Tyler Cowen thinks that at least part of higher education should be redesigned from scratch.
especially those tiers below the top elite universities. Completion rates are astonishingly low, and also not very transparent (maybe about 40 percent?). I would ensure that every single student receives a reasonable amount of one-on-one tutoring and/or mentoring in his or her first two years. In return, along budgetary lines, I would sacrifice whatever else needs to go, in order to assure that end.
I am not sure I would exempt the elite schools. Just because they have better graduation rates does not mean that the benefits exceed the costs. Depends on how you measure, especially “compared to what.” You can’t just take two people and say this one went to Harvard and that one didn’t and compare their earnings. You have to figure out what would have happened had both taken the same educational path. My reading of the relevant literature is that it supports the Null Hypothesis.
I think that the idea of close personal mentoring is the model that existed hundreds of years ago. It might still be the best model.
The academy sells the idea that people can advance themselves/their children economically and socially.* Its ability to make good on this promise is often highly questionable, but enough people want what it offers that tearing it down would be hugely unpopular. It is not unlike religion in this regard, and religion has been remarkably durable.
*Many universities also offer education for the tiny percentage of students who are interested.
When I taught high school students, my impression was that they were even more concerned by a slightly different possibility, “If you don’t go to college, you are screwed.” Lack of college means downward mobility and college is the only way to advance socially and economically.
If there was another path to advancement, colleges would quickly go back to what they were before the G.I. Bill – sleepy little backwaters for wonks. But until there is, they’ll have huge popular support.
Individual mentoring won’t work at scale- not many professors are famous enough that their endorsement would be widely credible, therefore valuable.
In some ways current employment law makes the downward mobility a reality. Warren Meyer of Coyote Blog has an article in Cato’s Regulation magazine, summer 2018 on ‘How Labor Regulation Harms Unskilled Workers’.
He offered the following reason employers favor college grads due to the risk of an employee causing offense to a customer
“One potential way employers can manage this risk is to shift their hiring from unskilled employees to college graduates. Consider the risk of an employee making a racist or sexist statement to a customer or coworker (and in the process creating a large potential liability for the company). Almost any college graduate will have been steeped in racial and gender sensitivity messages for four years, while an employer might have an hour or two of training on these topics for unskilled workers. Similarly, because
good information on prospective employees—credit checks, back- ground checks, reference checks, discussions of past employment and salary—all have new legal limitations, employers who hire college graduates benefit from the substantial due diligence universities perform in their admissions process.”
I don’t know what it’s like at elite American universities, but one of the benefits of Oxford and Cambridge are that you get several hours a week of compulsory one-to-two tutoring to monitor your progress and fix gaps in your knowledge and capabilities.
I don’t believe lower tier UK universities provide this (although there is sometimes voluntary provision of tutoring).
Quite apart from the higher academic ability of the student intake at Oxbridge, this has to be one of the big advantages of these Universities.
I did my graduate work at a respectable ivy-league American university. We didn’t have anything like that. When I was a teaching assistant (TA) in freshman chemistry, we had 1700 students, 1 professor, and about 40 graduate TAs, mostly first-years. The TAs major priorities, in order, were to gain admission to our desired research groups, complete our own rigorous coursework, perform our TA responsibilities well enough not to kicked out, and get drunk and laid. We had little understanding and less interest in matters relating to teaching. Many of us prioritized “drunk and laid” above “not getting kicked out”, with predictable results.
Generally American universities are organized by subject; a typical student will have five classes at a time, about a third of which are in his major subject. The remainder are related subjects (as a chemist, I took a lot of physics and math), general requirements (pretty much everyone had to take English Composition), and breadth requirements (we joked that our department offered “chemistry for non-majors”, “chemistry for poets”, and “chemistry for vertebrates”, and as undergrads we took similar courses in other departments). When it happens, tutoring usually takes place at the course or subject level; as a TA I held office hours once a week which were gloriously restful except right before a test.
I agree that the Oxford tutorial system is superior to the American system, which I experienced at Princeton and Stanford. I learned so much law at Oxford that I could teach at an American law school 20 years later with only a little review.
