The Administrator as Change Agent

Leading change, implicitly or explicitly, is an element of the portfolio of every senior administrator. You might be leading an effort to increase the diversity of the first-year class, better prepare junior faculty members to achieve tenure, reduce expenses associated with some process while maintaining quality, improve the culture of respect and inclusion on campus, or reduce the harm associated with students’ use of alcohol.

That last one was mine. As vice president for student life at the University of Iowa, one of my priorities was to reduce the harm our students experienced as a result of their use of alcohol. In the course of that work, and other work, I came to believe that three conditions are necessary and sufficient for the success of such efforts: proven practices, metrics that stakeholders embrace, and commitment.

The framework suggested by these three necessary and sufficient conditions can help answer some crucial questions, including: Should I pursue this change effort? What interventions should we use? Why isn’t our effort succeeding? How should I describe to our stakeholders what we’re doing?

I’ll describe the conditions, with examples from our alcohol harm reduction efforts, and suggest a few implications for your consideration.

Proven Practices

What constitutes a proven practice may vary from context to context, but I encourage rigorous consideration before you assign that label. In the alcohol harm reduction domain, practices supported by multiple published, peer-reviewed, longitudinal, randomized clinical trials might epitomize proven practices, and a good number of such studies exist.

For instance, there is a very substantial scientific literature about alcohol harm reduction among college students, and as a bonus, an expert panel evaluation of the literature. The College AIM, an authoritative guide to the effectiveness and cost (among other characteristics) of individual and environmental interventions, gave us at the University of Iowa a lot of confidence as we selected proven practices to include as tactics in our alcohol harm reduction plan. It also gave us confidence in excluding approaches for which little evidence of effectiveness, or evidence of ineffectiveness, exists.

Those kind of proven practices anchor one end of a continuum of confidence in the effects of any practice you are considering. Farther down the continuum we might find “promising” practices. Often administrators are drawn to interventions labeled (usually misleadingly) “best practices.” Such interventions have typically been used at only one or a few institutions and have been evaluated only informally.

Even farther down the continuum, you might find “plausible” practices. They are practices that seem like they should work, based on relevant theory and experience, but that haven’t been tried or have been tried but not evaluated.

You might choose different labels for points along the continuum. What is important is to recognize that, for many change efforts, you can find a menu of practices, ranging from those you have very good reason to expect to succeed to those you have less reason to expect to succeed. In some, probably many, cases, there are no genuine proven practices. You might consider going ahead anyway. Implementing a change effort based on promising or plausible practices carries with it two implications.

First, consider opportunity cost. Given that you have limited resources, does it make sense to use those resources to try to solve a problem that no one knows how to address effectively? In my setting, we recognized marijuana as a public health problem on the campus but could not identify significant proven population-level interventions beyond the ones we were using (mainly enforcement). So we chose not to devote more resources to an effort to reduce marijuana use.

“Resources” probably brings to your mind money and time, and those are certainly important resources. But another very important resource is limited, as well: you and your leadership team can only really attend to a relatively small number of change efforts at one time. It may be best to direct your attention to those for which you can identify proven practices.

Second, if you do choose to address a problem for which you can find no proven practices, be clear about what you are doing. You might better conceive of your efforts as research and development work rather than as leading a change. Research and development efforts call for staffing, structure and strategies that differ from those for change efforts.

For instance, you will probably not extend a research and development effort to the entire population of interest, given that you do not have strong evidence that it will be successful. Additionally, you might want to test multiple interventions independently to understand the effects of each. And you would probably evaluate the success of a research and development effort on the basis of the knowledge gained as opposed to the change achieved. Finally, you would manage stakeholder expectations differently for a research and development effort than you would for a change effort.

Metrics that Stakeholders Embrace

If you are going to invest in a change effort, it seems irresponsible not to have a way to know if it is working. That involves collecting evidence, usually over time. Most commonly, the evidence involves some numbers -- and probably some statistical analysis of those numbers.

But what numbers? The choice is crucial, since your change effort will be designed and adjusted to move those numbers. In our case, we chose three metrics to monitor. Using the National College Health Assessment, we watched the percentage of students engaged in high-risk drinking in the two weeks before the survey was administered, the mean number of drinks consumed and the percentage of students who had consumed alcohol on at least 10 days out of the 30 days preceding the survey.

When we published our first alcohol harm reduction plan, students, faculty members, community members and other stakeholders were skeptical. They had questions about the representativeness of the students who completed the survey and the veracity of their responses. We had considered carefully those issues, and over the course of time and with deliberate efforts, we were able to explain our considerations well enough to establish trust in our choice of metrics among our stakeholders. In many circumstances, it would have been better to start by involving our stakeholders in the development of the metrics. In our case, I was confident in our ability to explain our choice and felt an urgency to address the problem on our campus.

Whether you educate them about the metrics or have them help determine those metrics, the involvement of stakeholders is important for two reasons. The first is rather obvious. Change efforts in complex organizations such as institutions of higher education require the participation of many people on the campus and often those in the community beyond it. They are unlikely to support your change effort if they don’t trust the metrics you have chosen to evaluate the effort.

The second reason for involving stakeholders in the definition of metrics is simply to get them right. In our case, for example, students understood the realities of the campus drinking culture in ways that our staff members didn’t. If they had not found the metrics we chose meaningful, we would have had a duty to reconsider those metrics. The fact that our metrics didn’t capture the problem in a way that made sense to students would have been a strong sign that we hadn’t adequately conceptualized what we were trying to change. The exercise of defining metrics forces clarity in the definition of the change we seek. Here’s another way to think about it. If you can’t develop a shared understanding of what your change effort is intended to accomplish, you might not understand your goals well enough to design your change effort.

Commitment

Imagine that you have launched a change effort, know precisely what you are trying to accomplish (because of the process you have used to define metrics) and have identified proven practices to use.

What could keep your change effort from succeeding? All the answers fall under a single rubric I call “commitment.” Whether that’s your individual commitment or your institution’s commitment will vary from context to context. In what follows, you might take “you” either way. Keep in mind, though, that effective leaders take responsibility for the success of the work they lead.

Markers of your commitment to a particular change effort can take many forms. In the case of our alcohol harm reduction effort, we demonstrated our commitment by publishing our plan, assigning a budget to the effort, hiring a staff member to lead the effort and creating a governance structure for it. In the absence of those kinds of signs of commitment, your stakeholders might understandably begin to doubt your commitment to change. At the extreme, for example, if your effort has no plan, no budget and no leadership, stakeholders can legitimately conclude that your announced change effort is little more than window dressing.

Put another way, if you know what change you want to create (i.e., you have defined metrics), and proven practices exist to effect that change, but things aren’t changing, you have no excuses. If your explanation for the failure of a change effort is that the effort was underfunded, you might be saying that you weren’t really committed to the effort, at least relative to all the other things in your budget. If your explanation is that the person in charge of the change effort wasn’t up to the job, you might be saying that you weren’t adequately committed to the selection, training and supervision of that person. If your explanation is that priorities shifted, you might be saying that your commitment wavered. In short, if you knew what change you wanted to make and how to make it, but the change didn’t happen, you are going to have to take responsibility.

So there we have it. Three simple necessary and sufficient conditions for the success of a change effort. Perhaps a bit too simple. If you and I sat down over a cup of coffee, I suspect we could find a number of ways I have oversimplified. I would be willing to go first.

Still, I don’t think that the simplification invalidates a basic truth. If you want to make change, you will need practices that work, metrics that are meaningful to your stakeholders and a good dose of commitment.