Inequality, Heritability and the American Dream

Ever since people finished identifying “the American Dream” — the idea that in the US in particular and the New World in general somehow allowed people to escape the hidebound social structures of the Old World and better themselves via their own efforts — people have been worried that it is on the point of dying. Americans continue to show an an unusual degree of belief in the ability those who work hard to better themselves by their own efforts. For instance, in the 1999 International Social Survey, 61% of Americans agreed that “people get rewarded for their effort”, whereas only 41% of Japanese agreed, 33% of British and 23% of French. This belief has actually increased in recent decades. In 2005 the New York Times reported that while in 1983 only about 60% Americans agreed that “It is possible to start out poor, work hard and become rich” by 2005 nearly 80% of Americans agreed with that statement.

And yet, those who study inter-generational income mobility have been increasingly worried in recent decades that despite American’s belief that people can work hard and get ahead, that it is becoming increasingly difficult for people to actually achieve this in the US. In a lengthy report by the liberal think thank Center for American Progress, Tom Hertz of American university brings together a number of the recent studies on intergenerational income mobility in the US as compared to other countries, showing how people who are born into the lower income quartiles in the United States are less likely to reach the top levels of income than in other countries such as Germany, Sweden or Denmark.

To give an idea of what is meant by this intergenerational mobility, it helps to look at a particular study which Hertz quotes in detail. In this study, researches tracked 4000 children originally surveyed in 1968 and compared what the household incomes of their parents were in the 1967 to 1971 period to what those children’s incomes were in 1994 to 2000. The intergenerational correlation in family income between parents and children was .42, and the implications of that for the children themselves are shown in the this chart:

Family incomes are inflation adjusted.  On the left column you see quintiles of parental household income in the original 1967 to 1971 window.  In the column headers you see the quintiles of household income for the children.  (The numbers are higher despite inflation adjustment because the group as a whole was better off in the late 90s than in the late 60s.  On average, people within the bottom 20% had higher incomes in the 90s than people in the bottom 20% in the 60s.)

In regards to intergenerational mobility, you can see that of the children of parents in the bottom income quintile in the late 60s, 41% of those children wound up in the bottom income quintile themselves in the late 90s.  24% made it into the second quintile, 15.5% into the third, etc.  Only 6% made it into the top quintile.  Of those born to parents in the top income quintile in the late 60s, 42% were themselves in the top income quintile in the 90s, while only 6% were in the very bottom quintile.

Now, it seems to me that a lot of the question as to whether the American Dream still holds true relies on to what extent we can assume that there is an equal distribution of ability and effort among all children across all income ranges in a study such as this. Yet, when we start to look at this, we (particularly because as Americans we have a great attachment to a variety of ideas relation to the American Dream) run into all sorts of contradictory emotions.

For example, let’s take two typically American Dream statements:
1) If you work hard and save, you can work your way up and become rich.
2) If you get your kids a good education and teach them how to work hard, they will do as well as or better than you.

Let’s assume for the sake of argument that both of these are true, and look at what happens over two generations. In generation A, people work their way up and become rich to the extent that they work hard and save. They have children: generation B. Now, generation A does a great jobs of getting generation B a quality education and teaching them to work hard, with the result that the children of B who work hard and save all do as well as or better than their parents. A few children of people in A whose parents did not work hard or save also work hard and save, and they become rich too, so we see some movement upwards movement from the lower earning families in generation A — but given that the middle and upper earning families of A did such a great job of teaching their kids and giving them a solid work ethic, we have the appearance of very little social mobility — because people are doing a very good job of teaching their children the skills that allowed them to achieve their current space on the income ladder, and so the only room for movement is if some people who themselves did not work hard or save manage to teach their children to do differently.

So, if the kind of abilities and behaviors that result in doing well economically are heritable or teachable, then after the first generation we would probably expect to see less intergenerational income mobility. If your abilities and work ethic are fairly similar to your parents, and if your parents economic success was determined by the extent to which they worked hard and saved, then in all probability your success will be a lot like your parents’.

On the other hand, if there is a great deal of chance involved in how one’s ability to work hard and save translates into household income, then one would actually expect more intergenerational variability in income. If your parents worked very hard, saved, etc. but through bad luck or lack of opportunity made very little, and yet they taught you to also work hard and save, then if you experienced better luck in translating your hard work and saving into higher income, you would do significantly better than your parents.

Similarly, if until recently people’s economic success in a given country did not reflect their efforts, but then something changed so that in future greater effort resulted in greater success, one would expect to see a period with a lot of intergenerational income mobility, and then a settling out.

Of course, all this is working off the assumption that the traits which might result in higher earnings are heritable. Some characteristics such as measured IQ appear to be quite heritable. One’s adult IQ has a .75 correlation to the average of one’s mother’s and father’s IQs — a significantly stronger correlation than the one between one’s income and one’s parents’ in the US. But other determining factors in how much one makes (willingness to work hard, the amount one is interested in making more, the type of career one is interested in, etc.) are much harder to quantify and may quite possibly be less heritable.

All of which is to say: It seems to me that discussing whether or not Americans are right in believing that “anyone can succeed in America” with hard work and ability is much more difficult than simply looking to see how often people whose parents were in the bottom income quintile end up in the top income quintile. Moreover, the fact that one country has higher intergenerational income mobility in recent years does not necessarily mean that it is more of an opportunity society than the United States (though that’s one of the possible meanings of that statistic), especially if the two countries have significantly different histories in recent decades.

Share With Friends


Now an Ohio Catholic!


