Why social mobility is important

By Felix Salmon
July 30, 2012
Tim Harford is a fan of the clear way in which Alex Tabarrok has couched the debate -- which started with a Tyler Cowen post back in January -- about the desirability of intergenerational economic mobility.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Tim Harford is a fan of the clear way in which Alex Tabarrok has couched the debate — which started with a Tyler Cowen post back in January — about the desirability of intergenerational economic mobility. Or, in English, is it a good thing if quite a lot of poor people become rich?

The Marginal Revolution guys say that looking at economic mobility is overrated; Cowen, also in January, linked to a bunch of critics of that position, including John Quiggin, Brad DeLong, and Paul Krugman. Recently, DeLong resuscitated the discussion, and Krugman came back for a second go-round as well, all of which resulted in Cowen being rude about Krugman, and Tabarrok trying to clear things up.

Tabarrok’s post is indeed clear, but it’s clear in an invidious way. He basically starts with his conclusion, saying that if a high-mobility society has no better outcome, in general, than low-mobility society, then there’s not very much to choose between them. And similarly, he says, if both a high-mobility society and a low-mobility society have the same very good outcome, then again there’s not much to choose between them.

But this obtusely misses the fundamental reason why high mobility is a good thing: that it improves outcomes. A sclerotic society where no rich people become poor and where no poor people become rich is never going to be a hive of creative destruction. Cowen even comes close to admitting this, when he says that “if the general standard of living is rising, mobility takes care of itself over time” — except he has the causality largely backwards. If you have lots of social mobility, then the general standard of living is going to go up: you’ll have lots of poor people becoming richer, and you’ll also have the rich protecting their downside, in the likely event that they become poorer, by doing their best to improve the lot of the poor.

So when Cowen talks about economic mobility not mattering much “for a given level of income”, or when Tabarrok talks about “some simple societies” with fixed levels of income, they’re taking the variable in the equation and they’re turning it into a constant. What they should be doing is looking at two societies, equal in all respects except that one is high-stasis and the other is high-churn, then fast-forwarding to see which one turns out better. The answer, of course, is the high-churn society — which means, working backwards, that if you want growth, you also want social mobility.

As a result, it’s reasonable to conclude that anything which impedes social mobility — like rising inequality, say — also impedes growth. The effect might not be huge, but it’s there. And the only way not to see it is to effectively assume your conclusions.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

A society where some poor people get rich will have a lower probability of revolution than a society where no poor people get rich.

Posted by KenG_CA | Report as abusive

“If you have lots of social mobility, then the general standard of living is going to go up: you’ll have lots of poor people becoming richer, and you’ll also have the rich protecting their downside, in the likely event that they become poorer, by doing their best to improve the lot of the poor….The answer, of course, is the high-churn society — which means, working backwards, that if you want growth, you also want social mobility.”

You assert this as a statement of fact, but do you have any evidence for it?

Regardless, it misses the point. If you want growth, then *you want growth*. Are you even contesting the proposition that mobility per se has no value in and of itself? For example, given two policy regimes to achieve equal long-term growth rates, do you prefer the one with greater long-term mobility? If so, you haven’t said why. If not, then I think you are essentially in agreement with Cowen and Tabarrok.

Posted by Anonymous | Report as abusive

Wait, Krugman accuses Cowen of being “anti-american” while Cowen praises Krugman as a great economist but as a poor reader (with evidence) and it’s Cowen who is labeled “rude”?

“There was a time when that sort of sentiment would have been considered anti-American. But I guess that was a different country.” — Krugman

“Paul Krugman is a great economist. But of all the people in my RSS feed, in terms of his quality and skill as a reader, he is not in the top 90 percent.” — Cowen

Posted by xgene | Report as abusive

” What they should be doing is looking at two societies, equal in all respects except that one is high-stasis and the other is high-churn, then fast-forwarding to see which one turns out better. The answer, of course, is the high-churn society — which means, working backwards, that if you want growth, you also want social mobility.”

Thank goodness correlation = causation!

