N.B. You might want to grab a libation before reading. This is a long post.
What does a young researcher do when he is suddenly thrust into leading a top-tier computer science department? He learns to fix problems before others see them, outwit university bureaucracy, make decisions that can change people’s lives, and protect the research career that brought him there in the first place.
It was a steep learning curve.
The year was 1996, and I was the youngest full professor in the University of Illinois Department of Computer Science, then ranked sixth in the United States. I was 38 years old with a large and flourishing research group (the Pablo project) in high-performance computing. My full professor colleagues were at least a decade older than me, and a few were old enough to have been my parents.
Illinois Context
Why was I in the Illinois cornfields? Because in computing, there were few places like it in the world.
From the early ILLIAC computers, culminating with the groundbreaking ILLIAC IV supercomputer, through the seminal PLATO computer-aided instruction system, a computer-aided proof of the four-color theorem, and deep contributions to computer architecture, semiconductors, and software, Illinois had been a key architect of the digital revolution. (See Stored Program Computing: Ideas and Individuals Do Change the World.) It was home to not one, but two supercomputing centers – the National Center for Supercomputing Applications (NCSA), which I later led, and the Center for Supercomputing Research and Development (CSRD), where I had a research appointment.
The campus believed, with some justification, that it was not aspiring to excellence, but it was the excellence to which others aspired. I was proud to be a part of that climate and culture.
To say that I was surprised when one of my senior colleagues suggested that I become the next department head would have been a classic understatement. I had absolutely no doubt there were wiser, more experienced, and better qualified hands among my more senior colleagues. To my shock, the Dean of the College of Engineering, Bill Schowalter, agreed with my colleagues. With substantial trepidation, I agreed to take on the role.
At the time, College of Engineering department heads were exempted from teaching, and they received a $1,500 salary augmentation, which I learned was called the “necktie allowance.” It reflected the then reality that all the college’s department heads were male.
Computer scientists are not known for their sartorial style, and I was no exception. I did keep an “emergency tie,” in Illinois orange and blue, hanging behind my office door, donned to meet an unexpected visitor from outside campus. Of course, it rarely matched my patterned shirts and khakis.
Little did I know I would spend my later years, both as a science policy advocate and as a university administrator, wearing a suit every day. (See Geek Versus Chic.)
Role Separation and Research Sustenance
At the time, my mentor, the late Ken Kennedy, my friend Rick Stevens, and I were helping the NCSA Director, Larry Smarr, build the NSF PACI National Computational Science Alliance, a national consortium of computational and computer scientists. When I told Ken I was becoming department head, he laughed and said, “It’s been nice knowing you.” The implication was clear; he assumed I would be consumed by administration to the great detriment of my research career. I was determined to prove him wrong.
With the Dean of Engineering’s tacit approval, I instituted some ground rules to balance my competing lives of research and administration. The first was maintaining two disjoint spaces: my faculty research office and the department head’s office in the administrative suite. I would spend two days a week in the research office and three in the administrative office.
On research days, I would not undertake administrative tasks unless it was an emergency, and I was the primary arbiter of emergencies. I would conduct research on administrative days if there were not enough administrative tasks to fill the day. Of course, there were exceptions, particularly during faculty recruiting season, when candidates were visiting campus.
After adjustments, this division of labor proved effective, both for me and for the department. Nevertheless, I had to learn a lot quickly, particularly with the personnel and budget processes I had heretofore largely ignored. I made mistakes, some I still regret, but I did the best I could.
As I now look back thirty years to my days as a young department head, my memories fall into three groups. The first is the humorous and mundane, the minutiae that any seasoned academic leader would recognize. The second is the faculty and personnel issues that trigger difficult choices and sleepless nights. The third is the consequential decisions that have long-term, strategic implications for the unit. Amid the fray, one must be ever mindful of these distinctions.
Academia Minutiae: Act One
Academic leadership is occasionally depicted as translating vision into strategy and tactics, but reality is grounded in an ongoing stream of minutiae. These are the issues that never appear in a job description but consume one’s daily time and energy. Managing them requires a mix of bureaucratic legerdemain, good humor, and recognition that small problems left unsolved can fester and infect organizational culture. Herewith are a few of my more memorable ones.
