The Law of Unintended Consequences (Revisited)

Humans tend to invent clean casual chains when they predict the future—A will cause B—while ignoring the inevitable subsequent events—C, D and E. As the erudite French economist Claude Frederic Bastiat said in his seminal but overlooked 1850 essay, That Which Is Seen, and That Which is Not Seen, the forward thinking economist “account[s] both for the effects which are seen, and also of those which it is necessary to foresee.”

This is remarkably difficult, as noted early on in Western thought. Aesop tells the story of a poor widow who kept a faithful hen that laid one brown egg each morning. The widow began feeding the hen a double serving of grain to acquire more eggs, but the hen grew “fat and sleek and lazy” and stopped laying eggs altogether.

Let’s consider a modern example. Several years ago, Albuquerque, New Mexico, introduced an incentive system to eliminate overtime work. The deal was straightforward: if garbage truck drivers finished their routes early, they could go home and still receive pay for the expected eight-hour day. However, a number of problems ensued. Drivers filled up their truck over the legal limit, got into more traffic accidents, missed pick-ups, and avoided necessarily repairs to their trucks. Albuquerque was worse off.

In Hard Facts, Dangerous Half Truths & Total Non-Sense, a somewhat dry but informative book, Robert Sutton and Jeffrey Pfeffer discuss the perils of financial incentives. When we talk about financial incentives we often focus on the debate between intrinsic and extrinsic motivation—about employee or student performance—but the unintended consequences they produce as equally important.

New Orleans instituted a program that rewarded districts that reduced crime. Many police officers started reclassifying crimes to give the illusion that crimes were reduced. Royal Dutch/Shell compensated its executive with stock options; they started overstating the reserves to keep the stock price up. In Seeking Wisdom, From Darwin to Munger¸ Peter Bevelin (another erudite thinker) mentions this parable: “There was a problem with mice on campus. The solution for exterminating the mice was to pay students $1 for every dead mouse they delivered. It worked! Until the students began breeding the mice in order to make more money.”

Bastiat’s point is when we think about events and their consequences we focus on “that which is seen; it takes no account of that which is not seen.” This is Cicero’s point about silent evidence that shows up in Montaigne (1:14) in the story of Diagoras the atheist. That is, we can get closer to the truth not through confirmation but by searching for examples that contradict what the evidence suggests. Bastiat goes on to mention another example I enjoyed: in terms of governments, businesses, and other people, we cannot see their alternatives, which makes judging their visible decisions difficult.

Bevelin’s solution is to take a “holistic view of things” is correct but a bit general and vague. Instead, I’d steal from Nassim’s Antifragile and inject a dose of harm, stress, randomness into a system so it reveals it alternatives. (I actually thought of this by miss-reading Bastiait second paragraph.) Using my anti-lock example, let’s say instead of looking at the benefits of anti-locks, we get people to use them as recklessly as possible. Doing so will reveal the error.

More later


One Response to “The Law of Unintended Consequences (Revisited)”