Posted by MattJ | Report as abusive

Cowen was rude to Krugman? I don’t think that was a very fair minded characterization of that post. If Cowen was rude, what would you call Krugman every time he talks about someone he disagrees with?

I hate to get dragged into this as much as I have because I think the debate is so academic on both sides as to approach absurdity. Nonetheless, I am honestly puzzled as to how people seemed to be characterizing Cowen’s point.

I read Cowen as saying that he does not value mobility for its own sake but only when it is part of a growth producing activity. Nowhere does he argue that decreasing mobility increases growth. Nowhere does he argue that you can have growth without mobility. In fact, he is on record as arguing that people should be willing to move around more to better their economic condition.

A lot of the gotcha quotes that people are pulling from Cowen are part of a thought experiment to show that mobility in the absence of growth does not accomplish much. They were not arguments to prove that you could have growth with out mobility.

Like anonymous above, I don’t see how your point would run counter to Cowen’s and I am at a loss as to why you jumped on the bandwagon that seems intent on misrepresenting what Cowen was arguing.

Posted by apeman1 | Report as abusive

I suspect that maintaining social mobility requires greater public investment in education. Periods of low social mobility are likely typified by greater private investment.

Plenty of brains in all classes (most people just need to be taught how to use them), so I’m not too worried about the lack of mobility stifling creativity or growth.

But a stratified society is not politically healthy. That alone ought to encourage the public investment to support mobility.

Posted by TFF | Report as abusive


You call Felix on this statement.

” ” If you have lots of social mobility, then the general standard of living is going to go up: you’ll have lots of poor people becoming richer, and you’ll also have the rich protecting their downside, in the likely event that they become poorer, by doing their best to improve the lot of the poor…. The answer, of course, is the high-churn society — which means, working backwards, that if you want growth, you also want social mobility.” ”

” You assert this as a statement of fact, but do you have any evidence for it? ”

Yes. There is ample evidence. Two of the most advanced countries on the planet are the very proof you’re looking for.

Why do you think there are so many American of Nordic ancestry ? 1.5 millions Norwegians and Swedes moved to the US in the late 19th, early 20th century, 20% of the population of those two countries at the time. Abject poverty and sheer starvation in Sweden and Norway was the reason, just like for immigrants from Ireland or Italy.

Unknown to most, those two countries were not always the highly egalitarian, socially stable and prosperous countries we know nowadays. They were actually atrociously unequal all the way past World War I and suffered from widespread and extreme poverty, far worse and far later than nearly every other countries in Europe. Rural areas barely out of feudalism, extremely poor farmers held in debt indenture across generations, very static and repressive social structures overall. Those were not nice places to live in if you were born on the wrong side of society.

It really took the 1930s to finally bring the issue to a boil and get Norway and Sweden on track to their present state, through policies very explicitly designed to bring down inequality and promote social mobility.

Posted by Frwip | Report as abusive


Oh, and by the way, Norway doesn’t owe its very nice and very egalitarian social structure to oil, in case you wonder. Oil production only started in the 1970s. By then, Norway was already the advanced country we all like (except for rakfisk. Bloody weapon of mass destruction, that stuff…).

Posted by Frwip | Report as abusive

@Frwip “Oh, and by the way, Norway doesn’t owe its very nice and very egalitarian social structure to oil,”

No just it’s wealth. Oil and gas exports are 20% of GDP. Norway is a country with the population of greater Boston living on a land mass about the size of texas with thousands of miles of seacoast.

Still though there are lots of South American and African nations which have similar natural endowments that sadly manage to parlay them into squalid poverty for most of their citizens… what makes Norway different?

The distance from the equator. 25,000 years of Norwegian ancestors freezing to death if they didn’t put away enought dry firewood and food for the winter. They have a predanatural instinct to work and save.

