Olivier Sibony on noise
(First of four parts)
HEC Paris management professor Olivier Sibony, former senior partner at McKinsey and current senior adviser to Singapore-based venture capital firm Qualgro, is renowned for his work in decision-making and strategic thinking. Sibony coauthored the book “Noise: A Flaw in Human Judgment,” with Nobel-Prize-winning behavioral economist Daniel Kahneman and Harvard public policy professor Cass Sunstein. In this exclusive interview with the Inquirer, he talks about noise, and how family businesses can improve their decision-making processes.
Q: What is noise? Are there studies in the Asian context?
A: In diverse fields, from law and business to medicine and fingerprint identification, we identified this common phenomenon—the unwanted variability in judgments—that we call noise. Noise is happening in many places. Specialists in the US say there is noise in epilepsy, urology specialists in France say there is noise in prostate hypertrophy diagnoses, end-of-life treatment experts say there is noise in the decision to terminate care, which is painful and complicated. I hope that as the book gets more visibility in Asia, we will also get interest from researchers to study whether there are any Asian-specific features there.
Q: What about noise in family businesses?
A: By noise, we mean variability between judgments that should be identical. So noise is inherently a problem of organizations; the larger the organization, the more visible noise is going to be. It’s not immediately visible in a small business. In a [typical] family business, the first concern is not so much between-person noise, the variability between people.
Noise would be [on the level of individuals], their sensitivity to the circumstances in which they make their decisions, or to extraneous factors that should not influence these decisions [such as what they are feeling at the moment].
It’s difficult to actually study noise in a small organization. The only way to study [system] noise is to use large numbers of people. You can take a million judicial decisions—literally a million, because you need a large sample to see those effects—and measure whether the outside temperature or the time of day has an impact on the severity of decisions. Or take hundreds of thousands of medical consultations and see if doctors are more likely to prescribe opioids in the morning than in the afternoon (afternoon, in fact). Or whether doctors are more likely to prescribe cancer screenings or vaccinations in the afternoon or in the morning (morning, in fact). The only reason you can measure these things is because you have a very large sample.
Does it mean that there is no noise in a small organization? No. It just means it’s not the place where you can measure it, because you don’t have such a large base for comparison. But of course, noise is there. Say, there are large differences in people’s heights, depending on how well they’ve been fed. I can measure this in a large population, and it would be true. If I measure this within a family, you’re going to say it’s not true, because within the family you cannot see it. But of course it’s true, you just cannot see it in the family.
Q: That’s a good metaphor. Usually people just take the average, the mean. But averages are not always ideal, they can miss out on a lot of information.
A: Often people believe that the average is what matters—that if on average, you’re not making a mistake, then you’re okay. Suppose your insurance company needs to set the premium for policies for customers. On average, the premiums set are okay, but half the time you’re too expensive and half the time you’re too cheap. Too expensive and you lose business, too cheap and you lose money because you pay more in claims than you collect in premiums. You don’t cancel out your two errors—you make two costly errors.
Averages can reveal the existence of bias. But once you’ve addressed the bias, you need to reduce noise, to ensure that each individual decision is correct. Otherwise, you’re making costly mistakes.
(Next week: Reducing noise)
Get “Noise: A Flaw in Human Judgment” at Fully Booked. “Noise” will also soon be available at National Bookstore. Queena N. Lee-Chua is with the board of directors of Ateneo’s Family Business Center. Get her book “All in the Family Business” at Lazada or Shopee, or the ebook at Amazon, Google Play, Apple iBooks. Contact the author at email@example.com.