Upward mobility in the United States is not declining as many citizens think

Here is the new Raj Chetty paper that everyone is talking about (pdf);

We use administrative records on the incomes of more than 40 million children and their parents to describe three features of intergenerational mobility in the United States. First, we characterize the joint distribution of parent and child income at the national level. The conditional expectation of child income given parent income is linear in percentile ranks. On average, a 10 percentile increase in parent income is associated with a 3.4 percentile increase in a child’s income. Second, intergenerational mobility varies substantially across areas within the U.S. For example, the probability that a child reaches the top quintile of the national income distribution starting from a family in the bottom quintile is 4.4% in Charlotte but 12.9% in San Jose. Third, we explore the factors correlated with upward mobility. High mobility areas have (1) less residential segregation, (2) less income inequality, (3) better primary schools, (4) greater social capital, and (5) greater family stability. While our descriptive analysis does not identify the causal mechanisms that determine upward mobility, the new publicly available statistics on intergenerational mobility by area developed here can facilitate future research on such mechanisms.

Here is summary coverage from David Leonhardt.  The highly reliable David starts with this: “The odds of moving up — or down — the income ladder in the United States have not changed appreciably in the last 20 years, according to a large new academic study that contradicts politicians in both parties who have claimed that income mobility is falling.”

Confusing issues of equality and mobility remains rife in current discourse.


That link to http://www.equality-of-opportunity.org/ is fascinating - from its FAQ -

'While we do not focus on comparisons across countries in our research, prior work [Corak 2013] has found that the U.S. has significantly lower levels of social mobility than most other developed countries.'

Upward mobility is not declining, but is downward mobility not accelerating? this is not left vs right, GOP vs Dems, Socialism vs liberty. This is war against White people.

Why do hostile globalist elite defend Israel as a Jewish ethnostate with Jewish only immigration, but ravage White majority Europe/North America into a multi-ethnic, multi-cultural Gulag with non-White colonization?

The world is 93% non-White, only 7% White. But 3rd world colonizers, Muslims, Sikhs, Hispanics, are aggressively advancing their agenda to annihilate gullible Whites, just as China annihilates Tibet.

How long will gullible Whites cuckold for murderous anti-White elite, who suppress our fertility, confiscate our guns, infiltrate/subvert our banks/FBI/CIA, indoctrinate White kids in academia/mass media, plunder White jobs/wages, & butcher White soldiers in bankrupting wars?

"Native" Americans invaded from East Asia. Yellow & Brown races committed 10-times more genocide, slavery, imperialism than Whites. Since Old-Testament, Whites have been victims of Jewish/Crypto-Jewish, Turkic, Muslim, N.African imperialism, slavery, genocide.

Gullible Whites should reject subversive ideologies- libertarianism, feminism, liberalism- & reject hostile slanders of racism. Peace to all humanity, but White people must organize to advance their interests, their fertility, their homelands. Spread this message. Reading list: goo.gl/iB777 , goo.gl/htyeq , amazon.com/dp/0759672229 , amazon.com/dp/1410792617

China isn't annihilating Tibet by your criteria as the native Tibetan population has grown.

One of the great benefits of being poor is that you are at much lower risk of downward mobility. Seems like a first world problem.

I have long maintained that the law of diminishing returns is one of those things that are very obvious but apparently very difficult for the majority of people to fully understand.

Yup. Why I married a person of Asian descent.

Jews, Turks, and Arabs are Caucasian. Native Americans are from Siberia and Beringia not East Asia. Contemporary Europeans are actually mainly descended from invaders from the Middle East and Central Asia.


Grant comments like this no quarter and give them no voice. Please delete. It's possible to have thoughtful crazy presented thoughtfully. But your comments section shouldn't be a platform for this stuff. (What tipped the scales for me were the links. I won't follow them at work. But that can actually do damage if the wrong person follows the wrong link trail). And you have more intellectually diverse readership than you may realize.

ADL/SPLC thugs like you are the greatest enemies of Ist Amendment. Is truth anti-semitic? Are facts anti-semitic?

Yo, you goddamn supremacists, you've fallen off of consensus reality, your carefully planned "stealthy"/"acceptable" talking points are ridiculously obvious and cliche.

