David R. Henderson  

Raj Chetty on Income Mobility

PRINT
A right-turn for Switzerland?... Tony de Jasay, the greatest th...
Mr. Chetty and the economists he works with tackle problems that seem intractable, and offer hopeful prescriptions. Consider economic inequality--the income spread between rich, middle-class and poor. Mr. Chetty addresses the issue indirectly. He examines income mobility, which he defines as the ability to rise from the lowest 20th percentile of income distribution to the top 80th percentile in one generation. Climbing that ladder is more important than ever, he says, because the distance between the economic classes is greater than in the past.
This is from Bob Davis, "Economist Raj Chetty's Proposals on Inequality Draw Interest on Both Sides of the Political Aisle," Wall Street Journal, October 20, 2015 (electronic.)

Reread that last line:

Climbing that ladder is more important than ever, he says, because the distance between the economic classes is greater than in the past.

That's exactly wrong. A greater distance between top and bottom makes it less important to get from the bottom to the top. Now it's possible that Davis, since he doesn't give a direct quote, badly mischaracterized Chetty's thought. The important thing here, though, is to get the thought right. And that thought is wrong.

To see why, take an extreme example where the range of income in the bottom fifth is a family income of 0 to $50,000, the range of income in the second fifth is $50,001 to $100,000, the range for the middle fifth is $100,001 to $150,000, the range for the fourth fifth is $150,001 to $200,000, and the range for the top fifth is $200,001 to infinity. This is a case where the distance between economic classes is even greater than in the past. Now, the thought above, whether Chetty's or the writer's, is that it matters even more than it used to that the family between $0 and $50,000 make it to the $200,001 level. See the problem?

With the second quintile being $50,000 to $100,000, if you make it to that one, you would be doing pretty well. If you make it to the middle quintile, you would be doing great.

HT2 Arnold Kling.


Comments and Sharing


CATEGORIES: Income Distribution




COMMENTS (21 to date)
Jack PQ writes:

That's a great insight, but I think there must be a way to get the point across more easily.

Suppose you say: If there is no income growth in the economy, but there is greater inequality, then it is more important than ever to improve mobility because the poor are poorer than before and it is harder to get to a decent income level. But if there is substantial income growth in the economy (which is the case), together with greater inequality, then the poor do not have to "travel" as far across income levels to reach a decent income.

Hmm. That is not clearer. But you see what I mean.

Brad writes:

I like how he put it: "climbing that ladder is more important than ever…”

This implies that individuals should take positive action to advance themselves economically, as opposed to riding the government-transfer-payment-elevator into higher income brackets. I doubt he meant to convey this point, but it's worth mentioning.

Kevin Erdmann writes:

I will consider economics, as a discipline, a failure until someone finds a way for the bottom quintile to be less than 20% of the population.

Let's stop beating around the bush. That is the problem. And I don't see anyone addressing it.

I guess I think of the words "more" or "important" differently than you do.

When the income differences between quintiles is larger, it matters *more* which quintile you are in not less.

Consider the opposite extreme: if there were no difference in quintiles, then it wouldn't matter at all (at least in terms of income).

Perhaps you don't think climbing the ladder not very important for us to focus on. But is your argument really that climbing the ladder to a higher quintile becomes *less* important to individuals if it makes a *bigger* difference in their incomes?

Vivian Darkbloom writes:

I suspect that Chetty was misinterpreted or that his idea was badly paraphrased--or perhaps even that the paraphrase in the WSJ is being mis-interpreted here. Remember, the discussion is about income *mobility* (as set up in the penultimate sentence of that quote in the WSJ). Mobility is not the same as absolute economic well-being. The idea seems to be that being "mobile" is not synonymous with moving up only one rung on the ladder, even though moving up one rung might substantially improve one's income and economic condition. The idea seems to be that to be truly *mobile*, the ladder has many rungs that must be climbed.

If we are talking about a definition of *mobility* as the ability to move from the 20th to the 80th percentile (not an unreasonable definition for purposes of a discussion of mobility), then it is also not so important that everyone actually does climb that ladder, but that our social and economic system gives everyone the opportunity to do so.

ThomasH writes:

Seems like a different distribution would give you a different result. If the differences between the first four brackets were small and only the the top was much different from the bottom, then climbing to the bottom top would be more important than when the progression were more uniform.

