9 Comments

The admission graph is quite interesting and seems to suggest that Ivy plus colleges are at the very least able to give some additional consideration to students from more difficult financial backgrounds.

I wonder, however, if that may be misleading in that it could simply be a happy accident that results from racial diversity preferences in selecting students.

Also what do you think would be the impact of a lottery system on the way society looks at achievement? I may be wrong but I imagine most students feel that if they achieve at an extremely high level they will be able to get to their school of choice despite any implicit or explicit diversity quotas. If a student with a perfect GPA and SAT score, who is a concert pianist and does volunteer work on the weekend could still be deprived of admission to Harvard by a coin toss, could it not potentially reduce his enthusiasm for academic achievement overall? Almost akin to a what punitively high top marginal tax rate might do to the desire of someone to earn a lot of money.

If faced with a lottery system, I wonder if students may simply value those institutions less and focus on more on entrepreneurial ventures. Which could turn out to be a net gain for society.

Expand full comment
author

I think a lottery system would even out prestige at the top and elevate some of the really good state schools like Michigan Wisconsin and UCLA. Bob Frank book is great, we already have an invisible lottery system. I don't think that marginal tax rates have the effect you imply, top earners care much more about their relative ranking than absolute income. The lake house will still go to the highest bidder even if all bidders have lower post-tax income.

Expand full comment

Thank you for your reply, professor. Even if we grant that luck plays a major role in deciding let's say which students are even in a position to apply for top colleges, won't the negation of the perception of merit in the final decision have some sort of an impact on how ambitious students act?

If all students above a certain achievement threshold have an equal chance of getting into the top colleges, the incentive for the most ambitious students to dedicate time and effort in building skills and capabilities geared towards success in these schools would presumably be reduced. That would not necessarily be a bad thing but wouldn't it at least have an impact?

I must concede that the analogy with a punitive marginal tax rate felt a little flimsy to me even as I wrote it but wouldn't the impact of such a rate be dependent on what tax bracket is affected?

I would imaging if it the rate is high enough and affects an income level that is low enough, (not low per se of course) then the potential indicators of the difference in relative ranking of top earners might be low enough that they prioritise leisure time over achievement.

Sorry for really long response. I understand completely that this is not a class for you to address all points raised by students. Thank you for sharing your thoughts. I am enjoying going through your posts after watching your conversation with the brilliant Glenn Loury.

Expand full comment

I always thought the higher admissions rate for 50th percentile and below was because all of the elite schools give admissions preferences for being lower-SES and/or first-generation. For example, Harvard, while technically being need-blind, goes to great lengths to label applicants as “disadvantaged” based on whatever other information it can glean from the applicant’s profile (ZIP code, high school, parental education / occupation, etc.).

Also, that chart is difficult to interpret in part because I doubt the application rate is uniform across the parental income distribution.

Expand full comment
author

Yes, the gradual decline in admission rate as one moves up the distribution reflects changes in the interpretation of scores, and is consistent with the model in the forthcoming paper I linked to at the end of the post. This makes the sharp rise at the end especially troubling.

Application rate is very high at the top, you're right, but presumably would be lower if admission likelihood were lower (would be interesting to see MIT/Chicago versus Dartmouth/Stanford)

Expand full comment
Jul 25, 2023Liked by Rajiv Sethi

One advantage of using lotteries for the final selections for elite institutions would be to reduce the corrosive and corrupting effects of competition among candidates for these schools. If the top students at our best high schools know that meeting a quality standard to get into the lottery pool is more important that getting a higher rating than the student sitting next to them, they will be more inclined to work with their fellow students and improve everyone's learning.

Expand full comment
author

Agreed, excellent point.

Expand full comment

Another way to interpret the graph is that wealth doesn't really matter for the vast majority of applicants. It's not until one gets to very small 0.1% of wealth that we see an impact. I think that is a bigger surprise than anything else, but because it doesn't fit the NYT narrative, we don't hear about it.

Expand full comment
author

Likelihood of admission seems to be declining with parental income until you hit the really high levels. The forthcoming paper I linked to at the end predicts this, since any given score corresponds to lower ability when it is attached to a high resource individual. This makes the sharp rise at the end even more troubling. I think the NYT article is pretty straightforward reporting on the paper, I don't see any attempt to twist it to support a narrative.

Expand full comment