1 Stay on message; we deal with the genocide of White people and the perpetrators, anti-Whites. When talking to the general public don't go into a rant about Jewish conspiracies, banking families, NWO, etc. White genocide is the first step to establish a beachhead in the general population's consciousness -- we can expand on other ideas later.
2 Use the Mantra and our talking points. Everything anti-Whites say always leads to our genocide, and we have good ways to counter their talking points. You can customise our stuff and use it for yourself.
3 Use our terminology; "Anti-White", "Pro-White", "White genocide", "Anti-racist is a codeword for anti-White", "Africa for the Africans, Asia for the Asians, White countries for everybody".
4 "Anti-racist is a codeword for anti-White" should be the last thing in your post if possible.
5 Don't use their terminology; "racist" should ONLY be used in "anti-racist"; "nazi" should ONLY be used in naziwhowantstokillsixmillionjews"; "HATE!", "holocaust", "facist" "LOVE!" should be avoided; "diversity", "multicultural" can be pointed out as meaning less White people. Avoid using "The White race", "Caucasians", "The Whites", because it makes it look distant and turns it into a philosophical debate; use 'Our people', 'My people', 'White people', etc.
6 Capitalize "White", when you're talking about our race.
7 ALWAYS use "Genocide", NEVER use "Extinction", "Population Displacement", "Demographics shift", etc.
8 NEVER call them "antis", they are "anti-White" - they are against White people existing.
9 Use "As an anti-White" e.g. "As an anti-White why do you think more assimilation is required in ONLY White countries?".
10 Don't argue, make our point: ALL White countries & ONLY White countries are told by anti-Whites, who claim to be "anti-racist", that they must accept millions of non-Whites and 'assimilate' with them, which is genocide under UN genocide conventions. If people get annoyed of hearing our message that's good, it means they can remember it!
11 This isn't about anti-Whites; we are talking to the general public, USE the anti-Whites to humilate them and point out that they support White genocide. A lot of the general public are tired of "anti-racists" (anti-Whites) and WE have to show them how to defeat and humiliate them.
12 Be aggressive and take the moral high-ground; if they call you names say "You're only saying that because I'm White".
13 Think before you talk. Talking about commiting violence is a no-go. No talking about genocide tribunals because it makes us look dangerous. Using lots of bad language is going to make you - and us - look stupid. That does not mean treat anti-Whites with dignity.
14 YOU ask the questions. If you don't get a reply keep asking them - it means your opponent is embarrassed and is trying to get off the subject. If they demand you answer questions say "WE ask the questions, because YOU support/justify genocide of MY people!".
15 Use emotional language; "Why do you hate little White children?", "Why do you want to genocide little White babies?".
16 Winning isn't important; you should be winning all of the arguments with the anti-Whites if you do what Bugsters do, but don't worry if you lose a few of them -- this is about imposing our terminology and getting the public talking about White genocide.
17 Know your target audience and wrap our message around what appeals to them. Young adults want to talk about education fees and housing prices, Teens want to talk about music, Old people want to talk about pensions, etc. Try to connect with the people using their language/colloquialisms e.g. "Dude" "Mate" "Bloke" "Howdy" "Yeah" "Lol" "Rofl"
18 Flip the script. What if ALL Asian countries were told to bring in immigrants and assimilate with them? What if "anti-racists" in Africa demanded that Africans stopped patrolling their borders and kicking out illegal immigrants?
19 We're all part of the team. Remember to report any places that you post The Mantra or mini-Mantras in Where did you post the Mantra today? II Bugsters will offer you advice if you're new, ask them for help Swarming and they'll arrive on the link to help you out.

Dumbass gangbanger thugs and welfare queens are generally smarter than you. At least learn to be an interesting racist like Sailer.

Who's is killing off the white people? Where is this genocide going on ? I think that you may be living in an alternate reality! Maybe your post is twilight zone style proof of the multiverse theories of reality !

How interesting. I wonder what other differences there are ? Does your America have a black President like ours ?

good points. completely off topic.

But US beats Germany! USA! USA! USA!

It does - and in Germany, this is a subject of discussion and debate, something that is considered critical to change if Germany is to remain a prosperous democracy. The entire political spectrum sees this as a growing problem, much like German wages and anemic economic growth was seen as a problem in the later 1990s, back when Germany was seen as a basket case. One can see the result of that political discussion in current economic performance - or just think that instead of recognizing and remedying a problem, the German economy just got lucky.

Here is an English language taste of how the current discussion about social mobility is proceeding - http://www.fairobserver.com/article/social-mobility-opportunities-particularly-poor-germany

And here are some perspectives highlighted in that article.

From the Christian conservative party - 'Gröhe went on to argue that this was why "we are expanding early years education by offering all-day nurseries and schools, especially in deprived areas. At the same time, we want an education system that differentiates between pupils, but also allows movement between the various tiers."'

