Sentences about income mobility

So, if everybody’s income is 50 percent higher than their parents’,
that shows up as zero income mobility.  On the other hand, if average
income stays constant, but a lot of low-earning parents have
high-earning children, and vice-versa, that shows up as high income

That is Arnold Kling.  Here is much more from Brink Lindsay.  He writes:

…is measured income mobility between generations really on the decline
in the U.S.?  Not according to Gary Solon of the University of Michigan,
who is probably the leading American researcher in the field.  (He is on
the advisory board of the Economic Mobility Project.)  His most recent research
on the subject, in a paper co-authored with Chul-In Lee, concludes that
“intergenerational income mobility has not changed dramatically over
the last two decades.”

I once wrote (read the whole thing):

1. "Age-adjusted parental wealth, by itself, explains less than 10 percent of the variation in age-adjusted child wealth."

2. 20 percent of parents in the lowest quintile of the parent’s
wealth distribution have children who end up in the top two quintiles
of their generation. One-quarter of the parents in the highest wealth
quintile end up with kids in the two lowest quintiles.

3. The age-adjusted intergenerational wealth elasticity is 0.37.
What does this mean? If parents have wealth 50 percent over the mean in
their generation, the wealth of their children will be 18 percent above
the mean in the childrens’ generation.

That all said, I think there remains a very real disaggregated problem concerning income mobility.  Given the rising size of the college earnings premium, many more people should be going to college but are not.  Not every country in Europe has the same problem.  Here is my previous column on that topic.  Since I do not think there is a single unique bottleneck preceding higher education, and many of the issues are sociological, I do not think this problem will prove easy to fix.


Yes, what two letter concept could possibly explain why the return from education is high for some and low for other†¦

If only economist could spend 30 years more research on this amazing puzzle. In the meantime let’s assume it is due to non-existent liquidity constraint for the poor, and spend billions of taxdollars on it.

“ Not every country in Europe has the same problem.†

According to the OECD no country in Europe has a higher share of the adult population with tertiary degrees as the US. (Among the 25-34 population the four smaller European countries have countries have higher, but the difference is miniscule, 0,2-3,2%).

2004 OECD

1. Canada 44,6%
2. US 39,1%
3. Japan 37,4%
4. Sweden 34,5%

OECD average 25,2%
Germany 24,9%
France 23,9%
Italy 11,4%

Maybe the return to education is so high in the United States that 100% of the population should be encouraged to pursue advanced college degrees. Who could possibly be wrong with such a plan?

So why aren't more people going to college? Is it because...

a) They're not qualified and can't get in
b) There's not enough college slots available
c) College overhead costs (professor salaries, ahem) are too high, forcing tuition costs higher to cover that overhead
d) Some other explanation?

Part of the problem is that people frequently confuse the average increase in income with the increase in average income.

The best evidence, such as Card (2001, Econometrica), cited on this blog, suggests that the marginal returns are there as well, not just average returns.

I think (though some would disagree) that the literature is pretty clear on the earnings effects of college quality.

That is, the literature is pretty clear that college quality has only a small effect on earnings.

Rich, the high schooler doesn't have the $100,000 to invest when he's 18 unless he goes to college. (No one's going to lend a new high-school graduate $100K to invest as he wishes.) So in one case, he works 40 years, and earns $1,000,000, and in the other case he works 36 years and earns $1,800,000. Who wouldn't choose the second?

Not to mention the fact that in the second case he's also smarter, understands more about the world, is more likely to meet smart women, will be a more attractive marriage candidate, etc.

Kling's complaint about the metric is valid. A society in which everyone's lifetime income profile was randomly assigned at some point in childhood would have maximal income mobility (zero correlation between parental and child positions in the income distribution), but many would have a hard time calling such arbitrariness, even arbitrariness uniformly applied to all, just.

Over the years there has been widespread criticism of the various studies of intergenerational mobility, largely because they were snap shots of how it existed at time. Many conservatives, libertarians argues that they understated mobility. But recently there has emerged a school of research that looked at intergenerational mobility in terms of total lifetime earnings.
But these studies have found that intregenerational mobility was actually much lower then the snapshot type studies found.

If I am not mistaken these studies would suggest that your estimate that the age-adjusted intergenerational wealth elasticity is 0.37 is significantly too high.

On the one hand, with respect to college, interest rates would have to be much higher before an education that was worth buying was not worth buying with borrowed money.

On the other hand, almost all of us are akrasiacs, though the manifestation and the intensity of this akrasia vary among individuals. It may be that a large fraction of us ontologically cannot do higher education, even though we recognize it as the best thing to do. To improve the human capital of this fraction of the populace, preparatory education must cause them to not just become greater than they are, but other than they are.

The return on education is only overall welfare-enhancing to the extent that it results in an increase in the supply of consumption goods, either directly or indirectly. Otherwise, larger salaries for graduates only enables them to outbid non-graduates for existing consumer goods.

Regards, Don

He's going to save $100,000 in four years by doing that?

This is ignoring the point. A high school graduate will have certain expenses over the next four years of his life. He can either work and make money for those four years, or he can go to college, not make money, and accrue debt.

Four years later, I am presuming, the college grad's net worth will be a lot lower, because he has foregone $100,000 in income, and accrued college debt.

The question is whether the $800K accrued over the next 36 years compensates for the $100K+ foregone at the outset.

If you make realistic assumptions that there are no extra assets for the student who goes to college, the numbers are not obvious one way or the other.

Probably because economists have such high IQs, which largely determines success in life. People who are successful in life will not dare mention IQ in civilized debate.

This leaves only a certain group of people who will dare mention those two letters.

Rich B.,

You wrote:

Rich, the high schooler doesn't have the $100,000 to invest when he's 18 unless he goes to college.

Sure he does.

And you claim I missed your point? You want to argue that college is a bad investment, go ahead. Lots of data out there. But learn to do the math. A return of $25K/yr (Difference between $25K for HS grad and $50K for college grad in your example) for 36 years on a $100K investment is just under 25% a year. Make the investment $200K and you get over 12%.


Kling's complaint about the metric is valid.

No it's not. If we are talking about the degree of intergenerational mobility it is irrelevant. He is merely stating a simple arithmetical point.


The mistake you are making is to assume that the difference in income is paid as a lump sum at the end of the 36 years. It's not. It's paid on an annual basis. That makes the return much higher.

Or look at it like this. You invest the $100K at 5.5%. Your annual income is $5500 in inerest plus $25K in earnings, $30,500 total. Or, you go to college. Your income is $50,000 a year. Which do you like better?

The best way to calculate is to compare it in terms of an annuity.

If the High School student were to save the 100,000k and invest it in an annuity that pays out over 36 years, what would he get? Using a random annuity calculator (, assuming a starting point of 100k, a 7% growth rate on the 100k principal, and a payout over 36 years, this would net the high school student approximately 7.2k a year. In order to get 25k a year he'd need to get a return of approximately 33.5%, which is extraordinarily high!

The time-value of money makes a gigantic difference -- money upfront is worth oddles more than money far in the future. You can't just add income up -- it is non-sensical.

This is especially evident when you consider that it's far more useful for a young person to have money up front for all of life's expenses. :-)


Comments for this post are closed