Inequality, mobility, opportunity

Alan Krueger, Chair of President Obama’s Council of Economic Advisers, gave a talk a few weeks ago on inequality. Krueger described the sharp increase in income inequality in the United States since the 1970s and discussed some undesirable consequences it may have. One of those consequences is reduced intergenerational mobility (relative intergenerational income mobility, to be more precise). Krueger provided a graph showing that nations with greater income inequality tend to have a stronger correlation between the earnings of parents and their children (less mobility). This has sparked a wide-ranging discussion about the link between income inequality and mobility (Winship, Corak, Winship, Corak, Cowen, Yglesias, Wolfers, Smith, Winship, Quiggin, Cowen, Quiggin, Bernstein).

Here’s what the pattern looks like according to Miles Corak, the source of Krueger’s data (enlarged version here). The vertical axis in the chart is immobility; lower means more mobility. The horizontal axis is income inequality a few decades earlier.

Nations with lower income inequality tend to have more intergenerational mobility, and the association is quite strong. There are concerns about the data. But suppose the data are accurate, and suitable for testing this link. What does the association depicted in this chart tell us about the magnitude of inequality’s impact? How much would reducing income inequality in the United States help?

For most, the aim isn’t high intergenerational mobility per se; it’s low inequality of opportunity. Mobility serves as an indicator (not perfect, but not bad) of equality of opportunity.

Money ought to be good for children’s opportunity. Kids growing up in households with higher incomes are more likely to have good health care, low stress, learning-centered preschools, good elementary and secondary schools, extracurricular activities that promote cognitive skills and earnings-enhancing noncognitive traits, and access to a strong university. It would be surprising, therefore, if inequality of parents’ incomes did not contribute to inequality of opportunity among their children.

But how large is the effect? After all, money isn’t the only thing that matters; a good bit of our abilities and motivations when we reach adulthood stem from nonmonetary influences such as genetics, in-utero developments, our parents’ habits and behaviors, and peers. Also, there are diminishing returns to money; beyond a certain point, more parental income probably helps only a little, if at all.

Yet the graph suggests a large impact. Apart from data concerns, is there any reason to question this? I think so. First, let’s set aside the low- and middle-income countries (Argentina, Brazil, Chile, China, Pakistan, Peru, Singapore). Mobility processes in these countries may or may not be comparable to those in the rich nations. Next, notice that inhabiting the lower-left corner of the chart are the four Nordic nations: Denmark, Finland, Norway, and Sweden. These countries have been providing affordable high-quality early education to a substantial portion of children age 1 to 5 for roughly a generation. James Heckman and Gøsta Esping-Andersen, among others, have argued that early education is perhaps the single most valuable thing a society can do to equalize opportunity. These countries also feature late tracking in elementary and secondary schools and heavy subsidies to ensure college is affordable for all. These public services, rather than low income inequality, might be the chief reason the Nordic countries have such high intergenerational mobility.

What would that imply for the cross-country association between income inequality and intergenerational mobility? As the following chart shows, if we leave out the Nordic nations the association is still there, but the countries are widely dispersed around the line, suggesting weaker grounds for confidence that the association is a strong one. (For the statistically inclined, the r-squared is .47 with the Nordics included and .16 without them.)

Is it possible, then, for a country to have high income inequality but also low inequality of opportunity? John Quiggin is skeptical. He suggests the UK experience has debunked this “third way” notion. I’m not so sure. Imagine a rich nation with America’s income inequality and Nordic public services: affordable high-quality early education, K-12 schooling with late tracking and equal funding, and widespread access to good-quality universities. And perhaps also comprehensive prenatal care. Would its opportunity (mobility) structure look more like America’s or more like Sweden’s?

But, some will respond, you can’t get those services if income inequality is high. The rich will block the heavy taxation needed to fund them. Maybe. But income inequality has been rising in Sweden. In fact, in the late 1990s and mid 2000s the top 1%’s share of income (including capital gains) in Sweden was about the same as in the 1970s United States (see figure 7 in this paper by Atkinson, Piketty, and Saez). So far this hasn’t undermined Swedish taxation, though it’s probably too soon to draw any firm conclusions.