Here is what the Green party head says in response - 'Green party leaderCem Özdemir strongly criticised this standpoint: "The planned childcare allowance and a bureaucratic educational package are precisely the wrong responses. Instead, the funds should be put straight into the quantitative and qualitative expansion of nursery provision and all-day schools, in which young people from disadvantaged backgrounds in particular can receive the support they need in a socially and culturally stimulating environment."

Here is the perspective of the generally pro-business party - 'The general secretary designate of the liberal FDP party, Patrick Döring, echoed the remarks of his CDU coalition colleague in stressing the importance of jobs. "Good, strong growth," commented Döring, created "the best opportunities for work and upward mobility."'

The Social Democrats had this view - 'However, Andrea Nahles, general secretary of the centre-left SPD party, sees a pronounced need for action to be taken in the economy and employment. Speaking in advance of the debate, she again called for greater "regulation of the labour market with a boost to permanent, full-time, and secure employment while reducing the number of people on temporary, part-time, and insecure contracts."

I'll skip the Left Party, as they are boringly redistributionist, to note this is the conclusion of the link - 'Almost all party representatives agreed that education for the disadvantaged merited targeted support to boost social mobility opportunities in Germany.'

The interesting thing is that this was also the sort of consensus creation that occurred in the last decade concerning poor PISA results. And after years of effort, Germany has improved its rankings - essentially because of the emphasis placed on better educating students with lower levels of achievement.

Thank you for the response. I suppose we will see if the approach with demonstrated effectiveness (rewarding individual merit, loosening economic controls) wins out over feel-good approaches with no track record of success (early-childhood education, more money for schools). Similar to the debate here.

My parents have moved from bottom 10% to top 10% (maybe top 1% for a year or two) and everywhere in between during their life. Myself I've moved all over the place, happy to be in the middle class now but in 10 years I could go up or down depending on the roll of the die (fate does not care for a persons plan.) I'm sure though that the opportunities for those incapable (through lack of will or ability) of running a business or learning a high skill trade are less than they were 20 or 30 years ago.

The fact is a lot of our people in the west are no more capable than a Chinese laborer or increasingly robots/machines, therefore they will not be able to command the same wages. Due to globalization and free trade our capitalist classes have the opportunity to run worldwide rather than national corporations. They also benefit from the cheaper labor. I think that this is the cause of inequality.

What can be done to fix it? Protectionist policies maybe? Not politically popular and will raise prices on the poor. Maybe if we wait long enough the Chinese labor costs will reach parity with our labor costs and the western laborer will rise again as a result? This seems to be happening to an extent.

also... I question if anything *should* be done to fix it.

I thought the argument made by Krugman was that mobility was greater during the post-war period (my parent's generation). If this study only focuses on the "last 20 years" it won't really show that decline.

Selecting your dates carefully is an important part of getting the results one desires. Just ask Reinhart and Rogoff.

Prior, you may not like this, but we're going to have to bomb the sh*t out of Germany again if we're going to see thos post WW2 production increases again.

I just love when 'progressives' hearken for the good old days.

In a very cherry-picked way, of course.

Someone can like a particular aspect of American society at a certain period in time without wanting to completely return to the past. What an idiotic comment.

Thank you for providing the definition of cherry-picked, Jan.

Don't be so hard on yourself. You're comments are less idiotic than you think.

"High mobility areas have (1) less residential segregation, (2) less income inequality, (3) better primary schools, (4) greater social capital, and (5) greater family stability."

Do they often serve beans at dinner?

There's also the conflation of realized economic mobility and potential economic mobility.

Given the large cognitive variance in the U.S., assortative mating, and the heritability of cognitive traits, this is a crucial distinction.

Word. I guess Tyler has been trying to make a political point in his last couple posts, which of course is fine (although in my opinion somewhat misguided), but I'm surprised at the lack of interest in defining where mobility and inequality *should* be, at least from an economic perspective.

Upward mobility is a reality but more for some groups than others.

Also, different groups are more upwardly mobile in certain areas than in others.

Blacks are more mobile in sports, pop music, and entertainment.

Asians are more mobile in academics and classical music.

Jews are more mobile in finance, computers, law, media, etc.

Mexicans are mobile in buying more beans to make burritos.

From the report: Communities with higher local taxes to support education have higher social mobility and higher test scores. Poor whites in segregated communities have less mobility than poor whites in more integrated communities. Stay away from parts of the South and move to Utah or the rural midwest, or San Jose.

