Jerry Z. Muller, author of the “The Tyranny of Metrics” as well as a professor of history at the Catholic University of America in Washington, D.C., is fixated on metrics and how corporations tend to over-rely on them.
Rather than looking in the business world, he came up with the idea for the book partly by observing what he perceived to be dysfunctional patterns going on in public policy over policies like No Child Left Behind. “[Those policies] place a lot of emphasis on standardized measures of performance, and then rewarding and punishing institutions accordingly,” he says.
“The Tyranny of Metrics” looks beyond policy and covers the pressure — across business, education, government, medicine, and other facets of life — to measure based on performance metrics. Chief Executive talked with Muller metric fixation, how board directors and companies tend to lose sight of the bigger picture when zeroing in on metrics, and much more.
Below are excerpts from the interview.
Why do board directors, CEOs, and companies fall victim to metrics fixation?
It’s a combination of beliefs that seem plausible and even scientific and objective when you first hear them, but are often counterproductive when they’re actually put into effect. It involves the idea that human judgment is unreliable and that you can replace human judgment with standardized measures of performance.
That is turn is connected to the idea of pay and punishment for performance. The idea is that if you fulfill or exceed metric targets, this should directly lead to increases in your pay, or if you fail to meet them, it should lead to one or another form of punishment.
So, for example…you might recall that the CEO of Mylan was to be remunerated if she met very high targets for growth and profitability. That led to the whole EpiPen scandal, in which the firm radically raised the price of a necessary pharmaceutical product, which helped lead to the decline in the reputation of the firm. Or in the case of punishment, there’s the case of Wells Fargo, which told its frontline employees that they had to cross-sell various products or else they would be punished in one way or another, including losing their jobs. This ended with all sorts of corruption, as some employees opened up bank accounts and credit cards for customers, which the customers hadn’t requested, in order to game the metrics. The employees increased the metrics of performance, but in a way that was inimical to the ultimate purposes of the organization.
“Metrics do have propensity to make one lose sight of the bigger picture. They have a propensity to be too narrow, both in the case of metrics for individual performance and for those of the whole division of a firm or even the firm as a whole.”
Metric fixation is this combination of emphasis on standardized measures of performance, the minimization of judgment, and tying pay or punishment more or less directly to those metric targets. It’s a variation of the pattern of management by objectives. And often for a variety of reasons that I deal with in the book, this turns out to be counterproductive.
It seems like in a lot of these cases, [companies] almost just lose sight of the bigger picture, and zero in on those metrics that they forget kind of the larger impact of what could happen. What are some of the damaging effects of losing sight like this?
Metrics do have a propensity to make one lose sight of the bigger picture. They have a propensity to be too narrow, both in the case of metrics for individual performance and for those of the whole division of a firm or even the firm as a whole. There are some jobs in the world that do really have a single limited purpose. That’s sometimes the case in sales jobs. It’s the case when you have some standardized job like replacing windshields on a car or some similarly repetitive task that is not intrinsically interesting and doesn’t leave much room for initiative. In those cases, pay for measured performance can work, when the job is narrowly structured to a single purpose.
But in most organizations, people have multiple purposes in their jobs and there are multiple goals they ought to be trying to fulfill. And if you just measure one or two of them, then a couple of things can happen. Either people will actually just do the thing that you measure and reward, at the expense of other important elements of their jobs that are not getting measured and rewarded. In those cases, you have actually a narrowing of people’s focus in order to focus on the metric targets for which they’re being rewarded or punished. Alternatively, people may feel a sense of frustration, knowing that they’re being measured and rewarded for only part of what they actually are doing or need to be doing. And that can lead to demoralization of employees.
And then of course, there’s the problem on the executive level… of the propensity of narrow metrics like quarterly profits to lead to short-termism. If you tie CEO remuneration very closely to stock prices and if stock prices, in turn, move in good part based on the relationship between what the firm projected in its last quarterly projections versus what its actual quarterly bottom line shows, then you have a propensity to short-termism, that is trying to meet those metrics or surpass those metrics at the expense of the longer term interests of the firm, which involves things like investing in capital equipment or investing in human capital or investing in R&D. In all of those ways, metrics fixation can have a very negative effect on the actual performance of firms.
What are some strategies you’d recommend that would be more productive than simply measuring on data? How could data be used more wisely?
One of the most important ways is to distinguish measurement used for diagnosis and analysis by the practitioners themselves, from metrics that are connected to reward and punishment. It’s highly useful often to have measures that are used for diagnosis by the people actually engaged in the work itself or by their superiors. That’s different from measures that are connected to reward and punishment. When you have reward and punishment, then you have all kinds of incentives for gaming the metrics, or focusing too narrowly upon them, at the expense of what is not measured. So, that’s the first thing I would emphasize: is it’s not just a question of measurement or no measurement, it’s how the measurement is used.
And then there’s the question of who develops and evaluates the metrics. If the people who are being judged ultimately have a role in the development and the analysis of those metrics, that often makes them more effective. And thirdly, metrics are often useful as a tool because they do provide a certain measure of objectivity. And in that sense they can help to combat bias in bosses. But there’s a difference between using standardized performance metrics as a tool in which there’s also a role for judgment on the part of those using the metrics, versus using those metrics as a decisive factor.
[Also] when you’re looking at the measurements, it’s important to remind yourself of all the things that are not being measured, perhaps because they can’t be measured, and remind yourself that they may be at least as important or perhaps more important as the things that are being measured.