Top colleges should select randomly from a pool of ‘good enough’ (2005)

First, let's not mince words here -- this suggested policy is about lowering the bar for admission to top universities so that some previously unqualified tier of applicants has a better shot at getting in. These lower tier students get a better chance in a lottery, at the expense of people who through better work/aptitude/achievement/(and no small amount of luck, finance, or legacy also) would otherwise usually get in with high likelihood. (Note, those last exceptional categories are a small portion -- something like 10%.)

So what's one of the side effects here? Or let's call it incentives produced by such a system?

Well, perhaps it's that students, who throughout school are taught to study hard, do well, apply themselves, at the end face a selection that tosses that out the window and says, well, everyone is equally good and we're going to do a lottery. We're lowering the bar so that everyone who got above a B average has an equal chance of getting into Harvard.

Is that teaching them the right lesson? Is that incentivizing people to excel and be really good at something? How would you respond at work if at the end of the year, everyone doing a minimum amount of work were given the same salary?

I don't think you'd actually like that (most people reading these forums, at least). Now, I don't dispute that there's a lot of unnecessary pressure on students today. And part of it is a bigger problem that the population has grown while top spots have stayed relatively constant. So people are trying to fix this.

But I disagree with attempts to create artificial equality at the last step. We wouldn't accept this for anything else that matters -- you wouldn't prefer to be treated by a doctor who was just "good enough", or be flown by a pilot who barely passed the flight exam. In the name of lowering their level of stress, or making the doctor / pilot population look a certain way. And by the way, to make a difference in college admissions (according to the function I can imagine the author desires), you really have to lower the bar a lot.

I don't want these kinds of distortions for a system that my kids will be subject to. That's not the lesson I want them to learn. Students in other countries have a far harder and more competitive time than ours. I don't want them to wake up to that reality only when it comes time to apply for their first job and find themselves falling flat.

It’s using a ruler to aim an electron microscope.

Standardised tests measure academic potential, i.e. probability of graduating and likely value added to society from the educational experience, with a fair amount of error. Taking everyone within a standard deviation of the cut-off and randomly selecting therefrom admits this error and removes the stigma from those just below the previous, false cut-off. If one really wanted, and if the statistics merited it, one could use a weighted randomisation, with individuals scoring higher getting an advantage compared to those scoring lower.