Has anyone attempted to tease out the effect of Oxbridge style tutoring, e.g. by randomly assigning half of a cohort of otherwise-similar students at some “lesser” institution to receive it? If it really is good for long-term outcomes this seems like an extremely important thing to know, and it doesn’t seem any harder to study than any other often-studied educational intervention.
I suppose it depends on what you mean by the ‘top elite universities’ but in general, they offer a very substantial amount of personal and one-on-one instruction to those who desire it. Ox-bridge certainly still uses the tutor model. MIT uses the UROPs. Harvard has a diversity of models across the various disciplines, the houses, and so on; with mentors at the near-peer through senior tenured and emeritus levels. CS, pre-med, entrepreneurs, science… each handle the need somewhat differently. Some undergrads avail themselves, others not as much. Yale also has diverse tutoring options, but my experience is that student utilization isn’t as robust.
Why does the market fail? You’d think with such poor completion rates people would stop throwing money away.
Why do so many people go on diets, when a year later, the vast majority have
1) gained back most of the weight they lost, or
2) gained back all of the weight they lost, or
3) gained back all of the weight they lost and added a few pounds?
Imagine if you had Pell Grants and subsidized loans that would pay for pre-packaged meals, meetings with counselors, etc.
Throw guaranteed loans at 18 year olds with no sense.
No need to redesign, just adopt the UK’s higher education system which is much superior to that of the US. First, three years of schooling are all that is necessary, the 4th year in the US is superflous and should be discarded. Second, junk the arbitrary and subjective grading that is done in the US. All of the group projects and busy-work assignments are pointless and negatively impact diversity by imposing a burden that is harder for working class students who have jobs and other commitments to bear that it is for ruling class students who have all the time in the world for such tomfoolery. One end-of-term objective examination is all that is necessary. Third, junk all the class participation nonsense. Professors are being paid to profess. Bring back the sage on the stage. It never left the UK and the UK does not have near as much censorship and intolerance of alternative thoughts. Finally, take accreditation away from the education guild. Just as the judges in the US courts act primarily to boost revenues for the legal guild, so do academics who participate in shaping accreditation standards. It is a clear conflict of interest. Trump’s proposal to move higher education policy to the Department of Labor should be enacted swiftly and accreditation should be made the responsibility of objective, non-conflicted entities free of featherbedding academics.
It would probably be good if the two different sides of universities are separated
Research and Study institutions on the one hand where the scholars pick their own students to train
and professional training and certification institutions on the other hand where the students are customers
those two different jobs clash when given to a single entity
Yes, and solve the accreditation problem. I wrote a bit about this a few years ago. https://www.chronicle.com/blogs/conversation/2014/02/28/home-college-an-idea-whose-time-has-come/
Arnold, didn’t you post, sometime in that last year or so, about a study that found that, if you take a population of people who had been admitted to elite schools, there was no difference in long term outcomes between those who attended the schools, and those chose to go elsewhere? That would bolster the null hypothesis, I believe.
Having said that, I’m almost positive that one-on-one instruction will change anyone’s outcome – primarily because they will get some high quality study time during the one-on-one that they usually wouldn’t have had. They probably end up a little more motivated to study on their own than otherwise.
However, this is very inefficient and expensive to do. I’m not sure if there is an efficient way to solve the motivation problem, other than to have kids truly understand the consequences of not applying themselves in college – a task no one has yet figured out.
It’s been figured out, it’s just hasn’t been applied in that setting, across the board. The answer is a tournament: to make peers into competitors for selection with a guaranteed herd-culling rate. “Look to your left and right. One of you won’t make it.”
Elite academics (or any high-status field) has a highly Malthusian pathway of many selective steps, and many off-ramps for those who can’t or won’t do what’s necessary to get to the next step. Stressful, yes, but motivating.
As soon as leaders and authoritires start succumbing to the temptation to lower standards to increase passage rates, the essential elements of hunger and fighting spirit wane considerably.
