(Cross-posted at ProgressiveFix and Frum Forum)

Everyone’s approvingly linking to this Edward Luce piece on “the crisis of middle-class America”.I want to set myself on fire.

Seriously, it’s discouraging to see so many people who should know better (because they’ve argued these points with me before) promoting this article.I can’t think of another piece in the doomsday genre—and there are many—that gets it so consistently wrong.  I'll stipulate that none of the criticisms below are intended to minimize the struggles that many people are facing.  But it's important to get this stuff right.  Let me dive in, with Luce’s words in italics and my responses following:

Yet somehow things don’t feel so good any more. Last year the bank tried to repossess the Freemans’ home even though they were only three months in arrears. 

The share of mortgages either in foreclosure or 3 or more months delinquent is
11.4 percent, which, because 30 percent of homeowners have paid off their mortgage, translates into 8 percent of homes.So the Freemans’ situation is typical of about one in twelve homeowners, or just over 5 percent of households (since one-third rent).*Their son, Andy, was recently knocked off his mother’s health insurance and only painfully reinstated for a large fee. 

Luce is arguing that there’s a new crisis facing the current generation.About 30 percent of those age 18 to 24 were uninsured in 2008 when the National Health Interview Survey contacted them.I don’t have trends for that age group, but the share of Americans under age 65 without health insurance coverage was
14.7 percent in 2008, up from….14.5 percent in 1984.

And, much like the boarded-up houses that signal America’s epidemic of foreclosures, the drug dealings and shootings that were once remote from their neighbourhood are edging ever closer, a block at a time. 

Well, the violent crime rate in 2008 was
19.3 per 1,000 people age 12 and up, down from 27.4 in 2000 and 45.2 in 1985. 

Once upon a time this was called the American Dream. Nowadays it might be called America’s Fitful Reverie. Indeed, Mark spends large monthly sums renting a machine to treat his sleep apnea, which gives him insomnia. “If we lost our jobs, we would have about three weeks of savings to draw on before we hit the bone,” says Mark, who is sitting on his patio keeping an eye on the street and swigging from a bottle of Miller Lite. “We work day and night and try to save for our retirement. But we are never more than a pay check or two from the streets.”

The key question is, again, Is this worse than in the past?The risk of a large drop in household income has risen modestly, but people experiencing a drop end up much better off than in the past.For example, the risk of a 25 percent drop in income over 2 years has risen from 7 percent among married couples in the late 1960s to 14 percent in the mid-2000s (based on my computations from Panel Study of Income Dynamics data).But if you look at the average income of married-couple families after their 25 percent drop, it rose from $40,000 to $63,000 (in constant 2009 dollars).

Solid Democratic voters, the Freemans are evidently phlegmatic in their outlook. The visitor’s gaze is drawn to their fridge door, which is festooned with humorous magnets. One says: “I am sorry I missed Church, I was busy practicing witchcraft and becoming a lesbian.” Another says: “I would tell you to go to Hell but I work there and I don’t want to see you every day.” A third, “Jesus loves you but I think you’re an asshole.” Mark chuckles: “Laughter is the best medicine.”

Hmmm….just a typical American household…..

The slow economic strangulation of the Freemans and millions of other middle-class Americans started long before the Great Recession, which merely exacerbated the “personal recession” that ordinary Americans had been suffering for years. Dubbed “median wage stagnation” by economists, the annual incomes of the bottom 90 per cent of US families have been essentially flat since 1973 – having risen by only 10 per cent in real terms over the past 37 years. That means most Americans have been treading water for more than a generation. Over the same period the incomes of the top 1 per cent have tripled. In 1973, chief executives were on average paid 26 times the median income. Now the multiple is above 300. 

Adjusting for household size and using the PCE deflator to adjust for inflation, median household income in the Current Population Survey rose from $29,800 in 1973 to $40,500 in 2008 (in 2009 dollars, again based on my compuatations).Factoring in employer and government noncash benefits would show even more impressive growth.
In the last expansion, which started in January 2002 and ended in December 2007, the median US household income dropped by $2,000 – the first ever instance where most Americans were worse off at the end of a cycle than at the start. 

