Introduction
When the past Director of University Computing (Pat Urban) resigned her position this past summer, Interim Provost Gerald Francis formed the University Computing Study Committee and charged it with recommending an administrative structure for computing on campus. Provost Francis pointed to the University's needs both with respect to computing in particular and information technology in general, and asked the committee to consider what structure would be appropriate to serve our needs in both areas. He suggested that one possibility might be a Vice President for Computing and Information Technology, but that other possibilities should also be considered seriously. Such a position would be in line with recent trends, both in industry and academia, for a Chief Information Officer (CIO). Such positions carry a variety of titles; viz. Vice-Provost (U. Penna), Associate Vice President (U. Minn.), Assistant Chancellor (U. Wisc./Eau Clair), Vice Chancellor (U. Tenn.), Executive Director of Computing and Telecommun-cations (U. Wisc./Stout), and Dean of Information Technology (U. Wisc. /Madison).
At about the same time that this committee was appointed, CAUSE produced a study entitled The Chief Information Officer in Higher Education. Because that document has much to say that is directly relevant to the business of this committee, it is appropriate to quote from their assessment of the general problem to be addressed.
Higher education institutions have been grouped into three categories based on their perspective regarding information technology:
Category 1 institutions view information and supporting technologies as a strategic resource, and they view effective information resource management as a necessary condition of excellence....
Category 2 institutions view the management of information and supporting technologies as an aid to performing day-to-day functions. Information technologies are important, without doubt, but merely as enhancers of effectiveness and efficiency....
Category 3 institutions, perhaps the largest group, are confounded and confused by the role technology and information play in the strategic management of the institutions. For this group of institutions, the head of the information unit is expected both to be a technologist and to have executive capability, This information technology manager is expected to solve all technology-related problems, and this is frequently expected without the involvement of other senior administrators in the institution.
It is the concern of this committee that the University of Vermont currently falls in Category 3, while aspiring to membership in Category 2. The intent of our recommendations is to propose a structure that we see as offering an opportunity for UVM to move into the second category within a year, and to set the stage for moving to Category 1 within five years. If we fail to make this move, we stand at serious risk of becoming a lower quality institution. From a financial point of view we will probably not be able to compete for high quality applicants in a tight market. This will have very serious consequences for our tuition-based operating budget. We will also have difficulty retaining good faculty, who need better resources to do their work. Finally, our day to day operations cannot be expected to continue smoothly at our existing level of development.
The report by CAUSE also went on to outline the major issues faced by a "chief information officer", whether that person held the title or not, and whether that office reported to a President or a Vice-President / Provost. Ranked in order, these issues included "resources (how to pay for growing demands), networking, integrating computing into the curriculum, staffing issues, microcomputer support, organizational issues, desktop MIS for administrators, buying or building software, faculty writing software, ..." This is the same set of issues that arose repeatedly, with remarkably little variation, in our meetings with administrators and faculty. Finally, most of the responsibilities listed by the CIOs interviewed were technical in nature, although only a fairly small percentage of the people in those positions had a technological background.
We cite this report here because of the similarity between the issues raised in that study and the issues that surfaced during our deliberations. Whether or not the University fills the position with someone having the title of "chief information officer," this person will have to serve that role.
In addressing its charge, the Committee met on 11 occasions for a total of 27 hours. We met with most of those directly involved in one way or another with computing on campus. The following is a partial list of those with whom we consulted, either as a committee or in individual conversations:
UCS Staff (public forum)
Administrative Computer Users: Registrar, Development, C.E., Personnel, etc.
Dean Pinder
Gary Barbour
Pat Urban
Norm Blair
Fred Curran
Bill Ballard
Dean Brandenburg
Jim Kraushaar
Paul Oliaro
Charles Harris
Martin Freeman
Rolfe Stanley
Willy Cats-Baril
Larry Forcier
Al Cassell
Bud Myers
While the committee did not speak directly with people on other campuses, we did obtained some limited amount of information via listservers operating over BITNET. This gave some sense of problems experienced by others with respect to structural issues. We also made use of a listserver of our own, posting summaries of our meetings. Interested parties could read them and respond. (Although the response rate was not particularly high, the community was kept informed.)