One thing that would dramatically affect college completion would be banning remediation, preventing colleges from accepting students who can’t do the work.
Tragically, we’re going in the opposite direction. The Cal State university system is giving credit for middle school math, ending remediation entirely, over the objections of staff.
American schools do not, not, not want to break up age cohorts. If you entered first grade with a hundred other kids, the system will do whatever it can to make sure you’re going to be in twelfth grade with the same hundred eleven years later.
That leads to, among other things, “they’ll fix it later” syndrome. Lots of your fifth graders can’t read? Pass the kids on and “they’ll fix it in middle school.” Lots of your eighth graders can’t do basic arithmetic? Pass the kids on and “they’ll fix it in high school.” Your twelfth graders can’t put together three sentences that are grammatical and follow logically one after another? Pass the kids and “they’ll fix it in college.”
Some times it gets fixed. Some times it doesn’t. “High stakes tests” were supposed to change this. My impression is that they have largely succumbed to some combination of being made easier, having the passing grade lowered, and becoming less high stakes. “We don’t believe the score on one test is a fair and accurate representation of what a student has learned. Instead, we use a holistic approach …”
It is an interesting question why schools are so averse to splitting age cohorts. Part of it is simply not wanting to make kids feel bad. Don’t laugh; schools really care about that. That is one reason tracking is so disfavored. Then there is, “Students who are left back don’t learn any better than students who are failing but get promoted.” Which is largely true. Unsaid is that neither group makes much progress.
I think a major reason for preserving age cohorts is that teachers and administrators know that much of school is social. Most kids aren’t very excited about the learning but they do like to be with their friends and acquaintances. Part of getting them to do what you want them to do (and keeping them from screwing around and being “disruptive”) is to foster a sense of groupness (one reason for e.g., football teams). That means not splitting people up.
What has come to be regarded (in the U.S.) as “Academia” is now largely engulfed in the “Education System,” which, in turn, is, at its core a “Teaching System” that evolved from the learning facilities (one of Quigley’s “instrumentalities”) our society generated from about the mid-19th century.
What Tyler Cowen (and Sal Kahn) note is the need to re-establish **learning** facilities. The need arises out of the effects of the economic, political and social relationships (internal & external) that have come to comprise the structures of the extent “Education Systems” at all levels – and clearly at the post-secondary systems.
Examinations of those relationships (in all their complexities) in our present “Education Systems,” auger for quite different facilities, of different orientation, rather than attempts to “reform” the extent structures.
In my model the colleges went from regular classroom to smaller, less frequent discussion groups.
Most of the background work is online, done via a class group message group with reading links. Beyond that, the students occasionally travels to the professor and assistants for scheduled group and one on one.
“I think that the idea of close personal mentoring is the model that existed hundreds of years ago. ”
Not sure if the early University of Bologna is what you have in mind, but it does provide a useful reform model that would be entirely consistent with close, personal mentoring. The reason mentoring/advising in universities today is completely pro-forma and unrewarding is the arbitrary assignment and bureaucratic box checking involved. In the early years of the University of Bologna, the oldest operating university in the world, groups of students actually selected and hired their professors. This would have involved much interpersonal contact and communication that would in turn lead to more rewarding educational outcomes. According to wikipedia: ” University professors were hired, fired, and had their pay determined by an elected council of two representatives from every student “nation” which governed the institution, with the most important decisions requiring a majority vote from all the students to ratify. The professors could also be fined if they failed to finish classes on time, or complete course material by the end of the semester. A student committee, the “Denouncers of Professors”, kept tabs on them and reported any misbehavior. Professors themselves were not powerless, however, forming a College of Teachers, and securing the rights to set examination fees and degree requirements.” One of the highest performing high school programs in the Washington, DC, the HB Woodlawn program in Arlington, Virginia, incorporates some of these features by through a very substantive teacher-student mentoring system, inclusion of students on faculty hiring committees, student control in curriculum decisions, weekly town hall meetings etc. It is a very far cry from the “give us your money and we will arbitrarily assign you teachers and advisers” model so prevalent in the USA.