Clean Windows at Last
In fine computer science fashion, the old Digital Computer Laboratory (DCL), then home to the Illinois Department of Computer Science, had undergone two binary expansions since its original construction in 1958. The initial construction and the first expansion were the birthplace of the famous ILLIAC computers.
The second expansion, to address the growing space needs of the department, was a three-story wraparound extension, with black, cantilevered walkways overlooking an atrium created between the older and newer parts of the building. It was an architectural compromise, designed to create the maximum possible new space at the least possible cost. I later tried to hide some of the ugliness with colorful banners, to negligible effect.
Alas, the prime contractor absconded as soon as the building expansion was complete, failing to clean the exterior windows and leaving construction detritus to obscure the views. Two previous department heads had tried and failed to get the windows cleaned. As the new kid on the block, I was determined not to be the third.
I first called the campus Facilities and Services group and asked if I could schedule a window cleaning for our building. They happily agreed to put us on the schedule for a mere $5,000. When I innocently inquired about the date, they responded that the cleaning was scheduled early in the new millennium (i.e., at least five years in the future).
Sadly, I was not surprised. If one looked carefully at the entrances to the engineering buildings, there was usually a small, spray-painted stencil just inside the door. It was the date the building had last been painted, often ten or even fifteen years in the past. It was a gross understatement to say the campus infrastructure suffered from a backlog of deferred maintenance.
As I was later to learn in senior roles at the University of North Carolina, the University of Iowa, and the University of Utah, deferred maintenance was the academic norm rather than the exception. There is no large and ready pool of donors eager to support remedial construction, and states routinely underfund university requests for the same.
Still young and naïve, but not easily defeated, I opened the phonebook – this was decades ago – and called a local window washer. I asked if they could clean the windows of DCL. They said, “Yes, of course.” When I asked when they could schedule it and how much it would cost, the answer was a pleasing “Next week and $1,500.”
Fully aware of the peril of circumventing the campus bureaucracy, I called the Facilities and Services group again and reported the alacrity and power of private enterprise. After being placed on an interminable hold, the scheduler finally responded by saying that an opening had miraculously appeared. They could wash the windows next month and match the private sector price.
A few weeks later, we held one of our regular departmental faculty meetings. As it happened, the window washers were hard at work just outside the conference room where we were meeting. I paused the discussion to note that unlike my predecessors, I had succeeded in getting our windows cleaned. To my surprise, the faculty rose as one and gave me a standing ovation, the only one I would receive in thirty years of administrative service.
Fixing Invisible Problems
I was reading email in my third-floor departmental office one day, when there was a knock on my door. I looked up to see one of the department’s recent faculty hires in the doorway. I motioned toward a chair and came around my desk to join him. After a few pleasantries, he bluntly asked me, “So, what is it you do up here?”
Initially nonplussed by the temerity of the question, my immediate reaction was to respond in kind. Instead, I reflected for a moment and then replied,
My job is to fix problems before you even know they exist, allowing you to focus on building your career. If you and your colleagues are discussing what I am doing, then it means I have failed to address an issue successfully. It’s a good thing, particularly as a young professor, that you do not know what I am doing here.
The reality behind my response was simple. The role of a young professor at a top-tier research university is, in priority order, to build a research career, to do a decent job of teaching, and to engage in some modest amount of disciplinary service, such as serving on high profile program committees. That teaching is relegated to second place is a sad part of today’s academic reality. Teaching matters, but an otherwise excellent teaching record will not compensate for a middling research portfolio. (See Research, Teaching and Service: Dynamic Balance.)
Death By Memoranda
In the “old days” before universities embraced mass email to distribute policies and notices, most memoranda were delivered to departments by the campus mail service. Except for mandatory campus messages, department heads then exercised their discretion about which of the memoranda were worthy of photocopying and further distribution. Not surprisingly, many were of little importance, unworthy even of paper airplane status, and were relegated immediately to the trashcan. (Yes, all of us in computer science had long had email addresses, but many on campus still did not. Paper was still the coin of the realm.)
