How can a university system of over one-half million students, faculty, and staff, spread over 23 colleges or campuses, each led by its own president or dean, work toward similar goals, in a coordinated way, showing continuous improvement? This daunting question faced Matthew Goldstein when he became chancellor of the City University of New York in 1999. Until that time, although there were certainly policies that applied to everyone in the university, and goals had been given to individual campuses, the goals for the university had not been clearly stated and applied, and campuses had not been held accountable for their performance with regard to specific goals. In other words, until 1999, CUNY had been functioning more as a federation than as an integrated university.
Any complex organization must balance how much independence to give to individual units, and how much to ensure that those units follow a uniform path. This article describes how CUNY has struck this balance, and how principles from behavioral psychology have been part of this process, and could be used elsewhere.
Background: Beginning in 2000, Chancellor Goldstein, with the support of the CUNY board, set CUNY on the path toward becoming a single university in which individual campuses would play their own specific roles but work toward one set of overall goals. At the same time, campuses — and their leaders — would be held strictly accountable for their performance. We call this the Performance Management Process (PMP), which is now entering its 10th year (more information on the PMP can be found here. )
The PMP operates on an annual cycle of goal setting, implementation, measurement, and consequences. These components of the annual PMP cycle are similar to the principles of behavior change outlined by psychologists whereby target behaviors are identified, contingencies are instituted, behavior is measured, and consequences occur. Over time, the PMP has resulted in large changes in CUNY’s performance, although many challenges in the use of the PMP remain.
Although some other large universities, such as Penn State,  also have university-wide goals accompanied by quantitative indicators, CUNY’s PMP appears to be the most comprehensive, designed for a university system that, despite being large and diverse, encompasses campuses that are closely linked. Students often take classes at more than one campus at a time, are likely to transfer from one campus to another, and what happens at one campus can have a strong and immediate effect on another campus.
Goals: The PMP gives three overall goals for CUNY: raising academic quality, improving student success, and enhancing financial and management effectiveness. These three goals are, in turn, subdivided into nine objectives (e.g., one of the objectives under the goal of improving student success is improving post-graduate outcomes). For each of the nine objectives, based on the current Master Plan, and after consultation with the college presidents and vice chancellors, the chancellor sets specific, university-wide annual targets and indicators (e.g., an increase in the pass rate on the nursing licensure examination — the NCLEX, is one of the post-graduate outcomes indicators). Then each president, for his or her own campus, sets specific targets and indicators that are based on, but not necessarily identical to, the university’s. A campus’s goals are based on its results from the previous year’s PMP, such that PMP results drive change going forward. The chancellor approves the campuses’ goals, after reviewing them for appropriateness, including the goals’ magnitude. Many campuses use the PMP as part of their strategic planning processes.
Implementation: After goals have been set for the coming year, campuses spend that year engaging in activities designed to reach their goals. Campuses are generally free to achieve their goals in whatever ways they deem best, though advice and support are available from the university central office.
Measurement: At the end of each year, the university’s Office of Institutional Research and Assessment prepares extensive data reports concerning the performance of each campus (e.g., the pass rates on the nursing licensure exam for each campus that has nursing programs). All of these data are available to the leaders of every campus. The PMP director also collects qualitative information about the past year’s performance of each campus in various administrative areas, such as the progress that a campus is making on updating its nursing curriculum. In addition, each campus prepares its own narrative report interpreting and supplementing its results. A central office six-person team, led by the PMP director, then discusses and evaluates all of this information, and the PMP Director prepares summary reports for Chancellor Goldstein. Finally, the chancellor himself reviews all of the data, information, and reports, and meets individually with the President of each campus to discuss the results — both positive and negative. He subsequently sends a confidential letter to that president summarizing their discussion. Follow-up communications occur as needed.
