wage in real terms for these men; it's not possible to get a job at 20 dollars an hour in 2008 dollars.

And you'll be pleased to know that you won't see me here anymore. You're supposed to be an economist by training. Getting this wrong is inexcusable.

(And, of course, it comes on the same day that we discover the market demands that television journalists provide extensive, detailed false government propaganda.)

Posted by JayAckroyd | April 20, 2008 8:45 PM

Let's say my pay wage stays at $12 per hour, but healthcare alone goes from $3 to a $9 benefit. I am getting the same amount of product (healthcare) but it's costing more.

The fallacy here is assuming you're getting the same product in 1975 and 2008. If you restricted your current care to that available in 1975, the cost would not change much- maybe higher because doctors wages have gone up with inflation, but probably lower because so much has gone off patent. And even limiting yourself to 1975 care, the care would still be better, because we have more research on what works and what doesn't.

Posted by Mandatory Vacation | April 20, 2008 11:00 PM

As far as I know, the UAW first hit $20 an hour in the 1980s, by which point inflation had severely eroded the value of that $20. And AFAIK, the average salary of a GM line worker today is well above $20 an hour, and will be for the foreseeable future.

The average hourly wage of a worker in 1979 was a little over $6. The average hourly wage of a transportation equipment manufacturing worker was $13 in 1990, the first year for which data is available. I don't know where you're getting your information, and I am open to having made a mistake here, but I find it hard to believe that the UAW was getting the modern equivalent of upwards of $60 an hour as an average. I've known plenty of union workers in closed industries who make over $100K, but that usually involves some overtime.

Even if they had been getting that, however, an average hourly of $6 an hour would make them an extreme outlier, not the kind of widespread phenomenon that the Times paints.

Posted by Megan McArdle | April 20, 2008 11:06 PM

They are inflation-adjusting the $20 wage--back to the 1970s. A $20 an hour wage in 1973 would be about a $100 an hour wage in today's dollars. Even if you think that the bygone industrial days were a halcyon era of plenty, this should flunk an elementary gut check. $100 an hour translates into an annual salary of about $200K a year in today's money. Yet per-capita GDP has more than doubled since 1970.

No, "they" are not. Or rather Louis Uchitelle is not. He's talking about nominal wages for precisely the union workers that hit 20 nominal dollars in 1980. He's not talking about general manufacturing wages throughout the US. He's talking about nominal wages in the Rust Belt for formerly union workers. The article is very, very clear about that.

Whatever Senator Barack Obama meant by his less than artful remarks about small-town Pennsylvanians “bitter” over lost jobs, he certainly turned a lot of attention last week to the decline of the American worker, bitter or not.

The talk most often has been of shuttered factories, layoffs, outsourcing and other effects of globalization, especially in a state like Pennsylvania, which has lost tens of thousands of industrial jobs. But there is another way to look at blue-collar workers or their counterparts in the service sector.

Leaving aside for a moment those who have lost their jobs, what of those who still have them? Once upon a time, a large number earned at least $20 an hour, or its inflation-adjusted equivalent, and now so many of them don’t.

The $20 hourly wage, introduced on a huge scale in the middle of the last century, allowed masses of Americans with no more than a high school education to rise to the middle class. It was a marker, of sorts. And it is on its way to extinction.

Americans greeted the loss with anger and protest when it first began to happen in big numbers in the late 1970s, particularly in the steel industry in Western Pennsylvania. But as layoffs persisted, in Pennsylvania and across the country, through the ’80s and ’90s and right up to today, the protests subsided and acquiescence set in.

Hourly workers had come a long way from the days when employers and unions negotiated a way for them to earn the prizes of the middle class — houses, cars, college educations for their children, comfortable retirements. Even now a residual of that golden age remains, notably in the auto industry. But here, too, wages are falling below the $20-an-hour threshold — $41,600 annually — that many experts consider the minimum income necessary to put a family of four into the middle class.

Now, "middle of the last century" is an exaggeration. But he is clearly, unambiguously writing about the fact that you could earn 20 dollars an hour at a union job in steel mill or a car plant 30 years ago, and now you can't. You could earn that with only a high school education. Now you can't earn that in nominal terms in the region.

Later. Much later.

Posted by jayackroyd | April 21, 2008 12:01 AM

They are inflation-adjusting the $20 wage--back to the 1970s. A $20 an hour wage in 1973 would be about a $100 an hour wage in today's dollars. Even if you think that the bygone industrial days were a halcyon era of plenty, this should flunk an elementary gut check. $100 an hour translates into an annual salary of about $200K a year in today's money. Yet per-capita GDP has more than doubled since 1970.