Posted by y2kurtus | Report as abusive

@Frwip: while your anecdote about Scandinavian countries’ paths towards growth and greater equality is interesting, it is really just anecdote. As @y2kurtus points out, the population of Norway is very small and its circumstances are unusual. I’d be more curious to see plots of “social mobility” versus growth. This is a very tricky thing to measure, as far as I can tell from a bit of googling. First of all, what metric does one use for social mobility? Note that we are not asking, here, about *inequality* per se, but *mobility*. One measure that seems to be floated is intergenerational income elasticity. If you use that, and are interested in the effect on growth, then you need to analyze growth relative to a potential time lag of a whole generation. How does one control for demographic effects, or exogenous factors (like the start of oil production in Norway in the ’70s, for example)? It strikes me that there will not be any simple story to tell, although perhaps econometricians have already solved this problem and it is addressed somewhere in the economic literature. If there is indeed ample evidence — in the statistical, rather than anecdotal, sense — I really would like to see it, or to learn about references.

Posted by same_anonymous | Report as abusive

If there’s one thing this discussion proves, it’s that economics is an appalling tool for describing moral social outcomes.

Let’s say Cowen and Alex are entirely correct – that from a standpoint of of economic utility maximazation, high-churn and static societies are indistinguishable. Let’s go even fitter and say that Felix is flat out wrong about high social mobility enabling better growth outcomes (although I’m a bit surprised that so many people here are hostile to what is a essentially a “competitive markets produce better outcomes” argument. Not a fan of free markets, eh?)

Most people’s response would be: So what? We should, and do, value social outcomes that aren’t just economic. Most Americans would be appalled by the idea of a hereditary aristocracy, even if it was utility-maximizing. And we’d do so because ideas like morality and fairness have value, too.

That’s my bigger frustration with economics. It is a social science that has somehow co-opted an entire range of social debate. Can Cowen price that a static society is utility maximizing? I doubt it. Would it matter if he did? Absolutely not.

Posted by strawman | Report as abusive

@strawman: Let me respond to each of your points. First, what metric other than utility do you propose to use? If the metric includes social mobility as a constituent, then you are simply asserting rather than proving the value of social mobility. As philosophers would say, you are “begging the question”. Now one potential metric you might suggest is *fairness*. While this is vague, at least we all have some intuitive sense of fairness and perhaps some of us believe that a world with greater social mobility is fairer. This is an entirely plausible story, but nonetheless part of the Cowen-Tabarrok argument is that we can should question our intuitions here. When a poor person becomes wealthy, we are glad for her, but why? Perhaps because her life is now better, or because she must have created some value or accomplished something good in order to change her circumstances. But NOT simply because the absolute value of the difference between her former and present circumstances is high. It was not her mobility per se, but her upward mobility that pleased us, at least to my intuition. In contrast, when Richard E. Rich falls on tough times and runs his family business into the ground, I do not rejoice. Although he is socially mobile when he falls from from the upper income quintiles to the lower ones, this does not please me, except perhaps out of a sense of schadenfreude which I do not take pride in. What’s the point? Even if your intuitions tell you that you value mobility, perhaps you are misinterpreting them, and what you truly value is growth, or the win-win creation of more utility to go around for everyone.

To summarize the above discussion, here is a quote from the original Cowen post linked above: “For a given level of income, if some are moving up others are moving down.” A story about why more mobility is desirable — ceteris paribus — must explain why it is “fair” or “moral” not only for people to move up but for the same number of people to move down. And indeed, phrased thus, this idea seems inherently in tension with some of my own intuitions about fairness. To wit, I don’t think it intrinsically “fair” for a rich person to fall into poverty; the mere fact of his having previously been rich does not alleviate his subsequent suffering. (Indeed, as Cowen mentions, “frame of reference effects” may mean that his previous wealth makes him suffer from poverty more than one who has been poor all along.)

As for your “slippery slope” argument about hereditary aristocracy, this is just sleazy reasoning. By social mobility we are referring, I believe, to the observed level of intergenerational income elasticity. No one is suggesting a return to feudalism.

As an aside, your mention of hereditary aristrocracy does bring to mind the singular policy question that figures in the American political debate and perhaps relates to social mobility: the estate tax. There are all sorts of arguments for why this tax is a good thing. I think an argument based on social mobility is not very convincing (because I think the estate tax probably has little to no effect on social mobility one way or the other).