Based on my knowledge of property taxes there, Western New York has a great base for social mobility. It just needs more socially mobile people to move there.

"For example, the probability that a child reaches the top quintile of the national income distribution starting from a family in the bottom quintile is 4.4% in Charlotte but 12.9% in San Jose."

I pointed this out to Leonhardt back in July: A major problem with Chetty's study is that it looks only at income inequality, not standard of living inequality. Charlotte ranks at the bottom of Income Mobility because it has maintained relatively cheap housing over the last generation, while San Jose's housing has become wildly expensive. But that also means that incomes in Charlotte have not gone up all that much to compensate because the cost of living has remained modest. In contrast, San Jose has become vastly more expensive, and incomes have risen considerably to compensate.

San Jose has become vastly more expensive, and incomes have risen considerably to compensate.

I think you have it backwards. Home prices rose as a result of an influx of high paying jobs.

I see what you're saying, but no matter what the cause of the differences in wages, when you're looking at intergenerational differences in income there should be some controlling for differences in cost of living between cities.

My wife is from San Jose. As an example, she has a friend who went to an average-ish CA state college, got an English degree and now does some sort of non-technical gig at Yahoo. She probably makes $120,000. If she had been born in Charlotte studied at NC State and started a career in Charlotte she probably would not be pulling that $120k. However, back 1980, average family incomes in San Jose were probably not much more than in Charlotte.

If, say, she was making $80k in Charlotte, would her standard of living be higher or lower than in San Jose, or about the same.

Chetty's study last summer simply ignored such questions, which is why Charlotte came in worst in the country in social mobility, even though it has been attracting quite a few of the moderately ambitious.

I'm curious why 'mobility' is defined as moving from the bottom to top quintile? Is this a convention? Or something specific to this paper?

That seems like really a corner case. I don't think of mobility as being 'rags to riches' so much as the ability to improve one's family through education and hard work. I'm not sure how much this statistic tells us about that. What about a measure such as probability of moving up by at least one quintile, or by at least two quintiles?

This struck me as well. Seems like any movement higher (or lower) would be defined as mobility and not just the most extreme movement which would seem like the most difficult to achieve.

Has anyone else noticed a decline in the level of Prof. Chetty's scholarship since he left Berkeley for Harvard?

This is interesting. Before commenting on whether social mobility is declining or increasing, what is the ideal state? I think, overall, it looks pretty good in America. The middle classes have a lot of mobility, but the poor stay poor and the rich stay rich. Not perfect, but not terrible.

But where I think we need to look more closely is differences among races/genders. For example, this website discusses the Pew Trusts study on social mobility pretty well: there's a stark difference between the mobility of black people and white males: http://priceonomics.com/social-mobility-statistics-are-racist/

"there’s a stark difference between the mobility of black people and white males"

Probably because of their genetic differences in average intelligence.

Could be explained by regression to the mean, which may be different for blacks and whites.

Also, Page 30 of the Pew report. For the bottom quintile, at least, test scores account for basically all of the black-white difference in the probability of moving out of the quintile.

Yes, I pointed out to Leonhardt last summer that regression toward the mean is driving much of Chetty's results. For example, Chetty found that people born in extremely white West Virginia have fairly high income mobility (even though West Virginia is increasingly a dismal place to live according to most statistics). That's because a large number of West Virginian natives have moved out to higher paying places like the Washington DC area or Atlanta.

In contrast, Georgia shows up as low income mobility, even though it has attracted many newcomers, black and white.

Exactly right: regression toward the mean explains much of the regional results. See p. 31 of the linked report.

Areas that have small black populations have higher social mobility from the bottom up among their natives than areas that have large black populations. The simplest explanations are that in metropolitan zones that have large black populations, more of the bottom quintile are black, while in areas with small black populations, more of the bottom quintile are white, Asian, or Hispanic.

Blacks regress toward a lower income means than do whites. That's why Atlanta is shown as having terrible income mobility, even though Atlanta has struck many intelligent blacks as a good place to move to to raise their families. The bottom quintile in Atlanta is overwhelmingly black, and thus the bottom quintile regresses toward the black mean. In contrast, San Jose, the top city in income mobility, has virtually zero blacks, so the bottom quintile regresses toward the income means for American born Hispanics, American born Asians, and whites.