At one faculty meeting, a faculty member objected to the “blizzard of paper” filling their departmental mailbox. As the memorandum arbiter, I assured them I was a 99% filter; they were getting only “the good stuff.” That ended the discussion, but it was obvious my colleagues were not fully mollified.
Sometime later, a new campus memorandum arrived, explaining that it was time for the periodic renegotiation of the “pouring rights” contract for campus vending machines. In the days before bottled water and energy drinks, this was a Coke versus Pepsi distribution battle. The memorandum noted that there would be a series of public meetings where faculty, staff, and students could express their perspectives and opinions on the choice.
As mundane as the issue may seem, the choice did have some practical import. As part of the contract negotiations, the competing vendors were asked to promise a percentage of their revenue to support student scholarships. Beyond the fierce loyalties dividing Coke and Pepsi aficionados, the scholarship revenue was the competitive battlefront.
Recognizing my opportunity was at hand, I began crafting a new memorandum to accompany the campus missive. First, I acknowledged that faculty time was a precious commodity and that I was working assiduously to protect it. Hence, it was only the criticality of this campus issue that forced me to intrude on their valuable time. I continued by reemphasizing my staunch support for shared governance, the centrality of participatory democracy to our mutual success, and the institutional importance of debating one of the great issues of our time. I urged the faculty not to miss this opportunity to be heard. The result was a page of satirical, hyperbolic prose.
With quiet glee, I then directed the departmental staff to staple my satire to the campus memorandum and place both in the faculty mailboxes. Within an hour, I began to get humorous email messages from the faculty, and there were no more complaints about overflowing mailboxes.
Reed May Be a Flight Risk
Two years into my five year term, I had one of my periodic meetings with the Dean of Engineering. During the meeting, he asked me, as he often did, about the state of the department. I reported that, based on all I could see, things were going well, but I was worried about one of my colleagues. They were being wooed by other institutions and were a potential flight risk. My sources suggested this faculty member was being pursued as a potential dean or provost, but I did not think they wanted to leave Illinois.
The dean asked the name of the faculty member. I said, “It’s Reed. If he leaves, I’m screwed.” He laughed and asked what I thought would be needed to retain Reed. With a smile, I replied, “It’s not money; he is well-funded, but he seems overwhelmed. He needs time, which means offloading some work.”
The dean understood something I had not fully internalized. The problem was not Reed. The problem was me, the overcommitted department head, who needed to delegate more. A few days later, the Dean quietly transferred a recurring salary line, which allowed me to hire a full-time communications director and focus more of my time on departmental strategy and less on day-to-day communication tactics.
Academic Tenure and Personnel Management: Act Two
If managing daily minutiae tests a department head’s patience, personnel decisions test their courage. At best, you learn things about your colleagues you would prefer not to know. Meanwhile, you are seeking input while making life-changing decisions about faculty and staff – who joins, stays, or leaves the community – to safeguard the institution’s future.
Tenure Standards and Excellence
Not long after I became department head, the full professors, who functioned as the department’s tenure and promotion committee, met to consider the case of an assistant professor. Like a Ph.D. qualifying examination, it is an “up or out” decision. Either the candidate is granted tenure and promotion to associate professor, with putative lifetime employment, or they are asked to leave the institution. (See Research: When There Are No Answers in the Back of the Book.)
The department took immense pride in the fact that its departmental tenure and promotion decisions had never been overturned by either the college or campus processes. We had high standards, and we expected our colleagues, ourselves included, to meet a high bar.
Although I was new to the game, I believed this candidate’s portfolio was marginal. The teaching metrics were adequate, and the research portfolio was not without merit, but neither was it outstanding. The deadlocked tenure vote after ninety minutes of lively discussion confirmed my assessment. As a full professor, I could have voted to break the tie, but by tradition, the department head served ex officio, because the head was responsible for managing the candidate’s portfolio and guiding it through campus review.