Suppose income inequality continues to rise in Sweden but its public services hold up. Will Sweden’s intergenerational mobility a few decades from now look like ours does today? I’d predict no. I suspect opportunity-enhancing programs can overcome a good bit of the harm done by income inequality.

That’s not to say we shouldn’t also try to reduce income inequality. I think we should. The point is that if we want to reduce inequality of opportunity, reducing income inequality isn’t the only way, and perhaps not even the best way, to do it.

9 thoughts on “Inequality, mobility, opportunity

  1. Interesting post.

    I think that you could present a cogent and politically viable argument that increasing taxation on the super-rich, and using that to fund better opportunity creating measures, would work on both parts of the problem.

  2. Forget about those “immobility data”. The studies are never really comparable, and most the times plagued by fundamental problems (you have to wait something like 50 years to get reasonable results, who can wait that long?)
    Son’s income at 32 is not a good proxy of what he later really earns and owns (like after getting the heritage) (Corak 1999). Look at Corak “chasing the same dream”, 2009, Figure 2 and 3, the differences between Canada and the US are just the statistical noise of the Mazumder 2005 US study, even for the most extreme cases. And it suffers from top coding etc.

  3. You could also ask whether, if we want to reduce inequality of outcomes, increased mobility is the only (or the best) way. That is sort of the mirror image of your question (whether more equal outcomes is the best solution to low mobility). In your Swedish example, a society with high inequality and simultaneously high mobility would still present problems. A large distance between the top and the bottom means there are still a lot of poor people, and without separate action to improve their lot (ie to reduce the distance between the bottom and the top) you could have a society which might be mobile but which might not be sensibly labelled successful.

  4. Being Swedish having insight into the Swedish statistics;
    intergenerational mobility in Sweden has dropped dramatically during the last decade, during which income inequality has increased. The pattern becomes clearer in the link between children’s educational achievements and family income.
    (Privatization of public schools has lead to increased segregation and this is according to the PISA study one reason behind the sharp drop in international educational ranking that Sweden has seen the last 15 years).
    Can a high tax society be maintained in this environment? It looks like it can’t. Since 2006, Sweden has had a conservative government (if one can call neoliberal for conservative, that is questionable), which has drastically reduced the taxation for mainly upper middleclass incomes. This has been financed by cutting sick-leave insurance benefits and unemployment benefits. This has resulted in a decline or in the incomes of the poorest 10% resulting in a sharp increase in child poverty.
    The development has partly been counteracted by a rapid increase of private debt, which has grown 2-3 times as fast as the economy, and is now amongst the highest in Europe. Obviously this is not a sustainable development, as we have seen in other countries.
    So, the outspoken aim of the conservative government, that Sweden should be more like other countries, is a wish that has been granted. From the Swedish example it seems that inequality is a process that drives itself and without the constant efforts to reverse it, we will sooner or later face a situation of utter deprivation of the majority, with low aggregate demand as a consequence.

  5. Lots of good thinking. Thanks. A bit of a chicken-and-egg problem, of course.

    I’d be curious to see similar analysis and thinking that looks at wealth inequality, which of course if far more profound.

  6. @Steve and Lane,
    if you can not even get income (mobility) data straight,
    producing garbage results like this Corak, forget about wealth data.
    We have some fairly good idea about income distribution,
    but much less so about wealth.

    And then “measuring” the mobility of it, at least one order of magnitude more difficult.

    Tonight I am in not such a good mood, having to realize that another US economics nobel prize winner, I held in regard until a few hours ago, was just faking data to support his garbage analysis.

    Economics is just > 95 % garbage in, garbage out.
    Not science.

  7. You mention that income inequality in Sweden has remained the same as the US. However, it is worth considering whether services provided by the government should be considered separate from income in the first place (i.e., the income of the Swedes should include the market value of the services they receive from the government).

    I would imagine that for the purposes of this argument, they should be considered separately as you have, but I dont think it is necessarily a foregone conclusion.

  8. It comes down to measurement problems that stem from logical flaws in the neoclassical consumer theory. Incomes aren’t adjusted for quality, variety and convenience. You need a framework that formalizes these things before you can measure them. It can’t be done with indifference curves, as you’ll see in this video by Alex Gheg. This former PhD econ student discovered the dirty little secret of utility functions.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s