Is early admission a good idea?

Harvard just announced it would abolish early admission, and with Yana going away next year to college, this question has in any case been on my mind.

Marriage, of course, is run on an "early admissions" basis.  You ask one woman (man), get an answer, and you are then more or less bound to go through with it.

In my Mexican village, they run marriage more like we run college admissions.  Marriage-ready males propose to several women in one "season," and eligible women court more than one proposal.  At the magical moment, some of the prime candidates accept best offers and the rest of the market scrambles to clear.  But some men can, and do, propose "early admission" to their preferred females on an exclusive basis.

The musical chairs method makes sense if a) people definitely want to marry at a certain age, b) there is greater interchangeability of partners, and c) "being engaged" gives you an excuse to have sex with the person in the meantime or otherwise try them out.

Now let us say that one woman in the village is prettier than the others and also a harder worker.  She may wish to rule out "early admission" marriage proposals, out of fear she will have to turn down her second choice before having sounded out the interest level of her first choice.  (One question is whether a woman is a more populous community still might have enough market power to enforce such a policy with profit.)

Harvard says that early admission hurts diversity and the chances of minorities.  But, thinking about the marriage analogy, I wonder whether banning early admission a sign of Harvard’s growing desirability and market power…

Addendum: Greg Mankiw is happy about the change, Arnold Kling is not.  Here is another economic analysis, based on matching theory.