Finally, let’s discuss your complaint about the use of economic reasoning as a tool for analyzing the morality — or let’s say desirability — of social outcomes. I think you are being silly. No one is suggesting that *only* economics be used for this purpose. But this is an economics-themed blog, which you read, so surely you acknowledge that economics bears upon many questions — even moral questions — about our society and its organization. It strikes me as ludicrous in the extreme, for these purposes, to rule out the use of economic theory (such as building mathematical models, abstracting the roles of individual, idealizing human economic actors as “agents” with prescribed behavioral tendencies, etc… and yes, even abstracting the notion of “value” in some contexts into a measurable quantity called utility). Ignoring these tools would be like fighting a boxing match with your hands tied behind your back. No one complains that quantum mechanics is a physical science that has co-opted an entire range of chemical debate. Rather questions of chemistry lead inexorably to quantum mechanical considerations. In the same way, questions about the morality of our social institutions lead inexorably to certain economic considerations. Just as there is more to chemistry than quantum mechanics (and it would be naive to take an extreme reductionist view that there is no more to it than that), in the same way there is more to politics and ethics than economics. But, to sum up, no you are not free to reject economics as a valid language for describing moral social outcomes. I reject your rejection.

Posted by same_anonymous | Report as abusive

y2k, we’ve had this discussion before. If Alaska didn’t have oil, it would be just like the southern U.S. states – dependent on tax dollars flowing into them. What about Mongolia, Kazahkstan, Poland, Belarus? They’re pretty far north of the equator and remain relatively poor.

If we get rid of upward mobility, what excuse for spending and tax cuts will the Republicans use to replace the “anyone can make it big in America” dream?

Posted by KenG_CA | Report as abusive

Funny how some people interpret “social mobility” as the ability to move house to a different geographic location when in reality it is the ability to better oneself and move up the social ladder. Any country that has systemic bias towards keeping one group of people in control of the levers of power will not make the best use of its human capital.

A typical example is the layer of jobs that require a full University degree as the entry point; these jobs are in effect unavailable to anyone who cannot afford to go to University, no matter how clever they are or how good a leader or potential employee at that level they would be.

This shuts them out of high paying jobs and reduces the likelihood their children go to University. The rich continue to send their kids to the best Universities so they get the best jobs, and as we have seen from many financial and political messes, being rich doesn’t make you clever, and being clever won’t make you rich.

Posted by FifthDecade | Report as abusive

I note that the specific mobility mentioned is intergenerational mobility. On what basis would one expect that the relative income of different individuals is makes for sensible indicator of anything apart from what it precisely measures? I suspect that I am in a higher income decile than my father was decades ago at my age. But maybe I’m not. I never think about it except when I read such articles and suspect that almost no one else does, either. I can attest that the topic of intergenerational income positioning has never, ever come up in a conversation with anyone I have ever met. If few people are even aware of where they individually end up in such a metric why would we think the aggregation of individually unimportant data points influences people?

Posted by Eric377 | Report as abusive

@y2kurtus & same_anonymous

Bloody amazing! You ask for evidence for Felix’s proposition and I give you the closest thing to a control experiment there ever was. And then you go on how it’s irrelevant, not applicable, a special case, etc.

@y2kurtus (specifically)

No, you miss the point so completely, it becomes quite spectacular.

I knew the objection of oil and gas would come, regarding Norway.

But Norway’s transition to a highly egalitarian, socially mobile society happened BEFORE oil and gas became an important part of the economy. And that was my point, precisely.

And guess what? It gets even better than that! Not only you get the closest thing to a controlled policy experiment if there ever was one, but you actually get two of them wrapped in one. You actually get to test a counter-factual, with a strong result that it’s not resources or similar accidents of national luck that matters but which policies are applied.