This is entirely a function of changes in the population composition (more Latinos) and in the share of employee compensation going to health insurance and retirement plans.

Worse is that the long era of stagnating incomes has been accompanied by something profoundly un-American: declining income mobility. 

evidence is ambiguous, but the best studies imply that intergenerational economic mobility hasn’t changed that much in the past few decades.Intra-generational earnings mobility has increased since the 1950s, though it has declined among men.

Alexis de Tocqueville, the great French chronicler of early America, was once misquoted as having said: “America is the best country in the world to be poor.” That is no longer the case. Nowadays in America, you have a smaller chance of swapping your lower income bracket for a higher one than in almost any other developed economy – even Britain on some measures. To invert the classic Horatio Alger stories, in today’s America if you are born in rags, you are likelier to stay in rags than in almost any corner of old Europe.

Tim Smeeding’s research based on the Luxembourg Income Study shows that in general Americans have higher incomes than their European counterparts as long as they are in the top 80 to 90 percent of the income distribution.Below that, incomes are more comparable across countries, and the living standards of Americans look less impressive.The US has comparable intergenerational earnings mobility to Europe, according to
Markus Jantti’s research, except among men (but not women) who start out at the bottom.In terms of occupational mobility, David Grusky’s research shows we're as good or better as anywhere else, but this doesn't translate into earnings mobility because we let people get rich or poor to a greater extent than other countries do.Jantti and Anders Bjorklund have estimated that Sweden would have the same mobility as the U.S. if the return to skill was as high there as it is here.Finally, employer benefits further complicate how "bad" we look.

Combine those two deep-seated trends with a third – steeply rising inequality – and you get the slow-burning crisis of American capitalism. It is one thing to suffer grinding income stagnation. It is another to realise that you have a diminishing likelihood of escaping it – particularly when the fortunate few living across the proverbial tracks seem more pampered each time you catch a glimpse. “Who killed the ­American Dream?” say the banners at leftwing protest marches. “Take America back,” shout the rightwing Tea Party demonstrators. 

The rise in income inequality is mostly about the
top 5% of the top 1% pulling away from everyone else, and existing estimates overstate inequality and its growth by ignoring employer and government noncash benefits and possibly by ignoring different rates of inflation in different parts of the income distribution.

Unsurprisingly, a growing majority of Americans have been telling pollsters that they expect their children to be worse off than they are. 

Totally wrong.The key here is to only look at polling questions that ask people about their own kids, not kids in general.Here are the relevant survey results I could find:

General Social Survey (1994)—45% said their children’s standard of living will be better (vs. 20% worse)
General Social Survey (1996)—47%
General Social Survey (1998)—55%
General Social Survey (2000)—59%
General Social Survey (2002)—61% said their children’s standard of living will be better (vs. 10% worse)
General Social Survey (2004)—53%
General Social Survey (2006)—57%
General Social Survey (2008)—53%
Economic Mobility Project (2009)—62% said their children’s standard of living will be better (vs. 10% worse)(unlike GSS and PRC, asked only of those with kids under 18)
Pew Research Center (2010)—45% said their children’s standard of living will be better (vs. 26% worse)
BusinessWeek (1989)—59% said their children will have a better life than they had (and 25% said about as good)
BusinessWeek (1992)—34% said their children will have a better life than they had (and 33% said about as good)
BusinessWeek (1995)—46% said their children will have a better life than they have had (and 27% said about as good)
BusinessWeek (1996)—50% expected their children would have a better life than they have had (and 26% said about as good)
Harris Poll (2002)—41% expected children will have a better life than they have had (and 29% said about as good)
Harris Poll (1997)—48% felt good about their children’s future
Harris Poll (1998)—65% felt good about their children’s future (17% N.A.)
Harris Poll (1999)—60% felt good about their children’s future (15% N.A.)
Harris Poll (2000)—63% felt good about their children’s future (17% N.A.)
Harris Poll (2001)—56% felt good about their children’s future
Harris Poll (2002)—59% felt good about their children’s future
Harris Poll (2003)—59% felt good about their children’s future
Harris Poll (2004)—63% felt good about their children’s future
Pew Research Center (1997)—51% said their children will be better off than them when they grow up
Pew Research Center (1999)—67% said their children will be better off than them when they grow up
Bendixen & Schroth (1989)—68% said their children will be better off than they are
Princeton Religion Research Center (1997)—62% of men said their sons will have a better chance of succeeding than they did; 85% of women said their daughters will have a better chance
Angus Reid Group (1998)—78% said children will be better off than them
Washington Post/Kaiser Family Foundation/Harvard (2000)—46% said they were confident that life for their children will be better than it has been for them
Economic Mobility Project (2009)—43% said it would be easier for their children to move up the income ladder
Economic Mobility Project (2009)—45% said it would be easier for their children to attain the American Dream

