The Lure of Academic and Social Reputations Versus Athletic Success: Influences on Enrollment Yield at NCAA Division I Institutions

Abstract

Colleges and universities face pressure to maintain enrollments in a time of demographic shifts in the college-going population and reductions in state funding. One indicator of successfully maintaining enrollments is the percentage of accepted students who matriculate—the enrollment yield. Factors known to contribute to yield include school size, cost, research, and reputation. Of interest in the present study is the import of academic reputation as measured by U. S. News and World Report rankings and social reputation as measured by designation as a ‘party school’ relative to accomplishments of the school’s high-profile athletic teams. I use a 21 year panel to model yield for all institutions competing at the highest level of intercollegiate athletics. The results show that yield rates consistently respond to USNWR rankings, but being named a party school has a more sporadic influence. Athletic success has little effect on a school’s enrollment yield. The findings suggest that the signals sent by academic rankings are stronger and better received than the signals sent by social or sports accomplishments.

Appendix A: Alternate Specifications

Different measures of athletic performance or school characteristics can be investigated though at a cost of fewer cases, a shorter panel, or more missing data. Athletic performance can be considered from multi-year perspectives of varying durations. Table 6 shows results when football and basketball measures are constructed over six- and fifteen-year windows. (Note the loss of cases and school years due to the longer time frame for athletic performance.) A six-year window covers the time frame when many students in an entering class begin to consider college and form choice sets and a successful team might put a school on a student’s radar. The longer 15-year window is geared more toward a culture of high-level athletic achievement and that too could influence matriculation decisions.

Table 6 Models using alternate measures of athletic performance

The previous year’s USNWR ranking has an impact similar to that found in analyses presented earlier, regardless of sport and how performance is measured. Movement into the upper echelon of those rankings has no impact, nor does rank as a ‘party school.’ As in Table 3, a new designation as a top 10 party school now is found to increase yield by over 1% among the sample of schools offering football. When considering football records over 6 years neither the harmonic mean of winning percentage, the number of appearances in post-season bowls, the number of times the season ended with an AP top 20 ranking, or the number of times an FCS school played in the tournament are related to yield. As the performance window increases, the individual indicators become much more highly correlated and that collinearity is responsible for the significant, opposite signed, effects seen for the 15-year average winning percentage and number of bowl appearances in 15 years. When each football indicator is run separately, no individual aspect of football 15 year success is related to a school’s yield. A similar conclusion is reached when different time frames are used for basketball performance with one exception. Now the harmonic average winning percentage over 15 years is found to increase yield suggesting that a consistently successful basketball program can be a draw for some students. But the overall conclusions—that there is a consistent influence of USNWR ranking and little impact of football or basketball performance—found earlier is confirmed using a longer time frame to measure those performances.

The Integrated Postsecondary Education Data System (IPEDS) offers more direct measures of some school characteristics known to influence matriculation decisions. A measure of the average financial aid award from the institution (in 2004 dollars) comes closer to considering the actual price paid by students. That variable is correlated .90 with the tuition variable used earlier so tuition is not included in the models below. The instructional expenditures per full time equivalent student is a common indicator of a focus on education at a school and as noted by an anonymous reviewer allows for interpreting any effects for USNWR rankings as more likely due to the influence of reputation. IPEDS data for these variables are available from 2000 through the end of the panel and compiled by the Delta Cost Project (www.deltacostproject.org).

IPEDS originally allowed public schools to report financial data by system rather than institution (e.g. all University of Texas schools were grouped with the Texas at Austin campus). While this practice was discontinued after 2003–2004, the Delta Cost Project continued grouping schools to maintain consistency across years. Financial aid and instructional expenditures have been spread to each campus of grouped institutions in the current panel though this is at the expense of more accurate school-based measurement of those variables. With less than desirable measures and an increase in missing data in some years and schools, any results should be considered suggestive at best. (Membership in the Association of American Universities is not used as it is effectively time-invariant over the shorter panel.)

Conclusions reached from using the different school characteristics (Table 7) are similar, though not identical, to those from Table 2. Now some school-level characteristics are related to changes in yield rate while before only tuition had an (negative) effect. As the size of the institution increases, yield declines. State-level unemployment and per capita income reduce yields. In the full models, yield declines as more students receive financial aid. Being in a Power 6 basketball conference continues to be associated with a lower yield. More importantly the majority of the conclusions concerning the ranking and sports performance variables are reproduced: USNWR ranking has a small positive influence on the change in a school’s yield while party school reputation does not and few measures of athletic success are related to yield. Movement into the top 10 of the AP football rankings, while having an almost identical coefficient, is now significant at just the .1 level (the standard error has increased). In this shorter panel there is no penalty for participating in the football championship. In the basketball sample we now see a positive effect for participating in the NCAA tournament two seasons earlier, though this can be offset by a decline in yield when the team moves into the top of the AP rankings.

Table 7 Models using IPEDS measures of costs and instruction

Overall the additional analyses of the Appendix serve to confirm the conclusions reached in the main text about the consistent influence of USNWR rank, the overall lack of influence of party school designation on yield, and only a few consequences for yield rates that are attributable to a school’s high-profile athletic teams.