Is the American middle class really shrinking? If so, how far and how fast?

Perhaps this belongs in GQ, but arguments over how to define “middle class” are bound to land it in GD eventually, so we might as well start it here. Ever since the early '80s, commentators have been expressing fears about the decline of the American middle class. For instance, this article from the Atlantic Monthly of July 1983:

So . . . has it actually happened? A brief search turned up some more recent articles but they seem to be rather tendentious, e.g., this one from the Federation for American Immigration reform cough[racists]cough:

What are the real stats? First of all, how do we define the “middle class” – as a simple income range, or are there also cultural elements and other sociological factors? Second, what about this class has changed in the past 30 years? If it has shrunk, by how much? And how many of its members have fallen into the lower classes and how many have risen to the upper classes?

And . . . what, if anything, can be done about it?

I remember this discussion from the 80s. IIRC, the middle class is, or was, shrinking. This is due to a number of factors. Most of it is due to the fact that the middle class is moving into the upper class, largely due to the aging of the Baby Boomers. As they move into their peak earning years (45-60), they move up faster than they are replaced by those moving up from the lower classes.

You will note that your Atlantic article was published in 1983, just as the Reagan economic boom was getting started. You will remember that both the number of people living in poverty, and the percent of those living in poverty, decreased under Reagan. The poverty rate in 1981 was 14.0%, and the number of people living in poverty was 31.8 million. In 1989, the poverty rate was 12.8%, and the number of people living in poverty was 31.5 million. Cite. This despite the fact that the US population grew by almost twenty million during the same period. Cite.

What should be done about it? Essentially nothing, IMO. This strikes me as similar to a criticism made of quota-based affirmative action - that it “seeks to make statistical fluctuation a federal crime”.


Among other things, we should get illegal immigration under control cough[not racist]cough.

So you’re assuming class status is defined purely by income?

This is a rather convenient use of statistics which takes advantage of economic boom-bust cycles. 1989 corresponds to a local minimum for the number in poverty. 1981 corresponds to a time when poverty was rising rapidly (although it peaked in 1982 and 1983). Here is the full graph which gives a rather different story than just pulling statistics from specific years.

And, by the way, if you want to play games with specific years, look at the Clinton record from 1993 to 2001 where the number in poverty dropped from ~39 million to ~32.9 million. But, again, I wouldn’t credit this that much to Clinton as most of it had to do with going from the peak of the recession to just after the peak of the boom. What may be a little more significant is that the poverty rate in 2000 had dropped to a level not seen since the 70s (which was the lowest level since the graph starts in 1959).

What has, of course, happened in the last 20 years or so is that the income gap between the rich and everyone else has widened significantly. As an example, the after-tax income of the top 1% increased 201% in real terms between 1979 and 2000. For the middle fifth of households, the increase was only 15%, with the lion’s share of that (12 of the 15%) coming in the 1990s. In fact, the bottom two quintiles saw their real incomes drop between 1979 and 1989. Here is a report from the Center for Budget and Policy Priorities (a liberal think-tank) with references back to the CBO report that it is based upon.

So, I don’t think you would necessarily call this a shrinking of the middle class, but rather a significant widening of the gulf between the middle class and the upper class.

In this context, yes.

If you are arguing that people don’t change their mindset and mores away from the middle class when they move up into the upper classes, that is true, but the OP’s cites seemed to be basing a definition of "middle class’ on income.

FWIW, my definition of “middle class morality” would be a belief in the following:
[ul][li]Delayed gratification[/li][li]Investment over consumption[/li][li]A belief in the value of education[/li][li]Hard work is more important than luck[/ul]There would be a big component of “family values” in there as well, but that is probably so contentious as to be a distraction. [/li]
I would further contend that a commitment to the above set of principles generally correlates with a rise into the middle class and beyond. You may be interested in a book called “The Millionaire Next Door”, by Thomas Stanley and William Danko. The authors detail the common elements of how their subjects moved from the lower or middle classes into the millionaire range. They didn’t do it by winning the lottery. :wink:

Do you have a different definition?