As Handle said; motivation comes from peer competition. Tutors were not always able to instruct the children of aristocrats; many just weren’t interested, and no tutor can motivate an unwilling child. Of course those kids with a scholarly temperament indeed grew extremely well with 1 to 1 instruction. But in order to have a scholarly *culture* you need to put kids together and have them compete to be scholarly.
“You can’t just take two people and say this one went to Harvard and that one didn’t and compare their earnings. You have to figure out what would have happened had both taken the same educational path. My reading of the relevant literature is that it supports the Null Hypothesis.”
What relevant literature did you have in mind? I brought up your point yesterday in conversation with an academic who argued against your point – that is, she said the research is pretty consistent that attending a higher quality schools does increase earnings.
Upon searching this literature on google scholar, I am inclined to agree with her – the studies I found (about 10) unanimously supported the notion of a return to school ranking.
Of course if you believe the returns to education are attributable to signaling rather than skill acquisition, this doesn’t necessarily conflict with the null hypothesis. In my experience, as a graduate from a top 10 school, I do believe employers are impressed by my degree. I also know of many firms that only recruit from a set of top schools.
Look up the paper by Krueger and Dale on college selectivity
People who go to a more selective school do indeed make more money after college. It is also universally acknowledged to be true that people on basketball teams are taller than people who aren’t on basketball teams. This does NOT mean that you will become taller if you join a basketball team.
Selective schools are by definition selective. They take students who have done better at making themselves attractive to the college. That means the students had to start out pretty smart. But they also had to work hard and be conscientious. They had to “play the game” of college admissions. Which means they had to want the reward, had to learn what was necessary to get it, and then had to do those things. People like that are more likely to be successful in other things, too. And since every kid is told that they should go to college and that they are cutting off major opportunities if they don’t, most ambitious 13-17 year olds at least make a try at playing the game.
Successful graduates make the school look good, which leads to more and better applicants, more “selectivity” and a student body that is better at doing what is necessary to get ahead.
Students who graduate from selective colleges do better than people who don’t. That makes it sound like the four years and the graduation are what is important. A different implication can be drawn from the equally true statement, students who are admitted to selective colleges do better than people who aren’t. In fact, I seem to recall research that kids who are admitted to a selective school but don’t go do just as well years later as those who were admitted and did go.
I think Arnold is referring to this paper:
http://www.nber.org/papers/w17159
“We find that the return to college selectivity is sizeable for both cohorts in regression models that control for variables commonly observed by researchers, such as student high school GPA and SAT scores. However, when we adjust for unobserved student ability by controlling for the average SAT score of the colleges that students applied to, our estimates of the return to college selectivity fall substantially and are generally indistinguishable from zero.”
Arnold – Thanks for recommending the paper. The authors have 2 papers (2002 and 2014) which both support your interpretation. Many of the papers I had previously found were older, so perhaps the Dale and Krueger papers are more representative of the current though.
Roger – I am well aware of all of these issues. The literature I am referring goes well beyond finding a simple correlation. See for example https://link.springer.com/article/10.1023/A:1023058330965 https://www.jstor.org/stable/146304?seq=1#page_scan_tab_contents https://www.journals.uchicago.edu/doi/abs/10.1086/298375 among many others.
“the idea of close personal mentoring is the model that existed hundreds of years ago. It might still be the best model.”
It probably IS the best model, but the elite colleges won’t do quite do it. And personal mentoring is fairly expensive.
If done by humans.
One of the huge coming ed revolutions will be personalized AI “Tutors” for each student, interacting with the student and reporting, some, to the teacher as well as to the school. Unlike the difficulty of teaching creative writing, most teaching of knowledge has 2-8 min length “chunks” of knowledge, which the student must learn. Such chunks can be tested for.
With better IT, there will be tests for learning learned, even outside of college / school. It’s coming more slowly than most thought / hoped. We’re still not so close to a HAL quality vocal interaction, tho that is finally looking more hopeful.
AI robot personal tutors.