Consequences. Presidents’ raises depend, in part, on a president’s campus’s past year’s performance on the PMP. In recent years, the raises for presidents continuing at CUNY have ranged between 3 percent and 6 percent. Presidents are themselves encouraged to set raises for their executives according to the executives’ contributions to their campus’s success on the PMP. In addition, when funds are available, the university provides additional funds for each campus to use as "PMP Incentive Funds." These are professional development funds to be dispensed at a President’s discretion to ensure the continued high performance of any campus unit that has contributed positively to a campus’s performance on the PMP. The proportion of the university’s PMP Incentive Funds that a campus receives depends on that campus’s overall PMP performance, as well as on the campus’s size. In recent years, a typical campus might receive $50,000, with high-performing, large campuses receiving significantly more, and low-performing, small campuses receiving significantly less. Finally, when a president is evaluated, which occurs every four to five years, the PMP results for the past four or five years for that president’s campus are compiled and used as part of his or her evaluation. Over time, many presidents use their campus’s PMP results, along with the university-wide PMP goals, as levers for change on that campus.
Relationship to Psychological Principles of Behavior Change. Psychologists have conducted a huge number of experiments demonstrating the effectiveness of certain strategies in changing behavior. These strategies bear a striking resemblance to the components of the PMP. First a specific behavior is identified that needs change (e.g., the NCLEX pass rate), and a goal is set (e.g., an increase of 5 percent in the pass rate). Next, contingencies are put in place. In this case, increases in the pass rate will tend to be followed by a greater salary for the president, greater salaries for the president’s chief executives, and more professional development funds for other campus members, plus campus members and many other CUNY members will learn that the campus’s pass rate has improved. In other words, positive consequences follow improvements in the behavior. There is an absence of positive consequences (or aversive consequences — in the case of the PMP, a frank conversation with the chancellor) following lack of improvement or a decrement in the behavior. When psychologists apply behavior change principles, once an initial target behavior has been achieved, a new goal is set and the process repeated until the final goal behavior has been achieved, a process called successive approximation. Similarly, CUNY’s PMP seeks repeated incremental increases in the values of its campus indicators, a process known in management as continuous improvement.  Essentially, the PMP attempts to change a campus’s behavior similarly to how a psychologist attempts to change a person’s behavior. The PMP uses a system of rewards (and their absence, or punishers) to enhance campuses’ performance in targeted areas.
Using rewards and punishers to change specific aspects of campus behavior is not new. As just one very recent example, Kent State University agreed to a new provision in their collective bargaining agreement with the American Association of University Professors such that faculty will receive bonuses if first-year-student retention and research funds increase.
Psychologists have found evidence that when someone simply monitors his or her own behavior closely, the monitoring in and of itself can be sufficient to result in change in the behavior. With regard to the PMP, perhaps ensuring that a campus fully acknowledges its less-than-optimal performance indicators may be sufficient to catalyze change. Denial and avoidance of unpleasant subjects are behaviors that sometimes characterize campus administrations as well as individuals.
A concern that is sometimes expressed in using rewards to change individuals’ behavior is that use of such consequences will decrease a person’s own inherent motivation to change their behavior. However, psychologists have shown that if the target behaviors are very specific and if high standards are set, rewards can actually increase individuals’ own motivation to change their behavior. (See the work of Professor Robert Eisenberger. )
Although such procedures have been shown to be extremely effective in changing the behavior of individual people, it is an empirical question as to whether such procedures can also be useful with as complex an entity as a multi-campus university. The long-term effects of the PMP suggest that it can indeed be useful in this way.
Long-Term Effects. Since 1999, CUNY’s performance on many of the PMP indicators has greatly increased, and it seems highly likely that the PMP is responsible. For example, CUNY’s mean pass rate on the NCLEX has risen from 72 percent in 2000 to 87 percent in 2008, and is now substantially above the New York State mean. This increase is not surprising given the PMP procedures. As an illustration, the 2005 PMP results showed that a campus considered to be a CUNY leader in educating students in the health professions had an NCLEX pass rate (85 percent) that, although not extremely low, was not as high as appropriate. Therefore, directly as a result of the PMP, this campus brought in a new dean to oversee nursing and raise the NCLEX pass rate. In 2008 that campus’s NCLEX pass rate had risen to 92 percent.