No, "they" are not. Or rather Louis Uchitelle is not. He's talking about nominal wages for precisely the union workers that hit 20 nominal dollars in 1980. He's not talking about general manufacturing wages throughout the US. He's talking about nominal wages in the Rust Belt for formerly union workers. The article is very, very clear about that.

Whatever Senator Barack Obama meant by his less than artful remarks about small-town Pennsylvanians “bitter” over lost jobs, he certainly turned a lot of attention last week to the decline of the American worker, bitter or not. The talk most often has been of shuttered factories, layoffs, outsourcing and other effects of globalization, especially in a state like Pennsylvania, which has lost tens of thousands of industrial jobs. But there is another way to look at blue-collar workers or their counterparts in the service sector. Leaving aside for a moment those who have lost their jobs, what of those who still have them? Once upon a time, a large number earned at least $20 an hour, or its inflation-adjusted equivalent, and now so many of them don’t. The $20 hourly wage, introduced on a huge scale in the middle of the last century, allowed masses of Americans with no more than a high school education to rise to the middle class. It was a marker, of sorts. And it is on its way to extinction. Americans greeted the loss with anger and protest when it first began to happen in big numbers in the late 1970s, particularly in the steel industry in Western Pennsylvania. But as layoffs persisted, in Pennsylvania and across the country, through the ’80s and ’90s and right up to today, the protests subsided and acquiescence set in. Hourly workers had come a long way from the days when employers and unions negotiated a way for them to earn the prizes of the middle class — houses, cars, college educations for their children, comfortable retirements. Even now a residual of that golden age remains, notably in the auto industry. But here, too, wages are falling below the $20-an-hour threshold — $41,600 annually — that many experts consider the minimum income necessary to put a family of four into the middle class.

Now, "middle of the last century" is an exaggeration. But he is clearly, unambiguously writing about the fact that you could earn 20 dollars an hour at a union job in steel mill or a car plant 30 years ago, and now you can't. You could earn that with only a high school education. Now you can't earn that in nominal terms in the region.

Later. Much later.

reformatted.

Posted by Jay Ackroyd | April 21, 2008 12:06 AM

I think the author of the article is a bit imprecise in his language in terms of what income in what time period. He mentions late seventies, the eighties, the middle of the last century.

This phrase obscures his point, I think:

The $20 hourly wage, introduced on a huge scale in the middle of the last century, allowed masses of Americans with no more than a high school education to rise to the middle class.

Posted by Finn | April 21, 2008 4:15 AM

The debate above shows just how poorly written the Times' article is. The author writes:

Once upon a time, a large number earned at least $20 an hour, or its inflation-adjusted equivalent, and now so many of them don’t. The $20 hourly wage, introduced on a huge scale in the middle of the last century, allowed masses of Americans with no more than a high school education to rise to the middle class. It was a marker, of sorts. And it is on its way to extinction.

So, he claims, $20 per hour was a typical or at least common wage for a blue collar worker in about 1950. That would equate to $174 today. So a worker working 8 hours a day, 5 days a week and 48 weeks a year would have had an annual salary of $334,080 in todays dollars. Nice work if you can get it.

But GDP per capita was only about $9600 in 1950. So a typical blue collar worker was able to earn an annual salary more than 30 x the average income? That doesn't pass the smell test. it can't be right. The Times tells us:

Louis Uchitelle has covered economics for The New York Times since 1987, focusing on business issues, monetary and fiscal policy and labor trends, particularly the unwinding of job security and the spread of layoffs.

It's hard to believe this guy has been covering economics for two decades and yet can have botched a simple economic analysis so completely.

Posted by rwe | April 21, 2008 7:53 AM

Just to show more clearly how wrong Uchitelle is: he says a typical worker was able to earn $20 per hour in 1950. $20 per hour gives an annual income of about $38,400. This woud have been 15 times the median income for a male worker in 1950, according to Census data. So he's wrong.

And Belle Waring of Crooked Timber is also wrong. She infers from the article that since 1970 the typical worker has seen his wages fall by 2/3 in real terms. But a quick glance at the Census data shows that can't be true. The median income for men was higher in 2006 than in 1970 in real terms (though admittedly not by much). Unless men are working three times as many hours as they were in 1970, their wages couldn't have fallen by 2/3. And of course women have seen their incomes rise sharply.

