A really interesting read!
I would love to see this model re-run with some covariates around class format (at a minimum, a variable capturing % of classes taken in lab and seminar formats).
Interestingly (per quote below), it seems that offering college classes online could be beneficial in terms of access for lower income students. However, I'd caution that the low WTP of lower income students for social activities may be due to a crowding out effect if the reason they participate less in social activities is due to a need to work to finance their studies.
The subsequent rows of Table 4 show that even within subgroup there is substantial heterogeneity, with
the difference between the 10th and 90th percentiles of the individual-specific WTPs being more than $3,000 for almost every demographic group. The 90th percentiles are sizable: second-generation and higher-income students at this percentile are willing to pay nearly $3,500 for each of the two amenities (social life and in-person instruction). First-generation and lower-income students at the 90th percentile are willing to
pay approximately $2,000 for these amenities. There is even a meaningful share of students with negative willingness-to-pay for both in-person and on-campus social activities, leaving the median across most groups around $0.
What's promising is that information shows what tuition discounts for lower income students would be required to close the gap in access to social opportunities. This also shows the value of social connection and suggests that universities would do well to explore what aspects of on-campus socialization generate the most network value for students in the marketplace. It would be a unique selling point to distinguish programs, especially mid-market ones.
If I were a university competing for top students, I'd want to quantify expected impacts of participating in the school's social network on mental health, physical health, quality-adjusted social connection, employment prospects, earning potential, and success in program of study. Even with simple self-report scales, I think that would be a powerful value proposition.
Caveat: There's inevitably collinearity between student WTP for social activities and the monetary value of good social skills in the market (i.e. being inherently social and participating in social activities that create future career opportunities are not statistically independent) . But ratings could be normalized by student scores on 5-factor personality inventories.
I'm now laughing to myself imagining a PR campaign: "Come to University X: where introverts can succeed with minimal forced socialization!"
As previously mentioned, the value of campus social life may, in part, be due to the formation of social networks which provide advantages on the job market or insulate members from bad shocks. We find evidence
of this in our survey; students with higher social WTP expect to make more money at age 35, with a $1 per year increase in WTP associated with an average increase of $0.81 in expected annual earnings
Cool share, Bostonian. You'd best get on the phone with your local congressperson!