Some other examples of indicators that have greatly improved since institution of the PMP include the following increases: headcount in credit-bearing courses by 19 percent, enrollment of all types of transfer students by 40 percent, the pass rate on the liberal arts and sciences teacher certification examination by 9 percent, the percentage of associate degree students transferring to a CUNY baccalaureate program within one year of degree completion by 6 percent, grants and contracts funds by 105 percent, and donations by 270 percent. In the case of the last two of these measures, grant/contract funds and donations, it should be noted that they are part of an objective whose results are double-weighted in the PMP evaluations. Thus it is perhaps not surprising that the values of these indicators have increased so much since the institution of the PMP.
Challenges. The PMP faces many ongoing challenges and questions. These include:
- To what degree should targets consist of outcomes, as opposed to processes designed to achieve an outcome, and how are the terms outcome and process defined? What is an outcome to one person (e.g., the percentage of students passing the NCLEX examination) is a step towards an outcome for another (e.g., the number of nurses produced by CUNY that are still in the profession 10 years after their graduation).
- To what degree should campuses themselves determine how to reach an outcome as opposed to the central office providing guidance and even direction? Does the former promote campuses striving for outcomes that may increase their own quality and/or reputation, but possibly at the expense of other CUNY campuses or the university as a whole?
- Is it appropriate for the university to have one set of goals for all of the campuses, especially given that the campuses range between community colleges and doctoral institutions? Is the annual individual campus goal-setting, based on the university’s goals, sufficient to allow the campuses to express their individual situations and needs?
- To what degree should campuses’ performance be judged against the performance of other campuses versus their own performance in previous years? If the former, is it appropriate to compare campuses on the basis of their final outcomes given that the entering students are quite different for different campuses — should change scores be used instead?
- Does the annual cycle of the PMP inhibit campuses and the university working towards multi-year goals?
- How much of the data, information, and results for a given campus should be shared within that campus, with other campuses, and with the public? Should it be up to a President how much and how the PMP results for that campus get shared with that campus? Should the university ensure that all appropriate campus members are aware of a campus’s goals, are working towards those goals, are aware of the results, and benefit from positive results, or is that the campus President’s prerogative?
- Has the PMP had the most success with variables that are relatively easy to measure, such as NCLEX pass rates? How should indicators such as program quality best be assessed so as to promote future increases in quality?
- Is the amount of time (and thus money) needed to effect the PMP justified given the effects of the PMP? Many people each spend many days (and sometimes months) on the PMP: presidents, campus members involved in goal-setting and report-writing, staff in CUNY’s Office of Institutional Research and Assessment, the CUNY chancellor, the CUNY vice chancellors and university deans, the university PMP director, and the other five members of the PMP review team.
- Do the chosen indicators promote optimal campus behavior? When the PMP measured only NCLEX pass rates, one campus permitted only its top nursing students to take the NCLEX, thus ensuring a high pass rate. CUNY then added a PMP indicator for the number of a campus’s students passing the NCLEX.
- And, finally, are the quantitative indicators formulated and calculated properly? Are they valid measures of the targets? Do quantitative indicators that are easier to calculate tend to drive out more difficult (and thus more expensive) indicators?
Conclusions. The PMP has been instrumental in improving CUNY’s performance during the past decade. It has been a successful method for managing the performance of a complex university system. However, CUNY’s environment and CUNY are constantly evolving. In addition, there are continuing challenges and tensions regarding the ways in which the PMP is conducted. There is no one perfect way to conduct the PMP, and reasonable people can disagree about whether and how it needs to be changed. However, in its elements, the operation of the PMP is consistent with the principles that have been shown to induce behavior change in individuals: identification of target behaviors, measurement of those behaviors, and consequences for those behaviors. The PMP’s success appears to constitute an example in which the principles of behavior change apply to a complex organization as well as to an individual. When CUNY Chancellor Goldstein instituted the PMP almost ten years ago, he put in place a performance management process that has greatly benefited CUNY and will continue to do so. Other universities may similarly enhance their performance through application of the principles of behavior change.
Alexandra W. Logue is executive vice chancellor and provost of the City University of New York. She is also a professor of psychology at Queens College, and a member of the CUNY doctoral faculty in learning processes and behavior analysis