The Curious Case of Balanced/Equilateral Higher Education Institutions and Potential Implications for Academic Libraries

Our Institutional Directions Model of US Universities and Colleges, which we have developed as part of The University Futures, Library Futures project, offers a novel system of accounting for the main educational directions of 1506 public and nonprofit US higher education Institutions.

Simply put, our model calculates the extant to which a university’s educational activity is focused on one of the following: (1) doctoral level research, (2) liberal education in the arts and sciences, and (3) career-oriented education and professional training. For each institution within our project population, our model calculates the percentage of each of these three educational directions, all three of which always add up to a total of one hundred percent. Namely, an institution with a score of 33% Research, 33% Liberal Education, and 33% Career would have its educational directions equally divided. For more information about the model’s underpinnings, please refer to our data set and scoring formula.

Because our research is focused on institutional differentiation and identifying distinctive institutional types, we have generally looked at cohorts defined by strong, shared directional emphasis. For example, institutions with a strong directional emphasis on Research, or Career-directed learning. However, it is equally interesting to consider the case of institutions with relatively balanced educational activity in each of the directions we are examining.

To this end, we extracted a subset of institutions for which educational activity on Research, Liberal Education, and Career is between 28% and 38% (I.e., 5% above or below 33%). On a radar graph, these institutions would look like a set of nearly equilateral triangles. Interestingly, there are relatively few of these: just nine institutions in our population of 1506.

To investigate how modifying the upper and lower bounds of this range affects the count of institutions, we increased the range to 5% above or below 33%. The results are presented in the table below:

Research Liberal Education Career Count of Institutions Percent of Institutions in UFLF population
>32% to <34% >32% to <34% >32% to <34% 0 0
>28% to <38% >28% to <38% >28% to <38% 9 0.6%
>28% to <38% >28% to <38% 1% to 100% 11 0.7%
1% to 100% >28% to <38% >28% to <38% 9 0.6%
>28% to <38% 1% to 100% >28% to <38% 16 1.0%

The nine institutions that fall between 28% and 38% in all categories, or between 28% and 38% on the Liberal Education and Career directions are identical.

Cumulatively, then, a total of 36 institutions (or 2.3% of the project population) exhibit a relatively “equilateral” distribution. This is clearly a very small share of the overall population. While a majority of the population exhibits strong directionality toward one, or at most two, of the educational directions we are examining, it is nonetheless interesting to consider the significance of an institutional identity that is more evenly distributed across Research, Liberal Education and Career-directed programs, for example:

  • What does the small number of institutions (in our project population) that exhibit this pattern tell us about what it takes to succeed in a competitive higher education marketplace?
  • Is it by initial design that these institutions (or some of them) offer a balanced, generalist type of education?
  • If not by initial design, have they intentionally moved in this direction (toward the center) or were they crowded out of another more highly differentiated space?
  • What do academic library services look like in these institutions? Are they notably different from the library service bundle in institutions with a “spiky” or asymmetrical distribution?
  • Is it any easier (or more challenging) to support academic library services in an institution with evenly distributed educational directions vs. an institution with a stronger emphasis in one or more direction(s)?

We invite our readers to share their thoughts in the comments section below or by email ( or

We thank our colleague Brian Lavoie for reviewing an earlier version of this post and providing helpful comments.