Goodhart’s Law in the C Suite

Joe Nocera’s latest Bloomberg View column insightfully challenges the reigning dogma that a public company CEO’s only obligation is to increase the value of that company’s stock. While this is obviously one important concern, a monomaniacal focus on the stock price has negative consequences. In order to boost immediate returns on the market, a company will often under-invest in research and development. R&D only bears fruit in the fullness of time. In a well-functioning company there would be a mixture of investment in projects with immediate return and those with long time horizons. A short-term fixation, in which the quarterly earnings report is all that matters, can harm a company’s long term viability.

In making this argument Nocera draws upon the work of Harvard Business School professors Joseph L. Bower and Lynn S. Paine. In their article in the most recent issue of the Harvard Business Review, Bower and Paine present a detailed case about the perils of solely focusing on shareholder value. They point out that CEOs were not always judged on stock returns. Indeed, this became a widespread practice only in the 1970s and 1980s.

Yes: the linkage between stock price and perceptions of CEO performance, which now feels so ironclad, is relatively recent. Prior to establishing the tie between stock price and CEO performance, there were no clear metrics of CEO performance at all. Corporate boards needed some way to evaluate the performance of their CEOs, and evaluating stock price was as logical a place to start as any.

Goodhart’s Law, named after economist Charles Goodhart, states that: “When a measure becomes a target, it ceases to be a good measure.” I’m most familiar with this concept as it pertains to the impact factor, the mathematical ratio of how often a scholarly article has been cited in subsequent work. The impact factor made complete sense when Eugene Garfield proposed it in 1955, as a tool for helping librarians decide which journals to acquire. But since then authors and journal editors have gamed their practices in order to boost their impact factor. This is because so much prestige accrues to high impact factor journals and their authors — gaming becomes irresistible given these prevailing incentives.

The exact same evolution has occurred with respect to maximizing shareholder value as a proxy for CEO effectiveness. What made perfect sense initially has become degraded and distorted over time. Yet again a measure ceased to be useful as soon as it became a target.

Bower and Paine offer several ideas for making the evaluation of a CEO’s performance more holistic, and thus no longer tied to the single variable of the stock price. A truly holistic approach should be a buffer against Goodhart, as there would no longer be a single number or value to chase. Even though humans will find a way to game any system ever devised, at the very least these efforts should become harder.