scott-forest-thumb.png Technology has made many industries more productive. But there is a tendency in academia, because the software investment is usually made by administrators, for software to be more related to the work of administrators than instructors. Whether it succeeds in making them more productive is an empirical question. What is not a question, though, is that the place where you want a productivity gain in high education is on the instructional side. Unfortunately almost all of the software that administrators select has the effect of making instructors LESS productive. And all manner of other practices follow the same pattern - from the reports that must be submitted to the meetings that must be attended. And more often than not you can trace this to the problem of organizational visibility that James Scott described.
draft00.png
scott-forest.png

My colleague, Mark Henderson, reminded me recently that a natural urge for an administrator or manager is to implement practices and technologies that increase the visibility (or what James Scott called "legibility"1) of the organization's workings to a person who sits where they sit. Once implemented, such systems make it look to the manager like things work better than before because she can see them more clearly. But this increased visibility tends to generate blindness too.

There is a related phenomenon - people at the top of organizations often track some stuff that makes them blind in this way but then seem to deliberately leave other things murky and obscured in a way that lets them ignore inconvenient details.

Administrators are famous for treating faculty time as a free and infinitely elastic resource. Because faculty work is relatively invisible and because as professionals and educators faculty tend to absorb the work and always manage to get the teaching done, administrators as individuals might be forgiven for this. But can we develop some tools that will help both faculty and administrators recognize the shape and size of this core institutional resource in hopes of improving the way it is deployed to achieve institutional goals?

Thought experiment: how many faculty-person-hours do I have at my disposal if I am running a SLAC? Let's suppose we have 100 faculty members working full time and each teaches 5 courses per year with the equivalent of a sixth course in advising and committee work. Let's further make the most conservative assumption that faculty work only during 34 weeks of term time (about 1400 hours) and that the research vs teaching/advising/service breakdown during this time is 1:9 (allowing faculty about 1/2 day per week for research).

If we allocate faculty work by the 5-6 rule then the 1260 hours available is divided as 1050 for teaching five courses and 210 for advising/service/etc. If we make the unrealistic assumption to treat teaching weeks and non-teaching weeks as identical we have over the 34 weeks (2 x 14 instructional weeks + prep week + exams + grading) and that the five courses are spread evenly over the two semesters this yields

30 hours instruction
6 hours service/advising
4 hours research

If we assume a salary of $100k then the hourly rate is about $73. Thus, for advising and service we allocate about $400 per full time faculty member per week.

$2200 instruction
$440 service/advising
$294 research

Now imagine you install a new expense tracking system and it costs Z and permits savings of W by the waste it prevents and X by the audit costs it reduces and Y by the clerical work it saves. In addition, it provides reports that give decision makers the feeling that they have a better appreciation for what is going on in the organization.

Imagine too that part of this transition involves requiring faculty members to enter their own reimbursement for travels and expenses. They do not do it often so they develop no deep familiarity with the system and each time they must be reimbursed it involves one hour of data entry, correspondence, etc. Let's assume each faculty member does this three times per year. To the institutional cost should be added 100 x 3 x $73=$22000.

One might argue that most faculty reimbursement is for faculty travel and most faculty travel is research related and so the best place to allocate this is on that line. If we do that, then we have

(3x73)/(34x294)=2% tax on faculty research
or
(3x73)/(34x440)=1.5% tax on advising
or
(3x73)/(34x2200)=0.3% tax on teaching