Also, polls consistently show that Americans say they have higher living standards than their parents.

And although the golden years were driven by the rise of mass higher education, you did not need to have graduated from high school to make ends meet. Like her husband, Connie Freeman was raised in a “working-class” home in the Iron Range of northern Minnesota near the Canadian border. Her father, who left school aged 14 following the Great Depression of the 1930s, worked in the iron mines all his life. Towards the end of his working life he was earning $15 an hour – more than $40 in today’s prices. 

Thirty years later, Connie, who is far better qualified than her father, having graduated from high school and done one year of further education, makes $17 an hour. 

It’s not valid to compare her pay mid-career to her father’s at the end of his career—and also, how much work experience does she have relative to him?Did she take time off to raise kids?

The pace of life has also changed: “We used to sit around the dinner table every evening when I was growing up,” says Connie, who speaks with prolonged vowels of the Midwest. “Nowadays that’s sooooo rare.” 

Time-use surveys show that while parents spend more time working (because of mothers) than in the past, they do not spend less time with children.They spend less time doing things by themselves.

Then there are those, such as Paul Krugman, The New York Times columnist and Nobel prize winner, who blame it on politics, notably the conservative backlash which began when Ronald Reagan came to power in 1980, and which sped up the decline of unions and reversed the most progressive features of the US tax system. 

Fewer than a tenth of American private sector workers now belong to a union. People in Europe and Canada are subjected to the same forces of globalisation and technology. But they belong to unions in larger numbers and their healthcare is publicly funded. 

Though unionization has declined markedly in most of these countries, and their healthcare policies are increasingly becoming too costly.Also, most of the decline in unionization in the U.S. occurred
before Reagan took office.

More than half of household bankruptcies in the US are caused by a serious illness or accident. 

This is
bad Elizabeth Warren research—she counts a bankruptcy as being “caused” by illness or accident if one was reported, but the household could have been in serious debt before these occurred.At any rate, bankruptcies are exceedingly rare (under 1 percent of households—see Figure 13).

Pride of place in Shareen Miller’s home goes to a grainy photograph of her chatting with Barack Obama at a White House ceremony last year to inaugurate a new law that mandates equal pay for women. 

As an organiser for Virginia’s 8,000 personal care assistants – people who look after the old and disabled in their own homes – Shareen, 42, was invited along with several dozen others to witness the signing. 

Ah…another representative household…..

More and more young Americans are put off by the thought of long-term debt.


Had enough?  I have speculated that to the extent economic insecurity has increased, it reflects the impact of a negativistic media (amplified by gloom-and-doom liberalism).

Pieces like Luce’s—and the blog posts it generates—affect consumer sentiment.Ben Bernanke and Tim Geithner aren’t the only people who can inadvertently talk down the economy.

*Originally said "just under 3 percent", which was incorrect.  -srw
8/3/2010 07:19:09 pm

This doesn't surprise. Luce works in D.C., an area with prices which have inflated more rapidly than average. I've long noticed that national and international media are locally biased.

My bet: the largest concentration of "middle class decline" articles issues from urban areas with high Case-Shiller scores. That is, reporters located in rapidly gentrifying areas are biased, seeing the "squeeze" everywhere when it is really a LA-SF-NY-DC phenomenon.

8/4/2010 02:46:27 am

You make some good points. The article in question by Luce is weak in some points. But it is strong in others, and, while you refute each point in turn, your data selection is flawed.

