As a former member of the Alberta Health Services Board who had my own disagreement with the previous Minister of Health, the news about the recent dismissal of the entire AHS Board over a dispute about pay-at-risk for senior executives had emotional resonance. Some of those who were fired are my friends and have served Alberta long and well in many capacities. It doesn’t seem fair that this should happen to them.
However, life often isn’t “fair”.
Perhaps more importantly, this episode reinforced my growing belief that pay-at-risk in Canadian health care is almost impossible to do well, and to wonder whether we should abandon it all together.
The philosophy behind pay-at-risk is that if a proportion of an executive’s salary (typically around 20% in health care) is dependent upon reaching certain goals, this will increase the likelihood that he or she will focus on the goals that their Board deems important, and will encourage them to work hard to achieve those goals.
However, in my view there are a number of problems with this approach.
First, to my knowledge there isn’t much convincing evidence that 20% pay-at-risk is enough to change behaviour. Some like Dan Pink suggest that pay-for-performance works well for mechanical skills, but is useless or even harmful for the kind of cognitive skills we want our senior health care executives to maximize.
Second, establishing the metrics upon which to base the pay-at-risk is often difficult in health care. The kinds of activities that a Board often wants to encourage executives to focus on (e.g. reduced readmissions in the case of a hospital CEO) often require reforms in parts of the health care system that are beyond the executive’s control (e.g. improvement in the quality of primary care so patients don’t end up in the emergency department). Executives can rightly argue that it isn’t fair to make their pay-at-risk dependent upon parts of the health care system beyond their control. However, focusing pay-at-risk only on the things executives can directly control encourages the kind of siloed approach to health care we’d like to get away from.
Third, in the absence of hard metrics for pay-at-risk, there is a tendency for Boards to be quite generous with the amount of pay-at-risk they approve. “Stretch targets” are no longer that stretch. A good Board is serious about its job of holding the CEO and organization accountable. At the same time, Board members want to be encouraging to the organization, and some develop personal friendships with the CEO. Little wonder that the proportion of pay-at-risk that is paid out is often high. If most executives get 80 to 90 percent of pay-at-risk, it isn’t very “at-risk”.
Finally, pay-at-risk is a nightmare to explain to the public and press, even when done well. When Stephen Duckett was CEO of Alberta Health Services, he and the Board agreed on explicit metrics that were truly stretch targets, and they were posted publicly. Stephen only achieved about 50% of his pay-at-risk. However, this was reported by the press as a bonus and the Premier was scrummed in the middle of the Calgary Stampede and asked what he thought about the appropriateness of AHS paying Dr. Duckett a bonus!
I note that Minister Horne’s press release about his decision to dismiss the AHS Board talks about “pay-at-risk” in the second paragraph and “bonus” in the fourth. The two aren’t the same. In the former, part of an employee’s salary is withheld if he or she does not meet performance targets.
At the moment, the Ontario and Alberta governments seem to be taking very different approaches to pay-at-risk. The AHS Board was just fired for wanting to provide their senior executive the pay-at-risk that was in their contracts. In Ontario, the Excellent Care for All Act that was passed three years ago requires hospital Boards to make part of their CEOs’ salaries at-risk.
The reflections above are based almost exclusively on personal experience and not a review of the evidence. In a future article, Healthy Debate plans to review the evidence for and against the use of pay-at-risk for health care executives.
The comments section is closed.
Andreas, Pay for performance does not have to involve pay-at-risk. When I was CEO of the Ottawa Hospital Research Institute we introduced rewards for performance of scientists by making next years pay dependent on the previous year’s performance. If we determined that the standard pay raise (inflation plus advancement) was “x” then the Institute’s top performers (usually about 5%) received a raise of 2x or even 3x. Those who were not meeting expectations received 0.5x or zero. By doing this every year a small number of superstars had their salaries increase at 2-3 times the rate of the average, and the cumulative effect was based not on one good year, but on consistent superior performance. Also the low level performers became acutely aware that their performance was impacting their pay. Could this be an alternative to pay-at-risk? It does require a systematic approach to the evaluation of performance, and would have to be modified for unique individuals such as a CEO.
%featured%Excellent analysis. As a former member of a Health Authority Board, I find the author’s experiences consistent with mine.%featured%