I don't really blame Waring, though, The Times article is one of the worst and most misleading I've read. It might be time for Louis Uchitelle to retire.

Posted by rwe | April 21, 2008 8:54 AM

Jack, I'm sorry, but as far as I can tell you're just wrong. *Whatever* the wages at the UAW (which, incidentally, has zero members in the region we are discusssing, AFAICT), the article is clearly speaking of a widespread phenomenon, which the $20 an hour wage in 1979, much less "the 1970s" was not. The minimum wage of 1979 was $2.90. It is mathematically impossible to get to an average of $6.00 an hour if, as Uichetelle states, "23 percent in 1979" made a nominal $20; even if every other hourly worker in the country made minimum wage, the numbers would still come in too high.

Moreover, it was not *possible* that more than a handful of people enjoyed such a wage during the period that could reasonably be characterized as "midcentury"--i.e. before the Great Inflation. Uichetelle says "The $20 hourly wage, introduced on a huge scale in the middle of the last century" --a definite statement which no reputable paper would pass if it referred to a $20 nominal wage, as you are suggesting.

Uichetelle's article could be more clearly written. But the hard statements he makes about the wages simply cannot be squared with your interpretation.

Posted by Megan McArdle | April 21, 2008 9:06 AM

'The $20 hourly wage, introduced on a huge scale in the middle of the last century' --a definite statement which no reputable paper would pass if it referred to a $20 nominal wage, as you are suggesting.-MM

It's hard to sqaure Megan's interpretation with the language in the article. Uchitelle says "Once upon a time, a large number earned at least $20 an hour, or its inflation-adjusted equivalent, and now so many of them don’t." But that implies that $20 was not adjusted for inflation, meaning that it is nominal terms, not real terms.

Moreover, if he's talking about the real wage, what sense does it make to say:

"The $20 hourly wage, introduced on a huge scale in the middle of the last century, allowed masses of Americans with no more than a high school education to rise to the middle class. It was a marker, of sorts. And it is on its way to extinction... That basic wage blossomed first in the auto industry in 1948 and served, in effect, as a banner in the ideological struggle with the Soviet Union."

This passage is unintelligible if he's not referring to a $20 nominal wage.

I agree with Megan that what Uchitelle seems to be saying is crazy. $20 an hour was nowhere near typical in 1950 or in 1970.

But even if Megan's interpretaion is correct, the article is still one of the worst economics articles I've read. Not only is it poorly written, but it is also quite dishonest. I throws around a lot of statsitcs and anecdotes to mislead the reader into believing that there has been a serious erosion of wages for the typical worker since 1948 or at least since the 1970's. And the data just don't bear that out.

Posted by rwe | April 21, 2008 9:41 AM

I agree with you that it is unnecessarily vaguely written, but the other interpretation--Jayack's interpretation--is simply nuts. I can't believe that statements like a $20 an hour nominal wage became widespread in the middle of the century would have made it past even a very green fact checker. Between inept and obviously insane, I vote for inept.

Posted by Megan McArdle | April 21, 2008 9:51 AM

Megan, I posted a comment last night that I think got caught in the spam filter due to links, can you fish it out and approve?

Posted by Skullberg | April 21, 2008 11:18 AM

FWIW, I was a Tool and Die Maker (journeyman)at a UAW shop in 1972. That job classification (pay grade) was the 2nd highest hourly in the plant next to the "Master" T&D Makers. Hourly rate was about $8.32. We had just gone thru a long strike with the resulting new 3-year contract granting increases to 1975 that would result in that rate increasing to $9.31/hr and the Masters' would have gone to about $10.35+, IIRC. Today, that rate is around $27/hr with some plants paying up to $30 due to shortages of qualified personnel.
As an aside, I think the assembly line workers were at about $6/hr in '72. Benefits were pretty good as was the retirement plan, but I have no idea what those might have been worth then.

Posted by MikeinAppalachia | April 21, 2008 12:46 PM

Seriously... we were having an interesting back and forth about a real issue, then people start the name calling.

You can disagree, even strongly disagree, without breaking out the vitriol.

Posted by Toxic | April 21, 2008 5:03 PM

Brooksfoe - Benefit dollars have increased in real terms, not decreased.

Posted by Tim Fowler | April 21, 2008 9:04 PM

I agree, Toxic. Megan, could you do something about rwe? Give him a strike or something?

Posted by ScentOfViolets | April 22, 2008 12:12 AM