Major Issues
A number of important issues arose in our meetings with faculty and administrators. Since these represent problems with which the new person will have to deal, and since some of them may be resolvable when the position is defined, we feel that it is important to include them in this report. They offer a useful way of framing the situation with respect to computing and information technology on this campus. In fact they largely define the structure of computing at the University of Vermont over the next decade. The Committee is well aware that the University is experiencing serious budget problems. We are also painfully aware that the comments below call for spending money that is in very tight supply. To this point we would make two observations. In the first place, we see some expenditures as necessary both from the point of view of efficiency and from the point of view of attracting and retaining the kinds of students and faculty that we need. We will have serious difficulties extricating ourselves from budget problems if we can't retain students, as we have seen as recently as this semester. In the second place, we see the University as being at the beginning of a major planning phase, and if these issues are not raised now, on the grounds that we don't want to think about expensive things, then they will eventually be forced on us when it is too late to plan intelligently. We want these issues to be part of the strategic planning of the University, and we want a director who will be actively and effectively involved in that planning.
In this section we use the generic phrase "Director of Computing" in place of any specific title. It refers to the person we seek, regardless of the level and title of that appointment.
1. Campus Network
Almost everything that we want to do over the next few years will
hinge on the existence of a campus wide network, with smaller local
area networks linking individual colleges or departments.
The Committee cannot stress strongly enough the importance of a campus wide network. It underlies most of the major plans that are currently being formulated. While a complete network is not a prerequisite for implementing proposals with respect to information systems for human resources and student records, its existence would greatly increase the effectiveness of those systems and would give us a much better return on our investment. (To take a second example, the Alumni and Career Development offices share some of the same interests. Unfortunately they do not share computer files. The benefit to be gained by appropriate access to each other's databases should be apparent.) A network is essential for the research needs of the faculty. An immediate example are the efforts to tie together some research labs in the Medical School and the EBMA Computing Facility to allow for the processing of large amounts of information. The increase in interdisciplinary research programs also calls for greater connectivity across the campus. A campus-wide network will also underlie nearly all efforts to bring computing and information databases to our teaching activities. It will make possible such things as electronic discussion sections, faster communication with faculty, and a wider variety of instructional experiences. (For example, Psychology 1 is currently using the new Academic Computing network in Waterman to allow large numbers of students to have hands-on experience with psychological experiments and data collection.) Moreover, a campus-wide network can play a central role in future outreach activities toward the rest of the state. To the extent to which UVM takes its land-grant mission seriously, off-campus access to dispersed campus information services will play a major role. One faculty member recently commented that a network is to the computing of today what a hard disk was to the computing of five years ago. Its importance will rise at a similar rate.
This network will be one of the most expensive things that we do, when you count not only the expense of activiating the broadband cable, but also the greater expense of building local area networks in departments and colleges. This money will be spent one way or another, either through a centrally directed plan or by a myriad of poorly coordinated departmental efforts. We feel that the central administration must see the handwriting on the wall and commit to a campus-wide development of the network. Without this, other efforts will be seriously compromised and a great deal of money may be wasted. One person referred to the network as a "highway of information" and the various local area networks and routers as ramps and access roads. It is probably not too fanciful to think of our needs in terms of an interstate system and to realize the enormous benefits that the U.S. derived when that system was essentially completed in the late 1960s.
In establishing whatever position is decided upon, the University must make a clear designation of where the responsibility lies for coordinating activities with respect to networks. There are certainly differences of opinion over how a future network should be operated, but it is clear that someone is going to have to be assigned the role of cheerleader and guiding light for the installation of the broadband cable, and be given the authority for planning for and coordinating an overall network structure for the campus. This person would also be charged with developing mechanisms to help individual departments and colleges install and operate local area networks. Our proposed structure diagrams the network as a simple box under the management of a director of computing. That plan only designates functions that need to be carried out, it does not designate the structural relations among units at the same level.
2. Distributed Computing
One of the particularly important and hopeful findings that the committee made was the fact that within University Computing there is a serious desire to distribute computing and information to those constituencies that use the data. Thus, for example, the staff of AIS would like to see Personnel, the Registrar, faculty, and students have much greater access to information (within the obvious limits of confidentiality), and would like to work with people to make that possibility a reality. This would shift the efforts of AIS staff away from running the programs and handing the "client" the printout, and toward working with people to make sure that they understand the software and the database and know how to use the resources that are available. This may be a trend that has been going on for some time, but it came as a pleasant surprise to many members of the committee. Both the AIS and the ACS staffs urged us to propose a structure that would allow those activities to continue and to increase. We hasten to note that most of the activities around distributed computing are dependent on a functioning network. If you cannot move data around the campus on a network, it will be difficult to implement those changes that must take place in the next few years.
