Household size and stagnant median income

One loyal MR reader writes to me:

However, census Bureau data show that the size of the average US household decreased from 3.1 to 2.6 from 1970 to 2007…

The underlying question is whether figures for the median household are underrating the true growth in average living standards.  A few points are in order.

1. Here is one passage from The Great Stagnation: "Since 1989, the size-adjusted and size – unadjusted measures have been rising at roughly the same rate, and post-1979 the difference between the size-adjusted and the size-unadjusted median income measures is never more than 0.3 percent."  For more on this, see Lawrence Mishel, Jared Bernstein, and Heidi Shierholz, The State of Working America 2008/2009, chapter one. 

2. David Leonhardt writes: "In fact, households were shrinking more quickly in the 1960s, ’70s and ’80s than they are now – and incomes were growing."  Read the rest of this post as well.

3. That households are smaller decreases the aid and assistance available to those living in them. 

4. There are an increasing number of women in the labor force, and that factor biases the household number to be higher than true productivity growth alone would dictate.  Unmeasured household production is a mix of "lower than it would have been" or more harried than it would have been.  I suspect this is a large effect, not a small effect.

5. On the issue of students and retirees, see the adjustments performed by Lane Kenworthy, pp.37-38.

Overall I do not see that changing household size allows one to dismiss the notion of relatively stagnant median income.


This sounds to me more as a distraction from the real issues.

Make the issue some vague "stagnation", an assertion that people aren't doing enough now.

Don't let people notice that it now takes two workers to earn the same income per household, while the rich have been fighting a class war against the middle class to try to push them back to Malthusian wages and claim the wealth for themselves. As Krugman's recent posts show, the median would be significantly higher if inequality hadn't been growing.

Comments for this post are closed

The charts in (5) would suggest that stagnant median income in the face of rising mean income reflects changes in distribution, not stagnation in innovation. cf. Sweden over the same time period.

Comments for this post are closed

@boris, in the common sense of the words, concentration of gains to income among the top 1% of earners is a more dramatic increase in inequality than concentration of those same gains among the top 10% of earners.

Comments for this post are closed

The comment about swimming pools (@ris) rings very true. Note, though, that most of the cost of a new built-in pool today is imposed by regulation (at least in the Northeast).

Comments for this post are closed

Aren't the small differences between the adjusted and unadjusted totals compounded over time? Isn't a difference of just .3% of income in a single year a significant difference?

If households were shrinking faster when income growth was faster, doesn't that suggest that the notion that household size shrinking should be understood as an effect of slowing income is probably wrong?

What difference does the availability of aid have on the median income numbers? I'd think none at all.

Unmeasured household production is surely reduced, but without necessarily implying a reduction in living standards. In any case, we should note that unmeasured household production was likely falling faster before the great stagnation, because of increased automation and because households were actually shrinking faster then.

Comments for this post are closed

Does anyone other than Sailer ever realize that household size is diminishing because its so damn expensive to provide well for your kids.

Comments for this post are closed

complaining about lack of holding household size constant? was your loyal reader Thomas Sowell?

Comments for this post are closed

#5 is good, but it seems to show that a substantial portion of the stagnation is due to immigration and household composition (maybe half). It states that the value of benefits versus wages has not increased in the last 40 years, which surprises me greatly and I question. It also is not clear how inflation is adjusted for.

Comments for this post are closed

"Josh, Do you have evidence/a link for that? The only thing expensive about kids is child care, which used to be done by the mother and still can be. Otherwise, they just run around outside and play with boxes and eat tiny tiny portions of food."

Tuition for my 7 year old is 18k per year....we live in what is suppose dto be a good Boston suburb school district. After a year and a half of public school we realize that all three kdis will be in private school. The "no child gets ahead" philosophy that now drives our federalized school systems was not as dominant in decades past.

3 kids education = $54k/ after tax dollars...I thought we could afford more kids but we have decided to stop at 3...gonna go get clipped soon.

How is the decline in our school system quality being factored into the deflator?