That parents felt better about their children's future between 1980s and 2004 does not mean much for the past 6 years not included there. You use "average" income of married couples. If you mean mean, and not median, then that stat is useless as the income curve is largely skewed. Women improve in mobility since 1950, though men do not. Well. I think that means mobility decreased, unless you really think that the fact that women are more mobile now than they were in the 50s is a great accomplishment. That is a social change, not an economic one. This is anecdote vs anecdote with selective statistics and verbage flying all over.

8/4/2010 04:58:44 am

Two questions:

1) Why are globalization and technology always blamed for allegedly hurting American workers (though again, I'm sure that different cuts of data related to these very broad terms show different pictures), but not illegal immigration?

If there are *officially* 12 million illegal workers in the US (and unofficially perhaps 20 million, given that some recent estimates place 10.8 million in California alone), that's roughly 10-15% of the workforce. In other words, a large proportion of the workforce is people who cannot communicate with the majority of US citizens and among whom a sizable minority (40% according to Pew) have less than an 8th-grade education.

That large proportion of the workforce is attractive to employers for the sole reason that it works for wages that are illegally low. Meanwhile, welfare programs artificially raise the cost it takes to get a similarly low-skilled American to work. So we have large numbers of Americans not working and, increasingly, a huge new underclass of unskilled, undereducated Central American workers doing the jobs that they are hired to do precisely because they are so cheap (meaning that if they should be legalized, their attractiveness would drop precipitously for employers and, as potential welfare recipients, the cost to get them to work would rise). That, to me, is the biggest source of economic inequality, lower wages for US workers, and likely social problems in the years to come.

But leaving that out is just one of the many problems with Mr Luce's distendedly grim view of American life. Scott, you pointed out many of the others, but there's one more you missed: "Mark calculated they have paid $163,000 so far on a house they bought for less than one-third of that amount. It could all have been for naught."

So they bought a house for $55K decades ago. Anyone want to do the math to see what $55K in 1980 dollars is today? I don't have my Excel open, but I would bet that that $163K doesn't look quite as jaw-dropping when you figure out the inflation-adjusted amount they initially paid. Not to belittle the people, but I think Luce is being more of a left-wing (or just media/mega-rich/academic/political class-wing) advocate than a journalist here.

8/4/2010 04:59:54 am

^^ Clearly, the "two questions" framework got left on the wayside. The second question should've been the inflation-adjusted price of the family's house.

8/4/2010 05:02:33 am

Also, Scott, your math is a bit off here:

"The share of mortgages either in foreclosure or 3 or more months delinquent is 11.4 percent, which, because 30 percent of homeowners have paid off their mortgage, translates into 8 percent of homes. So the Freemans’ situation is typical of about one in twelve homeowners, or not quite 3 percent of households (since one-third rent). "

If 2/3 of people own their home and 8.5% of homeowners are in a similar situation to the Freemans', that means that 8.5% x (2/3) = 5.7% of homeowners are in a similar position, not 3%. (You multiplied through by 1/3 instead of 2/3.)

8/4/2010 05:10:23 am

There's a concept in economics called "opportunity cost." It's relevant to this discussion.

8/4/2010 05:36:31 am

Ben--good catch. I fixed it--thanks!

Also a good point regarding their mortgage. By my calculations, their $50,000 mortgage in 1989 was $80,400 in today's dollars. If the house is now worth $73,000, as the article states, and if they are underwater (which seems likely), then they must have taken out a crazy amount of home equity debt. That would explain the repo efforts.

I'm not sure the American taxpayer wants to bail out the Freemans....

8/4/2010 05:42:22 am

Though I am a technologist by education, I have some skepticism on relying excessively on statistical analysis to understand social dynamics. Hence, since you feel Luce's article is very atypical in describing middle class household economics, what, in your opinion, might be unconsidered factors in Freeman and Miller households that make them less than accurate exemplars of their income/wealth class?

8/4/2010 06:00:45 am

Not to nit pick, but a couple thoughts about inter-generational comparison of household income:

1. Do you take into account non-cash benefits, such as employer-provided retirement and/or health insurance? I have no stats, but it seems plausible these may have declined over the period from 1973 to 2008.