Related to the issue of distributed computing, we found strong evidence that the several groups on campus who have responsibilities in the general area of computing and information have a sincere desire to work together more productively. Apparently the lines between AIS and ACS are already beginning to blur, and it has even become acceptable to be seen cooperating with someone from Telecommunications. This is the style of interaction that must continue if the University is to make any strides in the area of information technology.
The phrase "distributed computing" also refers to the existence of
various satellite computing centers on campus, whether those centers
be in colleges such as Agriculture, EMBA, and Natural resources, or
whether they be academic or administrative departments. The issue
came up early of the independence of these several organizations. The
more we spoke with people, the clearer it became that these
distributed computing centers must continue to operate on their own
if they are to retain the flexibility and independence that they need
to operate effectively. Perhaps the EMBA Computing Facility is the
best example of this, but the issues are there in the other centers
as well. At the same time, there are going to be situations when the
important needs of these centers for flexibility must be balanced
against the computing needs of the campus as a whole. This might
arise in the assignment of standards for the integration of campus
wide networks. Or it may involve the availability of specific
computer facilities to those users who need them. There may be policy
reasons for restricting access, but there should not be techical
ones. We need an open network. This is not to say that the satellite
units must be brought under the control of some central computer
czar; in fact we think that would be a bad idea. It is simply to
point to an issue that may have to be confronted in the future.
3. Computers in the classroom
The University has not really taken serious steps to incorporate computers into the classroom. This is true in all of the colleges, even those in which students are required to buy a computer with which to prop open their doors. If this university wants to maintain, let alone improve, its academic programs and its stature as a public ivy, it must make major strides in incorporating computers into instruction. By this we do not simply mean that we need to use computers to flash foreign language words on the screen. We mean that we have to become aware of, users of, and perhaps producers of first rate educational software. In addition we have to work with our students to make them competent members of the "information generation." Within 5 to 10 years most students will have virtually unlimited access to unlimited amounts of information. What are they going to do with all of that stuff? We have to adapt our teaching to that situation, and we have to make use of our computing resources to make that happen. If we fail to do this, we will be failing in our obligations to our students. We are not going to be able to attract the kinds of students that we seek if we cannot offer them something better than they had in high school. We would hazard the guess that computers play a larger role in the classrooms at Colchester, CVU, and South Burlington High Schools than they do at the University of Vermont. Students with whom we spoke assured us that their high schools certainly made more use of computers than do we. We can hardly be proud of our record.
We see the new director of computing as playing a proactive role in making all of this happen. If we go on past experience alone, it is unlikely that faculty and students will develop what we need on their own. There are just too many competing demands on people's time. We need a leader who will encourage faculty, who has ideas and the desire to change current practice, and who will be an articulate spokesperson for computing on campus. We may very well have to move toward something equivalent to a computer-focused Instructional Software Center. Many universities have such a facility, and it is an important component of the academic environment.
4. Training/education
A common thread that ran through many of our discussion with the
computer-users on campus was the clear need for a greatly increased
training activity on campus. As we distribute computing to the
university community and see a higher level of computer functioning
on campus, the need for adequate training of faculty and staff
becomes critical. Such activities will not be cheap, but they are
essential. It is difficult to imagine any large American corporation
doing as little training of their staff as does the University of
Vermont. And we are basically an academic institution that presumably
knows the value, for other people, of education. It is not that we
have not been offering training for faculty and staff--the
minicourses offered by ACS are often excellent. It is simply that we
are going to have to do a great deal more.
Lest it appear that we are speaking here solely of non-academic
staff, let us hasten to point out that the faculty as a group is
probably no more computer-literate than the staff. There is no way
that we can take advantage in the current revolution in educational
software unless we know what is going on and have a nodding
acquaintance with computing. This takes us back to the comments in
the previous section about the importance of developing, as resources
permit, a training facility with respect to educational software.
5. Fostering a Cooperative Environment
In an earlier section we pointed to the sincere efforts of AIS and
ACS to work together. We have also seen evidence that
Telecommunications has been working well with those units and with
others, and we see that as encouraging. The only point that we would
make along these lines is to say that whatever structure is finally
adopted, it will be important to ensure that the University's policy
is that there must be cooperation among those, and related, units,
and that someone, at some level, will be charged with the
responsibility of seeing that cooperation exists.
In a similar vein, the three organizations mentioned in the preceding paragraph really serve a service function with respect to the Libraries, Media Services, administrative offices, and so on. We have been struck by how well they are filling that role, and only point to the need to make sure that such an attitude toward service continues.