FWIW, my definition of “middle class morality” would be a belief in the following:
[ul][li]Delayed gratification[/li][li]Investment over consumption[/li][li]A belief in the value of education[/li][li]Hard work is more important than luck[/ul][/li][/QUOTE]

Oh come on. Plenty of the poor have these same values, but don’t have the chance to act on them because of a lack of resources. Delayed gratification is great, but you can only delay dinner so many nights. Investment doesn’t work if you have nothing to invest. A decent education in crappy schools is an uphill battle and luck is pretty damn important if breaking an arm or getting sick can bankrupt you because you can’t afford health insurance.

Middle classdom isn’t a state of moral superiourity.

Well, I admit that income level is of more practical importance than anything else, and when people speak of the “shrinking middle class” they’re generally talking about a shrinking number of people within a certain income range; and if we are to feel any anxiety about the “shrinking middle class,” it’s income, not culture, that we should be worrying about. But many commentators have suggested class identity is more complicated than that. (Others insist there’s no such thing at all in America as social classes, in the Old World sense, but I’ve never bought that line.) The following is from Class: A Guide to the American Status System, by Paul Fussell (New York: Summit Books, 1983):

And the following, a more serious analysis which is directly concerned with the upper class or “White Overclass,” not the middle class, comes from The Next American Nation by Michael Lind (New York: Free Press Paperbacks, 1996):

Not so much “convenient” as “demonstrating exactly what I said it did”. I said the number of people living in poverty went down under Reagan, which it did. I said the poverty rate went down under Reagan, and it did. 1981 was the first year Reagan was President; 1989 was the last.

Actually, no, it doesn’t give a different story at all. The poverty rate and number of people living in poverty were both trending downwards until 1976-1977, when Ford left office and Carter came in*. Then both increased sharply, and continued up until Reagan, when both resumed their decline, interrupted by the recession of 1990 that got Clinton elected. Then they trended downwards thru out the 90s, which, as you and I agree, had little to do with Clinton’s policies.


*Remember the Misery Index? Carter said it was 15 during the Ford administration, and “no President deserved to be re-elected with an index that high”. During the debates of 1980, Reagan pointed out that the Misery Index was hovering around 20.

Clinton was re-elected with a Misery Index of 8.22. Last February, under Bush, it was 7.29. Cite.

BTW, I started a GD thread last year: “How many social classes are there in the United States?” – No consensus emerged, and some posters stubbornly insisted that there are no social classes at all in the United States, only income levels.

Yes, but as I pointed out, it had everything to do with economic cycles. And, if you applied the same reasoning to the Clinton years (1983 to 2001), you would find a much more dramatic decrease in both the poverty rate and the number in poverty.

What you have demonstrated is that you can quote a few statistics but can’t carry out any sort of intelligent analysis of them.

God, your analysis here is so completely full of it. Just so everyone can see the graph you are making up novel interpretations about, here it is again. I, for one, don’t see any trend upwards starting in 1976-1977. It started a few years later when the recession began. And, the increase in poverty continued under Reagan, before reversing but dropping only to the extent that the number in poverty was only marginally less in 1989 (which was a local minimum) than 1981. The difference in poverty rate was a little more pronounced because of the growth in population but still nothing like the drop in poverty rate that was seen under Clinton.

You then conveniently quote my statement about Clinton , which is motivated by trying to present an honest accounting rather than bullshit partisan spin. You clearly operate under no such constraints. (No matter how low I set the bar in discussions with you, you seem to crawl under it.)

The misery index is an example of what someone here at work calls “mononumerosis”. If you believe that a 3% increase in inflation (particularly when wages are inflating too) is as much of a problem as a 3% increase in unemployment, then I suppose that misery index is for you.

Honestly, Shodan, with your selective cherrypicking of data and statistics, you are well-qualified to serve with distinction in this Administration. Alternately, I am sure the editorial page of the Wall Street Journal would love to have you join their staff.