Interestingly black Americans get most of some of the most desired jobs, those being college and professional basketball and football player. Even though many might say the college basketball and football player is not a good job due to the low pay (free tuition), they are highly desired and sought after positions, people who have played college basketball (and some who haven't) talk about and gains status from it for the rest of there lives.

This leads me to doubt that blacks lack of movement up in income is due to prejudice but is more likely due to a lack of desire to move up.

As commentariette says, "mobility" is always defined as moving up. But it we're talking in relative terms -- such as movement from the fourth income quintile to the fifth quintile -- there must be a person leaving an income quintile for every person who joins it.

Thus, rather than improving primary schools for the poor, income mobility could easily be increased by addicting the upper classes to some life-damaging behaviour. Imagine if we could get students at elite universities to indulge in heroin rather than alcohol and promiscuity...

What in the paper justifies the post title? Is it the comparison to Solon (1992) on p. 1? Otherwise, I see lots about spatial variation in intergenerational mobility, but not much about temporal variation in intergenerational mobility.

The way people approach this subject puzzles me somewhat. If one thinks of a totally random dispersion among 5 quintiles, the chances of an adult landing in any given quintile is obviously 20%. If there is skill, effort and other factors involved, that suggests a variety of alternative distributions, most of which are conservative of the immediately preceding distribution given that skill, effort etc require the expenditure of energy and nature tends to conserve energy. So for example, a probability distribution for a member of the third quintile might look something like: 40% chance of landing in that same quintile; 20% chance of landing in each of the adjacent quintiles; and 10% chance of landing in each of the farthest quintiles. By the same logic, you would expect a move from bottom to top or v.v. would be the least likely to occur. It might look something like: 40% chance of staying in the one into which s/he was born; 28% chance of landing in the adjacent one; 17% the next; 10% the next and 5% the farthest.

What puzzles me is that so many smart people see these distributions as heavily shaped by institutional factors and I think a lot of that analysis is just being "fooled by randomness", and these distributions cannot be confidently distinguished from those that would arise from plausible combinations of skill and luck and the normal deviations that might occur in a large panel of outcomes..

Much also has to do with how compressed the ranges of the quintiles are. It is easier to jump over narrower quintiles that broader ones.

The big shift in the income distribution against workers at the bottom occurred in the 1980s before the oldest people in this study entered the labor force.


See also: http://jaredbernsteinblog.com/stable-income-mobility-not-a-muddle-at-all/

As the income distribution widens—as inequality increases—it requires an increased rate of mobility to move from say the 10th to the 50th percentile. In this regard, stable mobility amidst increasing inequality poses its own challenge. You’re no more or less likely to move up and down the income scale relative to your birth cohort, but the distance between you and those above and below you is a lot greater.

Second, the other concern that many of us in this research have expressed is that higher inequality, through a channel by which both parents and society are unable (parents) or unwilling (governments) to invest in their children, could eventually lead to diminished mobility.

What's interesting is that the study focuses on movement between quintiles, and not income changes for quintiles.

So, if real income declines for all quintiles except one, you would not notice it in the stats.

It sounds like the youngest cohort Chetty considers is people born in 1980-82. Therefore I would hesitate to put such a definitive title ("upward mobility in the US is not declining as many citizens think") as though one academic study closes the case on whether social mobility is currently in decline. If social mobility has in fact been declining since the recession or for say the last 10 years, when would you expect to really see it? The sort of things that would lead to a decline in social mobility (relative increase in importance of connections/money/schools/etc) have an impact earlier in life and pay off later in life, so all this paper says to me is that social mobility didn't decline much in the 90's. What I'd like to see is how parental income correlates with the number of 20-30 year olds unemployed currently. I'd bet that the well-off are employed or in grad school while the have-nots are more likely to be unemployed or working at McDonalds and that this will lead to an observed decline in mobility when the same methods Chetty uses here are replicated with older cohorts in the future. When I hear people claim that there is a decline in social mobility, its usually not "I'm 37 and have been working hard my whole life and don't have a job because I didn't have connections growing up," its "I'm 27 and have been working hard my whole life but I find myself unemployed and with 100k in student loan debt." Let's see how those people's incomes compare to their parents in ten years...

The rise in inequality is partly due to the market reach of people like Lebron James but he is also an example of a child of low income parents making it bog.

Sports figures like Lebron James are outliers in terms of their "market reach," as well as physical and intellectual talent, but they don't seem to bother people as much as corporate CEOs with high salaries. But perhaps the CEOs are similar to sports stars: outliers with gifts, skills, and talents particularly suited to running large organizations and earning money.

Comments for this post are closed