6. Standards
We have heard a great deal about the lack of agreed upon standards on campus when it comes to computing. This is a particular issue with respect to networking, where computers and software have to function together or the system will just not work. The director of computing must take the lead in establishing standards. That does not mean dictating what hardware or software people may buy, but it does mean overseeing a process to decide what the University will support and making it clear that if someone buys software or hardware that we don't support, then they are on their own when their network won't function properly or won't mesh with the broadband cable.
7. Reporting issues
We heard testimony from several sources to the effect that reporting lines have long been an issue with respect to computing. In the past, UCS has had responsibilities with respect to both the Academic and the Administrative sides of the University, but has reported only to the academic side--either the Academic Vice President or the Provost. There has been the feeling, justified or not, that the academic side has come out on top. However to the extent that this issue has created a problem, and the committee is not here to judge that issue, it is incumbent upon the administration to clarify the situation and come to some working arrangement. If there is a problem, now is the opportune time to remedy it. Several members of the committee who have had some experience with dual reporting lines firmly believe that such a structure can not only work, but is desireable.
8. Planning
As far as we are aware, there is currently no long range plan for computing on campus. Many people have ideas of what they want it to be, but there does not seem to be anything concrete. We urge that whoever is placed in charge of computing and information technology have as her/his primary task the development of a campus plan for computing. We are currently without most of the permanent senior level officers of administration, and without a plan for computing it is difficult to know what the University's priorities are likely to be in two or three years. This makes a clear public plan a matter of even higher priority.
A plan for the future of information technology on this campus cannot be accomplished by one person. We feel that it is essential that those most intimately involved with the use of computing and information be brought together to develop such a plan, under the direction of the new Director. We are not thinking of "one more committee" that will meet monthly to talk among themselves, but of a committee with a specific charge, a definite, and short, deadline, and one composed of people with serious involvment in the implementation of that plan. If we need a model, the Broadband Cable Committee is a good example.
It would not be correct to suggest that no planning has taken place. We are aware of planning currently being done within adminstrative offices, including personnel, payroll, development, and others. We know of departments that have carefully planned for the future development of computer technologies. And we know of planning currently ongoing within colleges. What we feel is lacking is more widespread planning at the level of the individual units and the existence of an overall campus plan into which to fit the more circumscribed plans.
9. Funding
There are several things to be highlighted with respect to funding, but the committee is not sufficiently arrogant to offer Solomonesque recommendations.
In the first place, there is a need for long range budget planning with respect to computing, just as there is such a need with respect to most areas of the campus. Computing is a very expensive activity, and that situation is not likely to change in the near future. Probably the most important challenge facing computing in the next five years is networking, since most of what we want to do as a university hinges on the availability of a campus-wide network. But beyond that the person that we seek needs to establish a stable funding mechanism for the things that we need so that we can plan our development accordingly. We need to establish dependable sources of funding, which very likely will include development work with hardware and software companies, as well as finding sources of support for incorporating computers and information sources into our classroom teaching. We would be so bold as to suggest that it might be sensible for a director to develop ways of bringing together the resources of other College, departmental, or administrative unit budgets to accomplish a common goal.
The Committee considered a number of possible models for structuring
computing on campus. We discussed these models with those who spoke
with us, and found a general level of agreement on what was both
desirable and practical. Before laying out the various alternatives,
we should point to the fact that nearly everyone (but not all) agreed
that we need a dynamic leader who will:
Finally, although we have listed many functions that this one
position is to serve, we do not see this person as "micro-managing"
the units under her/his direction. We already have quite competent
people who are capable of managing their own units. We want to see
someone who will do considerable planning, who will create an
atmosphere in which people and units will work together, and who will
be an articulate spokesperson for the issues raised above. We think
that it is a task that one person can do. We do not think that we are
piling too many responsibilities on that person's shoulders, although
the list above may give that impression.
The following are the four structural models that we seriously considered. At one time or another the committee leaned substantially toward each of them, but the weaknesses of three of them lead to their elimination.
Model 1.
Hire a Vice President for Computing and Information Technologies who
reports directly to the president and has responsibilities for the
current University Computing, Telecommunications, and Networking.
(Everyone seems in agreement that Media Services has sufficient
responsibilities in unrelated areas that it should not be include
here.)
The major appeal of this model is the fact that it would represent a serious commitment to computing and information technologies on the part of the University. It would also provide a strong voice for the needs of that area at a level where that voice would have to be heard. This would also be a position that could provide perhaps the best guidance in terms of long range planning, because the planning would be coordinated by one of the most senior members of the administration.
