"Moreover, any loss in alumni giving at the elite institutions would presumably be at least partially, if not fully, offset by additional giving to alternative institutions where parents of wealthy students might donate as an alternative, and the wealthy students themselves would become alumni. "
I think that paper over-indexes on exactly that argument - that having a relative attend the school is just a proxy for wealth, so they would give just as much anyway if they went to a different school. It doesn't stand to reason at all that the donations will be transferred to whatever other school their kid attends, donations like this are not fungible and do not scale linearly - if everyone in your family for five generations has gone to Harvard, and your kid goes to Arizona State, you aren't suddenly going to build a family tradition of loving Arizona State. You're more likely to simply dilute the strength of the family-college connection and not donate at all, or donate a perfunctory amount and put the rest into the opera you attend these days. I didn't look up the authors of this study, but they certainly didn't bother to mention any familiarity with the field of fundraising, which isn't as straightforward as they make it sound. Even if the strength of connection was the same, do public schools have the same skills and desire to request donations as Harvard does?
To test that hypothesis: they should check the Arizona State alumni to see if the number of relatives who attended Harvard is just as good a predictor of donations to Arizona State, because it's equally as strong a proxy for wealth when you attend Arizona State as when you attend Harvard. My bet is that it's a very weak predictor!
I think that paper over-indexes on exactly that argument - that having a relative attend the school is just a proxy for wealth, so they would give just as much anyway if they went to a different school. It doesn't stand to reason at all that the donations will be transferred to whatever other school their kid attends, donations like this are not fungible and do not scale linearly - if everyone in your family for five generations has gone to Harvard, and your kid goes to Arizona State, you aren't suddenly going to build a family tradition of loving Arizona State. You're more likely to simply dilute the strength of the family-college connection and not donate at all, or donate a perfunctory amount and put the rest into the opera you attend these days. I didn't look up the authors of this study, but they certainly didn't bother to mention any familiarity with the field of fundraising, which isn't as straightforward as they make it sound. Even if the strength of connection was the same, do public schools have the same skills and desire to request donations as Harvard does?
To test that hypothesis: they should check the Arizona State alumni to see if the number of relatives who attended Harvard is just as good a predictor of donations to Arizona State, because it's equally as strong a proxy for wealth when you attend Arizona State as when you attend Harvard. My bet is that it's a very weak predictor!