More generally, the importance of mobility (which seems more related to the fairness of the process of climbing) does not depend on the differences between top and bottom. Rather, the legitimacy of the differences depend on mobility.

James Hanley writes:

I'm intrigued thst folks akways talk about people climbing the income ladders without recognizing that thos requires others to fall down the ladder. And while that might sound fair if we're talking about Bill Gates, what it really means--speaking loosely--is that if I climb the ladder, mu kids or grandkids need to fall back down to make room for someone else. Ans how is that fair to them?

Focusing on percentiles turns wealth gains into a zero-sum game, and emphasizes relative status over absolute status. It has some analytic value, to be sure, but tells us little about people's well-being.

That's not to imply David is doing that, as he isn't.

Swami writes:

It would be true if he said climbing the ladder is more BENEFICIAL as the ladder now goes higher.

It would also be true if he said people are better off NOT CLIMBING compared to past non climbers as the high end of the bottom twenty percent is so well off compared to past eras.

I think both of my statements are true, and they shed more light on the situation. My wife and I both retired early (in forties) and live a wonderful life of leisure off savings though technically in the bottom twenty percent of earners. Several of our younger relatives thrive off the ample welfare nets with an occasional part time job. I could provide details, but I get the sense young males are starting to realize workers are either suckers or neurotics.

In general I find writing on inequality and mobility is mostly a reflection of a certain popular paradigm which does not hold up to scrutiny. There is no "problem with mobility" if people are not capable of or willing to do what takes to move ahead. Indeed, the system is pretty much designed to work this way and for good reason. You move up economically by solving problems for other people. If you refuse to solve problems for others, or are incapable of, or you try and fail to meet their needs, you are supposed to fail. This is how the system tells you to do something different.

With associative matching/mating and safety nets which discourage producing for others, I pretty much assume lower mobility is the attractor for families in the lowest quartile.

The solution is smarter (not necessarily smaller) safety nets, fewer barriers, egalitarian rules (impartial, fair), and lots of kindness and empathy for those with two bad parents, or even worse one bad parent and one missing.

Baconbacon writes:

@ Erik,

The quote specifffically says that the measure used is climbing from the bottom to the top quintile- the larger the gap between the two the larger the ignored gains of those whose move from the bottom 20 percent to anywhere between 20 and 80 percent.

Andrew_FL writes:

Note also that if people are as upwardly mobile in income as ever, then by Chetty's definition upward mobility will have declined. That is to say if the typical person in the bottom quintile in early period increases their income over their life time by a certain percentage, and the same percentage in the later period, but the incomes of individuals in the top quintile are higher by a greater percentage relative to those in the earlier period than those in the bottom quintile, by Chetty's definition their upward mobility has declined even though they increase their incomes over their life times by the same percentage they always have.

Kevin Erdmann writes:

Btw, I think everyone may be misinterpreting the quote. I read it as saying mobility is going from the lowest quintile to any other quintile (the top 80%), not just the highest.

Philo writes:

I think David Henderson's point is that as a society becomes wealthier, differences in money income matter less, because of the diminishing marginal utility of wealth. In a poor society where the bottom quintile income is, say, $2,000 and the top, say, $10,000, moving from the bottom to the top makes you much better off, though the increase is only $8,000. In a rich society where the bottom quintile income is, say $50,000 and the top, say, $1,000,000, the move from bottom to top is much greater in dollar terms but probably lesser in terms of well-being (happiness, utility). If we measure amounts of inequality *monetarily--say, in terms of (inflation-corrected) dollars*--there is likely to be more of it in a rich society than in a poor one, but it will probably be less *important* (it will *matter* less).

(I admit, it is not *clear* that this is his point.)

Philo writes:

@ Kevin Erdmann:

Bob Davis's statement about Chetty's definition of 'mobility' is garbled, but surely no one would think that rising from the 19th percentile to, say, the 21st was very significant. And the percentages Chetty gives for various cities are so low, he must have in mind rising from the range of percentiles 0-19 to that of percentiles 80-99.