And, finally, just to add a little more food for thought in regards to that CBPP report on income gains, one of the most interesting features is that the average real income gain over the period 1979 to 2000 was 40% but the median gain (or at least the average gain of the middle quintile…which ought to be close to the median gain) was only 15%. This gives you some idea of how much more dramatic income gains would have been across the spectrum if incomes had gone up by the same amount on average but without such a rapid increase in inequality…in which case all incomes would have gone up by 40%. One could debate how easily it would be possible to realize such a situation. (Of course, somemight claim that one has to formulate policies such that there is a dramatic increase at the top in order to get any sort of increase across the board.) But, I think it is a really interesting question to ponder whether we want to continue on this trend of having huge gains for the upper classes, small ones for the middle classes, and even less for the lower classes. And, if not, what might we do to change it?

Horatio Alger also published a lot of work along these lines.

In a word, horseshit. You asserted it. You didn’t prove it.

And since the article cited in the OP referred to the Reagan years, that would have been off-topic, wouldn’t it? I asserted known and demonstrable facts. Your desire to change the subject to what you admit is irrelevant (Clinton’s effect on the economy) is beside the point. I asserted, and demonstrated, that the number of people in poverty, and the poverty rate, both declined under Reagan. Therefore, the Atlantic article is false in its assertions. Which was the subject of the OP.

And if you are asserting that 1983 was the start of the Clinton years, I suggest you check your history, Mr. Analysis Expert.

Look at the chart again. Note that the numbers living in poverty and the poverty rate begin to increase before the two closely following recessions. Notice how they continue to increase until the end of the second recession. Notice how they then begin their decline under Reagan. Exactly as I stated they did.

If you characterize the idea that the economy did better under Reagan than under Carter as a “novel interpretation”, you are the only one in the Western hemisphere to do so.

In other words, what I stated was the exact and literal truth.

Not for me, for Carter. Or for that clown Kerry, who manufactured a new version that implied that the economy did better under Carter than Reagan. Just what the Democrats and the country need - another Democratic dolt who has someone else balance his checkbook for him.

If you are saying that my analysis is closer to that of the Wall Street Journal than Kerry, thank you. I would hope it is.


You can’t prove anything in economics. But, look at the freakin’ graph! It clearly shows a cyclical nature to the poverty figures, correlated to periods of recession and economic expansion, especially over the last 25 years.

No, first off the OP did not only discuss the Reagan years. He said, “Ever since the early '80s, commentators have been expressing fears about the decline of the American middle class.” Second, citing one statistic about something does not prove anything as I have explained. Economics is more complicated than that. One can always cherry-pick statistics. As I noted, using your same analysis shows that Clinton was much better at reducing poverty than Reagan was! However, I wasn’t ready to conclude that this was conclusive evidence of any sort because I don’t engage in such bullshit.

Wow, congratulations. You caught me in a typo. Obviously, I meant to type 1993. As is clear from the graph, my analysis with the correct 1993 and 2001 dates.

What you actually said is “The poverty rate and number of people living in poverty were both trending downwards until 1976-1977, when Ford left office and Carter came in*. Then both increased sharply, and continued up until Reagan, when both resumed their decline, interrupted by the recession of 1990 that got Clinton elected.” This statement is incorrect in a number of respects:

(1) The poverty rate was not trending downward in the period before 1976-1977. It was in fact pretty flat through the 70s with a few small ups and downs.

(2) The poverty rate did not increase sharply immediately after that. In fact, it did not increase at all until an increase between 1978 and 1979, and that was not a sharp increase…The increase got what most people would judge as sharp between 1979 and 1980. While it is true that the increases preceded the recession, this is not out-of-line with other recessions such as the 1990 one or the 2001 one. (Also note that the point are plotted only every year so there is some discrete jumpiness in the graph.)

(3) The poverty rate did not continue up until Reagan and then resume a decline. It continued up under Reagan, rising sharply from 1980 to 1981 and 1981 to 1982, and rising slightly from 1982 to 1983 before starting to decline.