Rather than prolonging an increasingly circular conversation, I suggested that we meet again in two days after a period of reflection. Alas, the subsequent discussion resulted in a second deadlock, albeit with two individuals shifting their positions in opposite directions. Finally, one of the most senior faculty members looked at me and noted that as the department head, it was my job to make decisions. I had to break the tie, much as the Vice President breaks ties in the U.S. Senate.
I did, voting against tenure. It was the first consequential decision in my young administrative career, and it was personally painful, because I liked the candidate. However, marginal was simply not good enough in a top-tier department.
I was then left with the painful duty to relay the committee’s negative decision. It was a difficult conversation, one marked by tears, as I explained, as kindly as possible, that the candidate no longer had a future at Illinois. Yes, there would be a year of paid transition, but it should be used to seek a new position, either at another university or in industry. Afterward, I went home, opened a bottle of wine, and spent the evening staring at the wall.
The candidate appealed to the provost, but the department’s decision was upheld. The gap between the weakness of the candidate’s portfolio and the reputational bar was simply too great.
Decades later, as a provost, I related the same rationale to deans when they were grappling with similar challenges. One must put aside personal feelings. It is about building and sustaining organizational excellence. The key objective is simple: leaving the organization stronger than you found it, while never forgetting that people are not chess pieces, but individuals deserving of your respect and kindness.
Staff Are Unsung Heroes
In a university, the faculty often believe themselves the center of the academic universe. It is not wise to challenge that illusion. In truth, staff are the backbone of the university. Every process – admissions, student advising, classroom management, human resources, payroll and budgets, research infrastructure, and yes, even the sometimes-maligned campus facilities support – depends on hardworking and diligent staff.
One of my favorite interview questions, “Tell me about the life of the person who empties your trash,” captures that ethos. Anyone who stammers or responds with a blank stare may treat the staff as either second class citizens or as interchangeable building blocks. Everyone has a role to play.
As department head, I was blessed with a cadre of tremendously dedicated and talented staff. I had two wonderful administrative assistants, Anda Ohlsson and Barb Armstrong, who managed my often chaotic schedule. My business manager, Pat Patterson, who later became the campus CFO, went far beyond the call of duty, building financial projections and what-if scenarios that allowed me to focus on strategy. The student office was masterful in fixing intellectual and emotional skinned knees, and the IT staff deftly managed the research and educational demands of faculty and students. Most importantly, my associate head, Bill Kubitz, was a fount of wisdom for a wet-behind-the-ears department head and an expert at navigating university administrative processes.
There were, of course, unsettling hiccups. One day I received a call from the College of Engineering regarding the proposed salary increase for the senior staff member in my research group. With no preamble, the caller asked, “Do you realize you are proposing to pay this person more than many assistant professors earn?” Without hesitation, I replied,
Yes, and I’d like to pay her even more. In my absence, she manages a mix of forty staff, post-doctoral associates, graduate students, and undergraduates. She is the anchor of my research group.
I was angry for two reasons. The first was that she was a woman working in a male-dominated culture, and I was determined not to let that culture undervalue her contribution. The second was the implication that a staff member was less valuable than faculty.
I ended by saying, “Please don’t ever call me again with a question like this.”
Consequential Outcomes: Act Three
Beyond the daily minutiae and the burden of personnel decisions lie the inflection points that alter an institution’s trajectory. Some are about facilities that attract faculty and students. Others are about the talent that sustains and expands reputation.
But It’s a Prison
My indefatigable alumni coordinator, Judy Tolliver, reached out to one of the department’s successful alumni, Tom Siebel, and invited him to campus to be the inaugural speaker in our entrepreneurial series. The students loved it, and so did Tom. To quote the movie Casablanca, it was the beginning of a beautiful friendship.
While sitting in my office, Tom and I discussed his career, first at Oracle and then as the founder of Siebel Systems, and how his Illinois education had been the springboard for this professional success. During that conversation, Tom looked at the latest expansion of DCL and remarked that its gray and black cantilevered balconies looked like a prison.