Brian writes:

I'm sure that Chetty was quoted accurately. The issue, as always, is whether relative income or absolute income matters the most. The former category is favored by the left, and Chetty likely falls into that category. His argument would be that with a larger gap between quintiles, it becomes harder to move from on e quintile to another, and that's bad if relative income matters.

Interestingly enough, Chetty's data implies that the U.S. is a perfect meritocracy. That is, the distribution he gets is almost exactly what one would expect if genetic endowments explain just less than 50% of outcomes. So the question I would ask of Chetty is whether, as an economist, he can in good conscience advocate for more mobility than talent and hard work will allow.

David R. Henderson writes:

@Kevin Erdmann,
I don’t think so. Not if the quote is accurate. Look at it again:
"which he defines as the ability to rise from the lowest 20th percentile of income distribution to the top 80th percentile in one generation.” Notice the top 80th percentile, not the top 4 quintiles. If he had meant quintiles, he presumably would have said it. Instead, he said “percentile” and notice that it’s percentile, not percentiles.

Kevin Erdmann writes:

If he had said, "top quintile" or "top 20th percentile", then I would agree. Lowest 20th implies starting at the bottom and counting up 20. It seems to me that top 80th would imply starting at the top and counting down 80. I think, out of context, if you gave me a bar and asked me to fill in the top 80th percentile, I would fill in the uppermost 80% of the bar.

I don't know. I may be wrong. At first, the wording seemed weird, and I thought for a second, and then that is how it seemed like it would have to be read. And, conceptually it seems more useful to think of the ability to become average as opposed to measuring a society based on moonshots.

Kevin Erdmann writes:

No...you're right. I should have been more careful before I commented. I looked at the article, and there are other references to getting into the top quintile. That is what he is talking about.

Nathan W writes:

In life as experienced by us individuals, a lot of the goods that matter most to us are positional.

So ... everyone can afford a TV, computer, internet, decent clothes and an OK apartment now. But EVERYONE has that stuff.

Since positional goods are very important, a greater gap between 20% and 80% implies greater difficulty to achieve status, which is actually what matters to a lot of people more than their creature comforts, which are relatively easily attained in the modern day.

I think it is important to never stop observing that everyone is gaining (so long as this is actually true), but this fact does not mean that lack of social mobility is not important.

Nathan W writes:

If I'm in the 0-20%, and positional goods are priced largely according to the incomes of the 80-100%, then when there is a greater difference between the 0-20% and the 80-100%, then it becomes all the more difficult to afford to go to bars, conferences, clubs, schools, live in neighbourhoods, etc. where the 80-100% congregate and network. And besides the fact that you miss out on the networking opportunities to help you land a fat job or rub noses with potential investors, the 80-100% will then be MORE able to price you out of scarce quality resources required to start a business, such as the services of a good lawyer, a good accountant, a good web designer, and the list goes on.

While this should only be framed in a context where the benefits of strong incentives (getting to keep a decent share of your hard-earned money) are understood as very important, this means that a "too large" income difference (defined as "too large" by whether or not it has the effects that I discuss) between the 20% and 80% is very likely to have a very direct and negative impact on social mobility. That is not desirable if we are to make the best of all talents, efforts and preferences in a society.

A "too large" difference between the 0-20% and 80-100% is not a good thing.

liberty writes:

it's the same point I made in my 2005 Luxembourg Income Study working paper.

R Richard Schweitzer writes:
He examines income mobility, which he defines as the ability to rise from the lowest 20th percentile of income distribution to the top 80th percentile in one generation. Climbing that ladder is more important than ever, he says, because the distance between the economic classes is greater than in the past.
(emphases added)

Once again we observe a study of what ought to be the cause of a perceived problem, in this case "ability" rather than study the factors that may contribute to it; and with what relevant effects.


Other Than from Charles Murray (and in a different mode, Putnam and Lynn)there seems to be no consideration of MOTIVATIONS; of the factors that affect the initial formation of motivations - and the factors that affect their subsequent development.

Perhaps Chetty is back where Sen was in 1978, at "capacity," simply re-labeled "ability" or possibly to bring in the effects of externalities.

Examining one generation should encompass the impact of the previous generation on the motivations of the one examined. Are the motivations formed in Fishtown different from those formed in Belmont. What are the externalities experienced by that one generation?

Comments for this entry have been closed
Return to top