Comments for this post are closed

Posted by Gary J. Mallast on 2/4/2011 10:09:38 AM

Inflationists are amazing. For 300 years going back to John Law's "Money and Trade Considered with a Proposal for Supplying the Nation with Money, " (Edinburgh, 1705) inflationists have claimed the ability to benefit and even enrichen people through producing fiat money and to control its effects. And for 300 years they have failed, dozens of times over, in every country from Argentina to Zimbabwe. In the process wreaking untold economic havoc wrecking the lives of millions of people. Of course the whole scheme could be refuted by the old expression, "You can't eat a dollar bill." Real wealth, the things people really need to meet their needs and desires, has to be produced. Those things can't be wished, decreed, printed, taxed into being, or brought about by creative book-keeping. Universal stealing whether conducted by a run of the mill stick-up artist, taxation, or inflation does not enrichen. It just moves goods around enrichening some at the expense of others. While improved productivity enrichens all. But the faith of the inflationist (and other statists) in their own ability to create food, housing, transportation, health care, knowledge, and more out of nothing by coercion and funny money seems totally unshaken. Amazing!

Central planners are amazing too. As F.A. Hayek pointed out, the knowledge necessary for production is dispersed over the entire population and cannot be concentrated in place. Much of such knowledge is time and place-specific. Much is trade or profession-specific, and much is uncertain so the acting person must take calculated risks based on his knowledge. And all this dispersed activity can only be rationally coordinated by a viable, i.e. free, price system based on honest money as Ludwig von Mises pointed out. But the statist planners, whether regulatory or monetary planners, in spite of failure after failure, still think they can micromanage everything better than all the people with that time and place-specific and profession-specific knowledge to make decisions which the millions of producing citizens have to bear the consequences of themselves. Amazing!

Comments for this post are closed


I'm not really sure how this argument is "new". Isn't the conclusion that we are seeing investment and gains elsewhere in the world because it is easier to make gains elsewhere in the world?

The novel part of the argument seems to be technological innovation has necessarily slowed (which is scary). That said, I don't think that portion of the conclusion is correct.

I'm don't think it has become harder to build a better mouse trap, simply you can make a crappier one in China and get a better return than building a better one in the U.S.. Accordingly, the only innovations in the U.S. that are worthwhile involve how to manage American capital overseas, i.e. assisting in running larger businesses: communication and data management. So our focus is not building a better mouse trap, but building a better way to control mousetrap production in China.

That would also seem to predict the rising mean income and stagnant median (i.e. our rich are becoming richer because they are less focused on the U.S. and it is easier for them to take risks in less well regulated economies).

If you accept the foregoing, the real meat left to argue about in the political debate is how best to react to the situation? Do we cut taxes and let the free market reign to let the rest of the world catch-up quicker? Do we try and ensure transparency abroad, a more equitable capital distribution in the U.S. and a tax code that lets more Americans exploit foreign markets as investors? Is responding to the stagnation now too late in any event because given the rates of growth elsewhere, we can no longer control it?

If the answer to the last question is "yes, it is too late" the right likely has some explaining to do for the past few decades.

While I agree encouraging innovation in the U.S. is a good idea, the point is trite in that we've been talking about it for years.

The real overarching issue is not how do we encourage innovation, but how do we preserve the American attachment to individualism and individual production, when our position in the world requires us to become a nation that has to invest in others, which is truly the humanistic trope that has changed.

Comments for this post are closed

The Great Stagnation is even reflected in this blog, where it appears all topics of discussion have been exhausted other than the book "The Great Stagnation." I'm just kidding. I love the discussion and don't blame you one bit for promoting it. Just couldn't resist the wise-*ss comment.

Comments for this post are closed

Why is "affording to buy a new car every few years" a sign of past wealth? Seems to me that's a sign of crappy cars and comparative waste vs. the present.

One factor that much of this discussion misses is that most of the goods people buy today are vastly cheaper in terms of the labor time needed to buy them and of MUCH higher quality. The fact that most cars get 100K miles and last for a decade or so with minimal repairs needed (compared to cars from the 50s or 60s) represents GAINS in real wealth that are missed by aggregate and median income figures.

The household with the median income today has more stuff and better quality stuff than did the median income household at any time in the past.

