<p>Firing and penalizing is great in theory. Since hosting doesn't happen that often, I'd question the practical results of it. Granting that it might help, though, here's why it's not a a good economic choice:</p>
<p>The number of students who come to an overnight is small, compared to the total number of admitted students. (It's a big chunk of people, but a good number of students never visit/don't stay overnight.) Of those, the number who are truly undecided is somewhat smaller, and of those the number who are so undecided that they will be swayed by specific personal overnight experience (as opposed to the general experience that can be taken in just by having the chance to stay overnight or the whole visit in general) is even smaller. Basically, not that many people are going to base the decision on something the host can control.</p>
<p>Then there's the kinds of hosts:</p>
<p>A. Those who love prospies, would host every week for free.
B. Those who love prospies, host when asked (for free) and put forth a reasonable effort to make it a good experience.
C. Those who'll host when needed, but not do much for the prospie.
D. Those who'll let the prospie sleep on their floor, but avoid them otherwise.</p>
<p>Paying hosts isn't going to change groups A and B at all. Group C might be a little better hosts, and Group D are the people who would get replaced (it'd be impossible to get enough info about Group C to "fire" them, and they would still probably be inferior to the folks who'd do it well for free). </p>
<p>You wouldn't eliminate too many from Group D anyway, based on the evaluation (there aren't that many overnights to provide opportunities for a bad evaluation), and their replacements aren't likely to be much better given that the most willing hosts are already hosting. Only a few of Group C might, out of psychological obligation, improve their hosting, but I doubt it would be too dramatic. Partly because being a good host is kinda hit-or-miss. If you are really quite different in personality/interest/etc from the prospie, your best move might be to find someone else to handle them. </p>
<p>In order for host improvements to have any utility, the would-have-been-bad-but-are-now-good hosts would have to get paired with the I'm-so-unsure-that-the-decision-will-turn-on-my-specific-overnight-experience propsies. Even if that happened at the average rate, and even if the host-improvement program worked (which I'm skeptical of), you'd be dealing with an extremely low yield. When you consider that the hosts would have to be bad at least once before getting fired, you're making the potential payoff even lower.</p>
<p>Paying hosts would be putting a significant amount of money (and the time of sorting out the evaluations, etc.) into something that, even if it worked, would probably not yield many more students. Given that the effectiveness is rather murky and the costs (both time and money) of making the system work well would be high, host paying would probably be a rather poor economic decision. The resources can be used much more effieciently on other methods of recruiting and more focused efforts by admissions, or perhaps financial aid.</p>