Tom was right; I had often imagined armed guards walking the halls, and I knew I was not alone in that visualization. I gently suggested to Tom that he could help us fix that.
That initial conversation—followed by a series of dinners, golf outings, and deep discussions—led to Tom’s $32 million gift. With additional funding from the State of Illinois, it created a new $75M home for the department: the Thomas M. Siebel Center for Computer Science. The Siebel Center, together with the new National Center for Supercomputing Applications (NCSA) building, formed a $100M IT quadrangle for advanced computing research and education. Twenty-five years later, it remains an anchor for faculty and student talent in one of the world’s great computer science departments.
And Still They Came
My tenure as department head overlapped with the dot.com boom. That was both a blessing – we had everyone’s attention – and a curse – everyone wanted a piece of the action.
Already a large department, the explosion of student interest during that period almost overwhelmed us. Fearing political and public backlash, the campus would not allow the department to cap enrollment, and every year even more students applied and were accepted. After considerable negotiation, the campus granted us the right to set an academic admissions threshold distinct from the college. Each year we raised the threshold, and each year more students applied and ever larger numbers exceeded that threshold.
With only forty faculty, we had nearly 1500 students studying computer science, of which roughly 400 were graduate students. In addition, we were teaching large service courses (introductory computing) for the College of Engineering and the College of Liberal Arts and Sciences. Not surprisingly, our teaching loads, as measured by student/faculty ratios, were among the highest, if not the highest, on campus.
Just before I became department head, my Hennessy-Patterson advanced computer architecture course, with a mix of undergraduates and graduate students, regularly had 150-175 students. Even graduate classes had enrollments of forty or more. Quite clearly, this was not a sustainable situation.
Strategic Faculty Hiring
Faced with a still growing student body and an exhausted faculty, departmental survival required more than incremental growth. We needed a step-function increase in faculty size, both to support the educational mission and to ensure Illinois remained a global leader in computing research. I told the faculty to grab something and hang on, because we were going to pull into the passing lane and put the hammer down.
I authored essays about the centrality of computing to the university mission. I gave speeches around the state. I pestered the dean, and I advocated with the provost.
I also made the mistake of telling the dean that in the 21st century there would be only two kinds of engineering departments, those that studied computing and those that applied computing. Time may have proven the truth of that statement, because if anything, I underestimated the societal impact of computing. It may have been true, but it alienated some individuals whose goodwill I needed. It was a well-deserved lesson in the importance of humility.
I sought a dramatic expansion of the faculty and associated staff, seeking at least twenty-five new faculty lines. The campus, in the form of a provost, Richard Herman, believed in me and a vision of what could be. With his political and financial support, we now had gas in the tank, and we were on the speedway to the future.
When I stepped down as department head to focus on leading the National Center for Supercomputing Applications (NCSA) and building the NSF TeraGrid, the combination of the new Siebel Center and strategic hiring had lifted Illinois past Cornell to number five in the national computer science graduate U.S. News rankings. Since then, the department has continued to grow, and thanks to the continuing generosity of Tom Siebel, recently became the Siebel School of Computing and Data Science.
Concluding Thoughts
My friend Ken Kennedy laughed when I told him I was becoming a young department head. His “It’s been nice knowing you” comment was sobering and apt, and he was right to be worried.
Administration did change my career. It took me from Illinois department head during the dot.com boom to director of NCSA, founder of the Renaissance Computing Institute (RENCI), a corporate vice president at Microsoft, university leadership, and national science policy roles.
However, it did not end my research life.
More importantly, it taught me something I have carried into every leadership role since – the best leaders are often doing work nobody notices. They remove obstacles. They protect talent. They make the hard calls – and sometimes they make mistakes. In both cases, they feel the personal weight and pain those decisions entail. And, when they are fortunate, they leave behind an institution that is stronger than the one they inherited.
I was extraordinarily fortunate to have learned those and other lessons at Illinois. For a poor kid from the Arkansas Ozarks, I have been blessed.
Please leave a comment …