Comments for this post are closed

Josh, you will never convince me that $18k tuition for a 7-year old is necessary. It would probably be cheaper to hire a tutor or just move to a better school district. Growing up, I never in my life thought "I hate this public elementary school so much, I wish I was dead." I would never not have a fourth kid because I wanted to spend $54k/yr to place my other three kids in a ritzy private school. Your response seems completely out of touch with reality.

Comments for this post are closed

Can someone provide some clarification about how you go about defining a "big idea" other than as one that produces a big change?

If the factor is novelty, why can't a small idea, e.g. a slightly better transistor, produce as many changes as a big idea, like a car? Further, how can you predict in advance that a car is going to be a successful big idea or the wasted capital in producing a bad idea will not be outweigh any benefits of a good idea required by similar regulation?

Comments for this post are closed

"you will never convince me that $18k tuition for a 7-year old is necessary."

You'd probably see a greater return on better/higher education the earlier the child is enrolled, as capacity for learning is highest at those ages. I'm not sure it would justify paying almost double the average per pupil spending of public schools though. Maybe if you are already in the elite class.

Comments for this post are closed


Gabe Harris responded to you before. Not me.

My child will go to public school, but the extra cost of a house somewhere in a decent school district may as well be tuition. Also, college is very, very expensive. How many kids am I supposed to send to college? I don't want my children to have ridiculous amounts of debt for the rest of their lives. Do you know how much my parents paid for college? ~$1000 per year.

Comments for this post are closed

Unless you hold age and work experience constant over different periods, changing population characteristics will unduly affect the median income number. Younger people with less experience in any field generally are paid less than more experienced workers and adjustments are necessary for the post WWII baby boom and several subsequent mini baby booms as these population waves move through the workforce and skew the median numbers. Also, workers who reenter the workforce after being out of the workforce for a number of years, such as married women and married mothers, while divorce rates were accelerating, would skew the median HH income data lower as they lack work experience of others their age and form their separate HH. Additionally, adjustments need to be made for changes in immigrant workforce percentages. The immigrant workers are usually adults with less skills and are paid less.

Younger workers, more immigrants and accelerating divorce rates will prevent median HH income from rising.

As to Boris who noted that the top 1% have a big impact on the inequality measures after 1993. The top 1% is not a fixed population set and is an ever-changing mix of individuals. It been shown that the income of this upper group has shifted from investment income to earned income with an increasing portion of one time gains from realized income of equity and option gains as a result of the changes in the tax laws that pushed executive compensation away from salary into deferred equity incentive compensation. Recent research has shown that a large portion of the excess of deferred compensation over a yearly salary is due to the extra risk associated with deferred equity compensation over a yearly salary payment.

Comments for this post are closed

Free massages

Comments for this post are closed

Yakov A.

Had a good time reading your post.

It is true that it is difficult to conquer the trick that will lead us to bigger and better innovation; but, surely, you will agree that we still live today from a few critical developments owed, essentially, to individual human beings which were knowledgeable, curious, ambitious, etc; and that these developments took place well over 60 to 80 years ago.

Today these people cannot really exist. They have been replaced by big private and public institutions.

These institution's priorities, grants and budgets are a far cry away from those in the minds and souls of the "lone" inventor of the past.

In spite of huge volumes of patent registration and related activities their real impact is only a fraction, if any at all, of the breakthroughs of the past. Very, very few exceptions if any.



Comments for this post are closed

In 1950, 50% of American homes did not have indoor plumbing - for Virginia data, see:

I feel that a large amount of US poverty comes from reduced working hours of single parents do to child care issues. We should remember that child care is a highly regulated industry and formal child care is very expensive.

Typing this on my iPhone after watching a DVRed TV show in high definition, I don't feel very technologically stagnated.

Comments for this post are closed

Does anyone other than Sailer ever realize that household size is diminishing because its so damn expensive to provide well for your kids.

The story of total (white) fertility rate in the U.S. since 1800 is that it steadily declined from 1800 until WWII, then ticked up from the end of WWII through the 1960s, declined again through 1980 and then experienced a small uptick in a mini-baby-boom.