2. How do you measure inflation? Basket of consumer goods? That works to a point, but certain expenses, e.g. health care and college tuition, have increase at a rate beyond general inflation.

3. Certain expenses may affect 2008 households at a much higher rate than 1973 households. For instance, college tuition. Even if the cost of tuition had increased at the same rate as overall inflation (which is not the case) there's still the matter of increased emphasis on college education. In 1973 the percentage of households that had to plan for financing their kids' college educations was likely much smaller than in 2008. So if inflation-adjusted median income had remained constant over this period, the 2008 household would still be "poorer" in comparison.

4. For two-adult households, what was the rate of one-income vs. two-income between 1973 and 2008? $29,000 with one income vs. $43,000 with two incomes could end up being a wash when you consider the opportunity cost of the second income. The two-income household sacrifices a great deal of time that could be spent by a non-earner to take care of daily tasks like cooking, cleaning, shopping, and child care. The two-income household will have the expense of having to outsource some of these tasks (especially child care).

5. Are your numbers for inflation-adjusted median household income measured before or after taxes? How has the total tax burden (federal, state, local) changed for households at the median income level changed over the period from 1973 to 2008?

8/4/2010 06:06:36 am


Illegal immigration is a convenient peg to hang many of America's social problems. However, I think, all the negatives of illegal immigration are also balanced somewhat by the excessively low inflation, especially in service sector, that Americans have got used to after the '70s. Without Ill Imm, we would have like seen much higher service sector inflation and much lower standard of living. Same with off-shoring & outsourcing too. Basically, as a nation, we implicitly exchanged short term elevated standard of living for the past 3 decades for long term social decay.

8/4/2010 06:16:53 am

What are the first featured people in this article spending their money on? They must be netting around $4,500 per month and have a mortgage payment of around $300/month.
Food and non food staples $1,000 / month
Utilities (incl cable & internet) $500 /month

2 car payments, alot of credit card debt for high-end electronics, medical bills, vacations, what?

Also, in any comparison of past and present on incomes, what about percent of income spent on food, utilities, housing (at the same home size)?

How much did my cell phone cost in 1973? Or my internet service in 1973? Wait a second, zero, cause I did not have that possibility at all.

Non-wage compensation growth (various insurances, etc.)over time has to be factored in.

8/4/2010 06:54:18 am


I fully agree with you with regard to the fact that illegal immigration has kept the price of services -- restaurant dining, landscaping, roofing -- artificially low, just as a minimum wage keeps the price of services artificially high.

I also agree that with that short-term service-sector price deflation we've taken longer-term economic and social problems. You provide amnesty and what? Suddenly, restaurant owners have to pay their workers min wage. Since they overhire based on artificially low wages (restaurants where I am in NYC always seem to have dozens of water-pourers), they'll be just as quick to fire when they have to pay workers legal wages. Since those workers have few skills other than being cheap, we may find millions of low-skilled workers on the dole in no time.

Contrast with Canada or Australia, which have as close to a market-and-merit-based system of immigration that you can get without having fully open borders. In those places, the newcomers are predominantly from Asia and rather than being an underclass of water-pourers and mulch-spreaders with intergenerational poverty issues, they're eating next to you in restaurants, working alongside you, their kids are competing with yours in school, and they're otherwise assimilating. No ghettoes, no (serious) talk of "floods." Instead, they get skilled labor that keeps prices down on IT and medical bills, rather than on shrubbery-cutting and restaurant meals.

My parents were immigrants, and so is my wife and many of my co-workers. I'd like to see a robust immigration policy, but if we've got the choice of doing so (which we do today) I'd like it to strive to raise the bar in terms of living standards, education level and earning power rather than not.

8/4/2010 07:13:08 am

Scott, in case you weren't aware, The Weekly Standard also cross-posted you:


8/5/2010 03:30:53 am

some good refutation, but the overall gist of the article is that the author has an agenda to shill, and the stats are just props. My favorite is that a nearly 2% rise in income over a generation would double that income -- let's see, using the rule of 72, a generation is 36 years. Well, who could have problems with that?


Your comment will be posted after it is approved.

Leave a Reply.