Sweden next door made exactly the same transition in the same time frame using essentially the same types of policies and, incredibly enough, achieved essentially the same outcome. And even double plus good better, Sweden has even managed to sustain this outcome since then, just like Norway, and still no oil and gas (save for a bit of shale gas).

Posted by Frwip | Report as abusive

Hi @frwip:

I don’t think your account of Norway and Sweden is a controlled experiment. Was there a specific date you can peg as a paradigm shift in their social organization, as regards equality? (Note that we are actually talking about social mobility, not equality. I’m not sure how you hope to separate that issue, but that is an entirely different point.)

If you look at figure 2 here: http://eh.net/encyclopedia/article/grytt en.norway
you’ll see their GDP has grown in close to a straight line (on a log scale) from 1830 to present day. The annualized growth rates in Table 1 at that link do indicate that growth has been a bit faster since World War II. But how can we tell whether increased mobility and/or egalitarianism is the dominant effect, as opposed to e.g. oil (after the 1970s), or increased European economic integration (since WWII)? Basically, your experiment *lacks* a control. If Sweden and Norway were identical in all respects, but one transitioned to a more egalitarian society and the other didn’t, then perhaps that would be a controlled experiment. But you suggest that both Sweden and Norway transitioned in the same way at the same time, so they don’t help to isolate the mobility effect, if there is one. Figures like the ones here: http://sun-bin.blogspot.com/2005/12/when -will-chinas-gdp-overtake-us.html, show that various economies have grown since 1960 at various rates relative to the United States. If one had comparable data on social mobility for all these nations during the last 50 years, then perhaps one could start to address the question of providing convincing empirical evidence. (I suspect some economic historians could provide the relevant data, but I can’t find it online easily.) But just saying “Sweden and Norway once sucked for the poor, now they suck less, and oh look their economies have grown” is not evidence. For example, this paper: http://people.anu.edu.au/andrew.leigh/pd f/ChinaIGE.pdf, finds that China has “strikingly low levels of intergenerational mobility”. But obviously the extraordinary economic growth in China in the last 20+ years has been one of the greatest boons to human welfare in world history. Does that count as evidence that “low mobility leads to growth”? Of course not; that would be an absurd conclusion.

Like most forms of scientific inquiry, economics is hard. That’s why I think it is important to call people (like Felix Salmon) out when they are blase about asserting causal relationships (e.g. between mobility and growth) without explaining the evidence.

Posted by same_anonymous | Report as abusive


There’s a lot of meat to yor reply, much I which I will have to beg off answering, for the sake of time. My apologies! I do not enjoy talking last another person in a debate – I don’t mean to do it here

However, I did want to touch on my “sleazy” reasoning about a hereditary aristocracy (I have never reasoned sleazily before – how titillating!). I’d refer you back to the examples Tabarrok provided in his post. In an example of a static society with an inter-generational income elasticity of zero, you would get: Total class stratification.

Now, I don’t think that was his point. Nor do I think it reflects reality. Nor do I think economics is useless as a tool for describing social outcomes – rather, that it can and should be supplemented with other considerations.

But here’s what I do think – Cowen is focusing on only one aspect of mobility – its utility compared to a hypothetical, static society. But if we want our theorizing to have real world consequence, it’s fair to ask what would a static society look like. What makes a society less mobile than others? And is that a good thing.

In my opinion, more static societies feature less access to public goods, less access to general education, or more capture of social institutions by the elites, all of which reinforces income levels across generations. That doesn’t have to be the right answer! Perhaps immobile societies are simply ruled over by genetic supermen. I’m happy to listen.

To put it simply: I feel a mobile society is important because it is a proxy for the ability of individuals to improve their circumstances. What immobile society can you describe which is entirely consistent with the tenants of market capitalism?

Posted by strawman | Report as abusive

@strawman Is that “tents of market capitalism” as in its basic principles, or “tenants of market capitalism” as in those who in the context of this discussion do not participate as capital owners, but as beneficiaries for the price of remaining so?

Posted by FifthDecade | Report as abusive

Of course I meant “tenets”.

Posted by FifthDecade | Report as abusive