  1. “This may reflect the effects of discrimination in the labor market, but may also result from factors such as the difference in the quality of schooling acquired by blacks and whites. Note also that while we have controlled for a long list of parental personality variables, we have not been able to
    control for that same list among the children. It is thus possible that given ostensibly comparable family backgrounds, African American and white children develop different attitudes towards economic success that are then reflected in their family incomes.”

    This seemed interesting given that mobility for Latinos was not significantly affected by their race. (It would also have been interesting if Asians were included.) One might hypothesize that schools are generally poor for Latinos as well as African-Americans. If that is the case, then increasing quality of education would not be as useful as improving “attitudes towards economic success.”

    This conclusion is supported by this which shows that increasing expenditure on schools does not necessarily improve outcome:


  2. As it turns out, heritability of IQ can explain only about 5% of the correlation between parent/child income (in other words, controlling for heritability of IQ reduces the parent/child income correlation from .42 to .4). The reason for this is that while IQ does appear to be substantially heritable, the correlation between IQ and income is fairly weak (.27).

    Granted, IQ is only one trait which is both potentially heritable and correlated with income. Things like willingness to work hard, for example, may be partly heritable. However, these traits are presumably also present in other countries where the parent/child income correlation is a lot lower.

  3. Blackadder,

    GNXP seems to be down at the moment, so I can’t follow your link, though I did find this someone interesting EconLog post linking to the same GeneExpression post:


    I don’t have a particular dog in the IQ fight — I have an inherent dislike for the IQ concept, have never had mine measured, etc. That said, it seems to me that we have a couple of pretty basic conceptual issues to sort out here:

    – People in the US tend to want to believe that “anyone” can succeed, and so that success correlates, to a great extent, with the extent to which people have striven to “get ahead”.

    – Additionally, people tend to want to believe that through some constellation of factors (genes, education, instilling “values”, etc.) they can prepare their children to have success equal to or greater than their own.

    Now, either one or both of these are totally false, or else one would expect, after a certain period, a country to become fairly “sorted” and to find that many people achieve success moderately similar to their parents — not necessarily because there are nefarious forces keeping their parents from succeeding, but because their parents success is a measure of their parents effort and their parents have been moderately successful in passing on the characteristics (whatever they are) that allowed their own success.

    Certainly, you’re right that the degree to which humans tend to inherit the characteristics of their parents is likely to be consistent across countries, and there is a good deal of variation across countries in regards to the degree to which parental income correlates with child income. However, that could potentially be the result either of the country being less sorted (and thus people’s parents abilities having less correlation to their incomes than is the case in the US) or to some difference in cultural attitudes and drives.

    The perception that success is primarily the result of chance is much more prevalent in some of those countries with lower correlations of parent to child income — and the lower correlation of parent to child income could, depending on what assumptions one makes about the heritability of whatever traits it is that result in success could in fact be the result of greater randomness rather than greater correlation of ability to reward.

  4. As I think about it further, there are really three beliefs which, like many Americans and classical liberals, I find myself strongly attached to:

    1) People are all fairly equal — there is not some “peasant class” which is inherently fit only for inhabiting the bottom levels of society.
    2) Hard work, saving, etc. will result in “getting ahead”.
    3) By bringing your kids up right and teaching them well, you can make it pretty likely that they’ll all do as well or better than you.

    The thing is, these three beliefs are not really compatible. If it’s true that via effort and ability one can “get ahead”, and if by bringing your kids up well you can assure that most of them will do the same or better than you, then necessarily people are not all that equal (as shown by the fact that not everyone “gets ahead”.

    As a result, people necessarily end up de-emphasizing at least one of these three in order to try to make some sense of the thing. (Or else mindlessly asserting all three without thinking about the contradictions.)

  5. Darwin, your three beliefs are logically incompatible only if you assume all parents are bringing up their children right and teaching them well, and that is demonstrably false. I am reasonably confident in the accuracy of your first belief as long as “fairly” is underscored. Most people with multiple children observe remarkable aptitude variances notwithstanding the same gene pool. The second is most certainly true as a generality. The last belief is problematic. Children of achievers seldom equal or surpass the achievements of their parents. There are several reasons for this, which are reasonably obvious upon modest reflection. Nonethless, like most Americans achievers share (or at least want to share) your last belief, but the odds are they will be disappointed even assuming sound child-rearing. But it is certainly true that good parenting produces better results than poor parenting, but that is a different point.

    In the end, the real debate is over the role of luck. Even assuming away dogmatic biological materialism, liberals tend to over-estimate its importance and conservatives tend to under-estimate it. It is a very important factor (indeed being born into a loving family and with a brain wired for aptitudes the market values is luck), but ordering a society around the assumption that luck is the dominant if not dispositive factor (which at bottom is what many liberals would like to do) is extremely pernicious in that it will fail to reward prudent behavior and punish imprudent behavior thereby making society much worse off for all.

  6. I agree with your three points, Darwin. Point number 2 is the funny one. Some people will tend to balk at that statement, reading into it a cold and (dare I say) Calvinist worldview. As Mike pointed out nobody denies the presense and influence of “luck”, it’s a matter of to what degree it plays a part. Clearly people of our mind allow for good fortune and misfortune, but the exception prove the rule.

    As a test we can ask the question of the naysayers: how far do you suppose someone will get in life if they don’t work hard or make no effort to save or at least limit discretionary spending? The answer to me is self evident and I suppose it would have to be to them as well. I also think you can rework point 3 that way too. I doubt you would get many people to disagree with point 1, but unfortunately I think there are a number of people who subscribe to it in practice. Oddly enough they’re likely to be the same people who have no use for points 2 and 3.

Comments are closed.