Incentives are Superpowers
Though it took me a while, I finally finished the book ‘Poor Charlie’s Almanack,’ a nod to the old Ben Franklin almanacs from early U.S. history, this one a collection of wisdom from Berkshire Hathaway’s 96-year-old vice chairman Charlie Munger, a Franklin devotee.
I would call this less of a book, though, and more of a miniature museum, as it contains many photos and tidbits in the margins that are quite similar to what you’d see in an exhibit. And, the book itself has a beautiful hardcover and glossy pages, every one of them full color. It’s probably the most expensive book I’ve purchased (non-college textbook edition, that is), somewhere around $70. But I kept seeing it recommended by some of the finance people I follow, and I’m a big fan of the Buffett/Munger duo, being an Omaha resident and Berkshire shareholder. (Mind you, my wife and I own one ‘B’ share apiece, to get an annual meeting credential if we want to attend.)
Since this is a book I did not want to mark up, I didn’t underline or take notes, thus I did not remember much (see this post on ‘transmissionism’ and the problem with lectures and books).
One idea did stick, however: ‘Incentives are superpowers.’ Much of what Munger says sounds like common sense when you read it, but you quickly see how deeply he thinks about his ideas. In discussing incentives, Munger provides a few examples - as he does with every lesson - to show how the superpower effect works. The one I enjoyed most is the account of the FedEx overnight workers. FedEx needed the workers to get shipments on planes, with zero errors and zero delay; prompt shipping is the company’s value proposition, after all.
How did FedEx incentivize its workers to quickly get the right cargo on the right aircraft? Based on Munger’s telling, it sounds like more than one option was tried. If you’re paying workers by the hour, the incentive is to wait to load the last parcel until the last minute of the shift to maximize the paycheck.
The solution in the end was simple: let the workers go home as soon as the shipments were loaded. The workers went home with the same pay regardless if it took two hours or eight; whenever the cargo was ready, they could leave. I would guess that these workers even came up with some creative ways to optimize the process so they could be home sooner. A huge win for FedEx, and a huge win for the workers.
Warren Buffett discusses incentives often in his annual letters, and tells of many a CEO whose compensation was tied to performance that Berkshire was seeking with the subsidiary. No compensation consultants necessary, the agreements were/are famous for being under one page.
It sounds so simple and smart, but I imagine this is all very difficult to pull off in reality. The age old principal agent problem is not easily solved, when complex people are incentivized to engage in many different behaviors. Some are motivated by dollars, some by less work, some by status, etc, etc.
I think one of the reasons the U.S. tax code is so complex is because of the incentive to find loopholes. Who likes paying taxes? Entire industries such as tax accounting, tax software-ing, and tax lawyering have become fairly lucrative enterprises by finding ways to minimize taxes for individuals and corporations. They want the code to be complex. (It’s one of the reasons I suspect efforts to simplify the tax code have failed, like this one in California.)
In the ‘Almanack,’ one of Munger’s speeches talks about how stock option compensation in the 90s incentivized bad behavior by corporations. Since stock options weren’t paid in dollars, but in possible future redemption by employees, companies lobbied - successfully - to not show stock option compensation as an expense on its profit and loss statement. If you don’t pay in cash, and don’t need to show stock options as an expense, you can greatly reduce the total expense of the company, and thus increase profits, which will often increase the share price, which will trigger the redemption of stock options by workers...it’s a cycle which, Munger argued, would eventually lead to companies taking shortcuts in other areas to keep the inflated profit numbers increasing. (He was right; this speech, I believe, was given before some of the accounting fraud of the late 90s/early 2000s was uncovered.)
We encounter incentives every day. For me, I have incentive to automate as much as I can at work so that the cognitive effort isn’t high for certain tasks, like processing payroll. I have a checklist, my payroll system spits out a journal entry, and I have a way to make that journal entry almost entirely thoughtless, yet correct.
A humorous example is seen in the behavior of my dog Linus. He’s a very energetic pup and needs a lot of play time to work off his energy. For much of his first two years, that energy was often channeled in less-than-endearing ways. Ripping open pillows. Stealing hanging towels. Running off with shoes and socks. The problem for my wife and I is the potential damage to these items (and in some cases, possible danger to the pup). Linus was incentivized to continue these behaviors, because few activities are more fun than a game of keep-away.
Partly due to his ‘growing up,’ and partly because of our efforts to change the incentives, a good portion of this misbehavior has stopped, as Lisa and I stopped chasing and indulging in the game. Now, when he grabs a pillow off the couch, we first ignore it, and that often leads to a fairly adorable image of Linus just sitting there, head on the pillow. Sometimes, he still wants to rip the pillow apart, but I would guess that we’ve successfully eliminated a good 50% or more of the misbehavior by ignoring the attempt to start the game.
So how do we better align incentives for organizations? How do we change the incentives to modify our own behavior? There aren’t always easy answers, as in the FedEx example. And my own bad habits are very difficult to break. Reputational and social effects can be very strong, and create incentives to engage in or avoid certain behaviors.
As Munger points out, people behave the way they do because it works. If incentives are changed so that a misbehavior no longer works, we’re very likely to stop. The difficult part is determining how to come up with the right solution.
Pictured: my buddy, being good (taken 6/30/19), followed by my buddy, that very same day, enjoying a game of shoe chew:
