The problem with social mobility
Intergenerational mobility is not a great way to asses higher education. Don’t get mad at me yet. I explain why.
Originally posted on Review on May 24, 2021.
Ripping off the Band-Aid
Social mobility (as conventionally understood) is not a very good way to assess higher education.
There, I said it. Probably no one will notice but I will explain why just in case anyone objects.
Before getting into specifics, let me attempt to immunize myself from being seen as a jerk. I do think inequality and social reproduction in higher education is a problem. A big one. I’ve written about it, including in these ungated venues here and here. I’m not trying to do some devil’s advocate hipster-contrarian thing.
We need more social mobility. We need less social reproduction and more social mobility. Higher education has a role to play. I just don’t think the way we normally talk about social mobility is a good measure for assessing higher education.
What do I mean by social mobility?
When I say social mobility, in this context, I actually mean intergenerational income mobility - that is, where you sit on the income distribution relative to your parents. Say my parents were in the bottom 20% of the income distribution, but as an adult, I am in the top 20%. In that case I’ve experienced substantial intergenerational income mobility.
Income mobility is a commonly used measure and it’s what most people mean when they say “social mobility.” Income mobility is the measure Raj Chetty and the Opportunity Insights project use in for their Mobility Report Cards for colleges and universities in the U.S.
So when I say social mobility is a not-so-good measure of higher education performance, what I mean is that intergenerational income mobility is not such a good measure of higher education performance.
Why do we like to measure higher education by how well it does on social mobility?
Why do we like intergenerational income mobility to assess college performance?
The simple answer is that it makes intuitive sense. Going to college is supposed to provide opportunities for a better life. Colleges that enroll low-income students who go on to be middle and upper-income earners have done a good job on mobility. These colleges should be celebrated. That’s the intuition behind the Mobility Report Cards and I agree.
We should pay attention to which colleges and universities enroll large numbers of lower-income students and help to propel those students into meaningful and remunerative careers. Supporting institutions that support Black, Native, Latinx, AAPI, and other students’ that hold marginal identities is good policy. We should fund CUNY, and Cal State, and Maricopa Community College.
Taking the mobility imperative to the logical end gets complicated.
Building from the Mobility Report Card approach, the Postsecondary Value Commission developed concepts of value for assessing higher education. The Commission identified moving to the 60th, or better yet 80th, percentile on the income distribution at the third of five “thresholds” for determining postsecondary value.
The Commission’s report is suitably nuanced about establishing income-bracket attainment as a value threshold. Some students might have aspirations other than income. It also makes clear that attaining a 60th percentile income is a threshold only for students who originate from lower portions of the income distribution.
Meeting Threshold 3 requires that students attain the income necessary to enter at least the fourth income quintile (60th percentile). This was inspired by measurements of economic mobility introduced by the Opportunity Insights mobility report cards, though their metrics focused on reaching the top income quintile, which may not reflect all students’ personal or career aspirations. However, we recognize that for students already in the fourth- or fifth-income quintile, achieving this level of income represents economic stability, rather than mobility. The intention of this threshold is not to move already high-income students into lower-income brackets, but instead to raise the bar for earnings overall and encourage upward mobility for students from low-income backgrounds.
So, this threshold is about mobility of income – where one falls on the distribution – more than it is about income itself. It’s about making sure that college contributes to social equity. Seems good.
But it gets complicated quickly. I started to think out loud about this on Twitter.
1/ This “threshold” of value from the Postsecondary Value Commission seems odd to me. “Economic Mobility: This threshold measures whether students reach the level of earnings needed to enter the fourth (60th to 80th percentile) income quintile, regardless of the field of study.” Why?
2/ By focusing on a position in a distribution and not on harder to measure but probably more meaningful indicators about living wages, personal satisfaction, political and social enfranchisement the commission has boxed itself in.
3/ First, college graduates cannot get to >60 or >80% of the labor force for this measure to be possible. Second, it implies that if any meaningfully number is people without a college degree have incomes over the 60 or 80th percentile then college isn’t worth it.
Let’s pause here and think about what ensuring that all college graduates hit at least the 60th percentile for income means.
Today, about half of all adults in the U.S. (49.1% in 2019) have an Associate’s degree or higher and nearly 4 in 10 U.S. adults (38.7%) have a Bachelor’s degree or above. Now, we are a long way off from 60% of adults holding a degree, but if and when that mark is reached, to meet the Commission’s definition of value we’d have to have an income distribution that is perfectly segmented by educational attainment. Everyone with a degree would have to be in the top 60% of earners and everyone without a degree would have to be in the bottom 40% of earners. That seems both unlikely and also probably undesirable.
Even if you take the position that the degree should result in a 60th percentile or above income, which on its own seems right, a couple of problems arise quickly. First, people whose parents are wealthy are much more likely to go to college. We’d need to bank on more people who come from the top of the income distribution either not going to college or realizing downward mobility from college. We might want this to happen (the Commission does not call for it), but it currently seems unlikely.
The other issue is that the 60th percentile value threshold is in passive conflict with the completion agenda. As soon as you hit 50% plus 1 person of the adult population with a degree, then someone with a degree is necessary in the bottom half of the income distribution. In the U.S., somewhere around 70% of high school graduates, which of course are not all people in the country, go on to college (though this number seems alarmingly to have dipped in the most recent years). Over time, to ensure the 60% threshold for graduates comes students who enter college should complete college. Yes, for sure, we will never see 100% graduation rates, but it seems odd to me to define value based tacitly on the assumption that graduation rates won’t go up substantially.
In South Korea, a whopping 70% of adults aged 25 - 34 have a degree. Does that mean that South Korea’s higher education system is not producing value? Should South Korea work to lower the share of the population that gets a higher education? I am being a little bit flippant here, but you get the point.
Population with tertiary education, 25-34 year-olds / 55-64 year-olds, % in same age group, 2019 or latest available
We care mostly about upward mobility. That implies that college participation rates need to stay below a certain threshold, and that there is little point of going to college if it is not going to improve your income relative to all other earners. Both of these things mean the supply of higher education should be kept below truly universal access.
4/end I agree that economic mobility is important, but mobility measures this way is always 0 - sum. Holding ourselves to a threshold like this means that higher education participation will always be limited and sees it as fundamentally transactional.
Some other ways to think about it.
If I’m going to point out the limits of mobility as a measure to assess higher education I’d probably better say what I think is better. Here are a few ideas.
Economic measures are relevant.
As the Postsecondary Value Commission observes (it’s their “Threshold 0”), going to college should not make anyone worse-off. It is important that wages and other sources of well-being (now that is hard to measure!) provide as much benefit as the cost (tuition and fees, debt service, opportunity costs) of attendance for individuals. You will be glad to know I am not for trapping students with student loans.
A new report from the Pew Research Center shows that first-generation college students don’t see the same outcomes. First-generation college students don’t graduate at the same rate as continuing-generation students, they don’t see the same academic gains in college, they don’t go to the same type of colleges, and they don’t earn as much or accumulate as much wealth as do continuing-generation students.
Higher education might better to contribute to social mobility when there is no or little association between graduating from college and intergenerational income mobility. I know this sounds nuts, but hear me out. When graduates from rich families have the same odds of going down the income ladder as graduates from poorer families have going up the income ladder, then outcomes will become more equal. In a sense, higher education could do just as much for equity by providing a pathway to make some people poorer than their families as it does by providing a pathway to make some people richer.
Non-economic individual outcomes.
So far as I know, we only get one life. We ought to enjoy it. Education can help us enjoy our life. It can provide the resources, opportunity, and space for people to shape themselves. That seems like a good and valuable outcome to me. I hope that going to college sparks a life of the mind, an appreciation of cultures and of the natural world, and install an obligation to others. Measuring these things is hard, but I think we should assess higher education on these criteria.
Non-economic collective outcomes.
We should assess higher education value by assessing how well higher education contributes to social ideals. In the European and American political contexts, this might include strengthening democracy, civic participation, and social cohesion, reducing racism and xenophobia, and improving ecological sustainability. Powerful foundations, industry groups, and government could do more to emphasize collective returns. Measuring them is hard but can be done. For one example, we could justify tax transfers to higher education in part by how much an advanced education does to reduce graduates’ carbon footprints. Learning how to live a greener life seems like valuable education.
Let’s take an educational turn.
The short of it is that I think we need an educational turn in the way that we evaluate higher education. A shift away for understanding higher education as an organizational box in which human capital transactions are located, to a place that is central to civic, cultural, and cognitive development and expression. Ok. Enough.
A favor, please.
If you read all the way to this point, then you probably liked or hated what I had to say (the internet has lots of hate reading). Either way, please consider subscribing to my newsletter blog thing-y and sharing this post on social media. Cheers!
Edited to correct typos on may 25, 2021 at 10:45am.