(4) The “interruption” you speak of in terms of the recession of 1990 actually resulted in a rise in poverty for a four year period from 189-1993. [This was your less gross mischaracterization.]

I never claimed what you said was not literally true but rather that it didn’t really tell us anything useful. What I said exactly was: “This is a rather convenient use of statistics which takes advantage of economic boom-bust cycles. 1989 corresponds to a local minimum for the number in poverty. 1981 corresponds to a time when poverty was rising rapidly (although it peaked in 1982 and 1983). Here is the full graph which gives a rather different story than just pulling statistics from specific years.”

Well, this is the one point we might agree on then…namely that both the original misery index and Kerry’s version are basically silly. One has to look at the economy more holistically. Under any reasonable comparion, however, the economy is worse off than we were 4 years ago…and while this is not entirely Bush’ fault, at some point he does have to get blame, e.g., for doing so little to spur job growth and squandering a hell of a lot of money doing it!

In fact, even in 1989 when the poverty rate reached a local minimum, it was higher than it had been during much of the Carter Administration (specifically, higher than it had been in 1977, 1978, or 1979).

All in all, Shodan, your description of the graph is so deceptive that it bears little resemblance to the reality which is why I keep linking to it so that everyone can see.

Obviously, “189” should be “1989” and “less” should be “least”.

Everybody’s arguing over what the poverty rate was under what administration and which president deserves the credit or the blame for it – but what we need to be concerned with in this thread is the long-term net effects. Is the American middle class shrinking or not? Has it shrunk since the 1970s or not? If so, why? How many people have risen from the middle class into wealth, how many have fallen into poverty?

Oh, I HAVE to take exception to this. There is no way you would find any reasonable economist who would state that the economy was better four years ago than it is now.

Four years ago, the economy was sliding into recession. Today, GDP growth is a very robust 4.2%.

In fact, if you go down the list of economic indicators, almost all of them are higher now than they were four years ago - with one exception: unemployment. It was near historical lows four years ago, and it’s about a point and a half higher now.

However, any honest evaluation of the state of the economy would not be a static snapshot of indicators, but would include clear trends. And all the trend lines were pointing down in 2000. The tech bubble had just busted. Unemployment was still low because unemployment is a trailing indicator.

But there’s a reason why the Democrats keep harping on the unemployment rate, even though it’s historically very good already and getting better. But it’s all they’ve got. Because if you look at other indicators like manufacturing investment, productivity growth, and GDP growth, they are not only good, they are excellent. In some cases near 20 year highs.

The economy is a losing issue for Kerry, unless the Bush team is as totally inept at showing how good the economy really is.

The topic is, not how healthy is the economy right now, or in the recent past, but how does the American socioeconomic pyramid, right now, compare to how it was a few years ago, and to how it will be a few years from now?

I was hoping to include a link to Stephen J. Rose’s American Profile Poster (W.W. Norton & Co., 2000), but I can’t find an online version. It is a graph using symbolic round-headed figures of various colors to show distributions of wealth, income and occupational categories in our society, and across racial groups and genders. There should be an online version.

Sorry to have gone off with Shodan so far in his hijack, but I wasn’t about to let him get away with his distortions. How about my link to that CBPP study though? It doesn’t speak exactly to your questions but it does show how the chasm between the most of the population and the rich has considerably widened in the last 25 years.

You have a strange way of looking at this. By your standards, the economy would be considered better as it was starting to climb out of the Great Depression than it would be at the peak of an economic cycle. Sure, trend lines are going up now as we finally climb out of a recession that supposedly ended a long time ago but led to a recovery that was so anemic that it is only now beginning to create some jobs…And, has still left us behind where we were on jobs a few years ago.

And, for whatever contribution that the stimulus has made to the recovery (which would have largely occurred on its own anyway), we only have to pay about a $3 trillion mortgage! With bargains like this, our nation will be broke in no time.