There is one major drawback to this model. In the eyes of many with whom we spoke, including deans, administrative officers, faculty, and staff, hiring "yet another vice president" at this time would not be politically astute, to put it mildly. A second drawback at this particular time is our sense that any administration would be extremely reluctant to hire such a person at least until we have identified the next provost, and perhaps the next senior vice president. We would be extremely reluctant to see the filling of this position delayed for an additional year. Finally, we think that we can accomplish much, if not all, of what we want to accomplish by filling that position at a lower level.
Model 2
Hire a Director of Computing and Information Technologies who reports to one or more vice presidents (or the Provost) and has responsibilities for the current University Computing, Telecommunications, and Networking. This is one step down from the previous model, differing only in the reporting line for the position. (As noted earlier, reporting lines need to be thought out carefully, and serious attention should be given to the possibility of dual reporting.)
There are several advantages to this model. In the first place it is clearly more practical and politically acceptable than Model 1. At the same time it would also pull together the major areas of interest, allowing for greater cooperation among the important organizations. With broad responsibilities and an effective reporting structure, we would anticipate that this director could accomplish all of the goals that we have set out earlier.
One additional advantage of filling the position at this level is that it puts the person much closer to the staff with whom he or she is working. This should lead to a closer day-to-day coordination of the organization. We see long-range planning as being one of the major responsibilities of the job, and an argument can be made that planning can be carried out more effectively when the individual is closer to the activities for which planning is being done.
The major disadvantage of this model over Model 1 is that the director would not have quite as powerful a platform for arguing for resources and future direction. This is an important issue, and by putting the position at this level we would be banking more heavily on the enthusiasm (and right-mindedness) of the Provost.
Model 3
This model is the same as Model 2 except that the director has responsibilities only for the current University Computing (and presumably some responsibilities for networking, but that hasn't been spelled out).
The committee could think of no advantages for this model over either of the two previous models, but could think of several disadvantages. In the first place it would leave University Computing much as it has been ever since ACS was brought back to Waterman. The director could speak only for University Computing (and networking), and would not be in a position to develop long range plans for all information activities on campus. A good director might give us a good computing center, but we need more.
Model 4
The final model, to which we gave serious consideration, was to hire no one, and to leave the system as it is. If we could sort out reporting lines and the budget, with some work it could be made to work. Under this model University Computing would continue to function much as it has for the past several months.
One advantage of this model is that it would allow us some breathing space to see the direction that will be set by the new provost. We could then determine where the University was going and either fill the position at that time or maintain the status quo. The second advantage is that such a move would save a considerable amount of money, which might then be plowed back into computing.
This model has several disadvantages. In the first place the University would be stalling for yet another year on where it is going with computing. By that time we would be even further behind. If we are going to maintain, and even improve, our position as a first rate educational institution, we need to do something now.
The second disadvantage of this model is that it leaves coordination and cooperation up to the good will of the individuals involved. Although the people currently in leadership positions at University Computing are doing a good job and have made some very serious efforts at increased cooperation, we feel nervous about a structure that depends so importantly on the particular people who happen to be filling the positions at any particular time.
Finally, maintaining the status quo leaves us without anyone in the position to be a strong advocate for computing and information technologies. That statement is in no way a criticism of the people who currently fill those positions, who also happen to be on the committee, but simply a statement that for truly effective advocacy we feel that we need someone at a higher administrative level.
Conclusion
The Committee recommends the adoption of Model 2, which calls for a Director of Computing and Information Technology with clear lines of responsibility and reporting. We further urge very serious consideration of a dual-reporting structure, because this person really does have responsibilities on both sides of the University. We have appended a possible organizational structure, but would hasten to add that our diagram speaks to function more than to structure. For structural reasons it might make sense to combine Networks and Telecommunication, or to think about the relation between Users Services and Academic and Administrative Computing. Those are decisions for which we have no recommendation. They are, in fact, probably decisions that should be postponed until after we have recruited a director who will be working with the structure.
It is not too extreme to say that an important part of the future of this university depends upon how we blend information, data communications, computing, administrative functions, and, most importantly, teaching. We are well behind many institutions with which we are familiar, and with which UVM compares itself, and we must make important decisions now as to where we are going and how we will get there. The person that we seek to fill this position will have an important role to play in those decisions and their implementation.
Fred Dicesare
Fred Evering
David Howell (Chair)
Roger Lawson
Sue Mackin
Rebecca Martin
Joe Patlak
Bob Thigpen