Fertility in the U.S. is high by developed country standards and, of course, extremely low by historical and/or poorer country standards.

Comments for this post are closed

Tyler writes,

Overall I do not see that changing household size allows one to dismiss the notion of relatively stagnant median income.

You and Lane Kenworthy are using the Census Bureau's current official measure of income, which is "money income excluding capital gains." As the Census Bureau itself notes, this measure has some very serious problems. Among those problems is that it does not account for the effects of capital gains and losses, health insurance, taxes, and cash and non-cash government transfers. The Census Bureau provides income data for alternate household income measures that account for these other factors.

Unfortunately, the historical data cover only the period 1979 to 2003. But that is enough to show how much things change when the other factors are taken into account. The data is here:
I just did some quick calculations. Using income definition 1 (the definition you and Kenworthy are using), median real household income increased by 12% between 1979 and 2003. But using income definition 15, median real household income increased by 20%. That's quite a difference.

If we make an adjustment for the change in average household size, the difference is even greater. In 1979, the average household size was 2.76. By 2003, it was down to 2.57. Using these numbers and income definition 15, I find that median household income per household member increased by 28% between 1979 and 2003. Again, this is a huge difference from the 12% increase that you get using income definition 1 and failing to adjust for household size.

Another way to illustrate this is by looking at the data on income shares ( Using income definition 1 (again, the definition you and Kenworthy are using), the middle quintile share of total income fell by 2.5 percentage points, from 17.2% in 1979 to 14.7% in 2003. But using income definition 15, the middle quintile's share fell by only 0.8 percentage points, from 17.2% to 16.4%.

This is another illustration of why I just don't believe that the median income data you are relying on provides a realistic picture of changes in the typical American's standard of living since the 1970s.

Comments for this post are closed


I do not undersrtand why you believe innovation which primarily benefits the U.S. is important. Not sure I believe such innovation ever existed. Can you give me any examples?

Comments for this post are closed

To be more concrete: try finding a block heater on a fiat.

Comments for this post are closed

Yakov, as a NZer, I'm surprised at your claim. For a start, NZ has seen continued increases in farming productivity. NZ, Australia and most of Western Europe doesn't get as cold as Canada, so car development benefits all these countries along with the US. Plenty of NZers and Aussies hunt with handguns, so improvements in hunting equipment benefit them.

Innovations in story-telling - the British are great at that, eg, Monty Python, Tom Stoppard (well, Czech, but raised in Britain), Andrew Lloyd Webber, the Beatles, Terry Pratchett, Douglas Adams.

Innovations in sport - football, cricket, rugby, are mostly not American games. How do you measure innovation in sport anyway? I note that America no longer always wins the America's Cup in yachting.

What are these diseases that affect American citizens, but not people in the rest of the developed world? I know, or have known, too many non-Americans who have had heart attacks, or developed cancer (and sometimes survived it), or schizophrenia, or had obesity fixed by that new stomach stapling treatment.

Comments for this post are closed

Last sentence should read: "Innovation and a well functioning economy may not be enough with globalization to continue to see the massive technology based improvement in living standards in the United States we have seen in the past."

Comments for this post are closed

Population still may have something to do with it. If economics is based on the idea that value is created when 2 people exchange goods and services, and even more value is created when 3 people exchange goods and services, then since the population growth rate dropped after the 1960s, you might expect a drop in the rate of increase of median income to occur when those people would have been expected to enter the workforce in the mid-70s to 80s.

People at the top may be doing relatively better because they effectively trade with the entire world, where population growth hasn't slowed.

Since the top 20% has 80% of the money (order of magnitude), then GDP growth wouldn't see much of a drop in growth rate.

Comments for this post are closed

Aspects apparently not taken into account might say a lot of things about the potential subjects for this study. One may focus on home surface, big rooms, spacey halls, others would rather have the latest fashionable releases in terms of
bathroom vanities even if they own smaller household size. When these three factors come together, there is one social studies statistical law that says there might appear an interaction effect between two factors that distorts the particular effects of these two upon the third one.

Comments for this post are closed

Comments for this post are closed