Flexible Self-Control

Sunday, March 25, 2012
By dreeves

The Beeminder bee walking down the yellow brick road to akrasia-free bliss

This article is cross-posted on Messy Matters.

The problem of self-control may be a ridiculous first world problem but it’s the granddaddy of first world problems and I want to solve it. We live amidst a deluge of opportunities for instant gratification, especially in the form of food and entertainment, and most of us don’t handle it well. The general problem, known as akrasia, is this: you understand your own best interests when you consider them dispassionately, but in the moment your decision-making is distorted. The best time for, say, a workout is always “tomorrow”. And the best time to start working is after “just one more” clip of [redacted; save yourself!] on YouTube. Knowing that the problem boils down to time scales (hyperbolic discounting) implies the fundamental solution: commitment devices. For example, many people buy cost-inefficient individual-sized packs of cigarettes or candy, paying a premium to throttle their future consumption. Deleting games from your computer or going somewhere without internet access to get work done is another common example. You need to lock yourself in to your chosen course of action, like maintaining a healthy weight, eating right, or getting blog posts written every month.

When we wrote about this for the inaugural post on this blog, we included a straw poll with a slew of commonly used commitment devices. As we pointed out then, the tricks and lifehacks on that list are surprisingly toothless — they’re mostly band-aid solutions. Economists have been wondering for at least half a century why we don’t see more serious use of commitment devices in the real world. But they’ve had an answer for equally long. As Robert Strotz speculated in 1955 [1], the reason is risk and uncertainty, “both as to future tastes and future opportunities.” [2] This should be obvious to our data-nerd readers. We sometimes call it the Principle of Delayed Commitment: the simple decision-theoretic fact that you shouldn’t commit to a course of action until you have to. Flexibility is valuable; laziness is a virtue. For example, you shouldn’t toggle the state of the toilet seat after you use it because you don’t know what state the next person will want it in. What if you expend the effort toggling only to have the next person to use the bathroom, perhaps you, toggle it right back? Such waste! I’m sure I’m preaching to the choir here. [3] Or take going to the gym. You shouldn’t pay for a membership if you can, for a similar price, pay as you go. After all, in the future you may decide not to go to the gym, making pay-as-you-go cheaper, in expectation.

“Two forms of irrationality can cancel each other out”

But you know where this is going. Yes, you might well decide not to go to the gym, but it probably won’t be a rational decision. The solution, ironically, is more irrationality. Commitment devices blatantly violate the principle of delayed commitment, removing future flexibility for no reason other than to thwart your future self. But if that impetuous version of yourself will undermine your own goals (such as getting in shape) then that’s quite rational from your current perspective. Two forms of irrationality can cancel each other out. Which is to say that if you irrationally over-weight immediate consequences — if you’re akratic — then you should “irrationally” pre-commit.

A gym membership is a literal commitment contract, albeit a weak one. It turns the gym fees into a sunk cost, committing you to pay whether you visit the gym or not. Commitment devices may be an old idea but they’re enjoying a technology-fueled renaissance in the form of a crop of new web services. GymPact had the clever idea to give the standard gym membership commitment contract more teeth, by charging an additional fee for missing workouts. StickK.com — a pioneer in bringing commitment devices to the web — generalized the idea and offers commitment contracts on anything you can (pre)specify. Fighting irrationality with irrationality is easier than ever.

But we’d like to make those two irrationalities cancel each other out while minimizing the collateral damage. In other words, bind yourself with a commitment device that thwarts your akrasia while retaining maximum flexibility as to future tastes and future opportunities. [4]

The Akrasia Horizon

Here’s a now well-tested way to do that, by focusing on data tracking:

Akrasia Horizon: The time horizon beyond which you can make rational decisions, undistorted by akrasia

  1. Retain the flexibility to change your contract in light of new information [5] (like, 40 hours of actual focused work per week is damn hard!). That sounds like it defeats the point of a commitment contract, but recall the fundamental problem of akrasia: over-weighting immediate consequences. To beat akrasia you only need to bind yourself for whatever the horizon on “immediate” is. Based on a study on grocery-buying habits — when buying groceries online for delivery tomorrow people buy a lot more ice cream and a lot fewer vegetables than when they’re ordering for delivery next week — and our own accumulating evidence, we’re taking that akrasia horizon to be one week.

Example Beeminder graph

  1. Just commit to progress. You don’t have to know what you’re committing to when you commit, which also sounds (oxy)moronic but what I mean is this: Commit to keeping your data points on a path to your goal (the “yellow brick road”) that you have control over as you go, subject to the akrasia horizon. That is, commit to something like “work out more” or “lose weight” and then decide as you go what that means based on your data.

So a maximally flexible self-control tool is one that commits you to keeping all your data points on a path to your goal that you specify and can change the steepness of at any time. The only part of the path that’s fixed is the upcoming week. [6]

Easy Street

You may be wondering how anyone could ever fail to stick to a commitment that’s this flexible. Here’s how: if you’re highly akratic. Such a person may well find it a daily struggle to stay on track. Yeah, you can always choose to wuss out and flatten the yellow brick road, but only starting in a week, which you don’t want to do. You want to wuss out Right Now, dammit! I mean, just for now, while you eat this pie, and then you’ll behave again. No such luck though.

The daily struggle to stay on the road does not induce you to touch that road dial. You always want to make it easier “just for today” — which the akrasia horizon doesn’t allow — and you always think you’ll get your act together by next week.

 

Eight Hundred Dollar Postscript

For three years now we’ve scrambled to make sure a Messy Matters post got out the door at least once in every calendar month. Such is the powerful psychology of “don’t break the chain”, also known as the Seinfeld hack. But the Seinfeld hack’s greatest strength is also its fatal flaw: Once you do break the chain, all the motivation it provided bursts like a bubble. You’ve got to somehow motivate yourself to build up another long chain to not break. Until then you’re on a “one more day won’t matter” slippery slope of sloth. (Notice how we missed our end-of-February deadline and how many days into March it now is!)

Well, Beeminder, as we say, is safety rope for slippery slopes. So it’s time to do this right! We shall henceforth publish on Messy Matters as often as the following yellow brick road dictates, or give one of you $800. We’ll get the added flexibility, of course, to change the rate of posts as needed, subject to the akrasia horizon. I think it can also add a different kind of flexibility and help prevent end-of-month all-nighters: If we publicize a draft of a post (to the hard-core Messy Matters readers who follow our twitter feed) that will count as half a post (publishing a draft will count as the other half).

Beeminder graph tracking posts on Messy Matters

I’m quite serious about this $800 offer, subject to some fine print [7]. As the dozen or so people who have collected on our $20 typo bounties will attest, we’re good for it. [8]

 
Illustration by Kelly Savage. Thanks to Bethany Soule, Sharad Goel, and David Yang for comments.
 

Footnotes

[1] This is the earliest treatment of akrasia and time inconsistency that I know of in the economics literature: Myopia and Inconsistency in Dynamic Utility Maximization, by Robert H. Strotz in Review of Economic Studies, 1955-1956.

[2] The uncertainty about how our preferences or our circumstances may change is a very legitimate reason to eschew commitment devices, but there’s evidence that we go overboard in trying to retain flexibility / keep options open / not close doors.

[3] I’m ignoring the signaling aspects of the toilet seat debate, of course, but (a) that’s a boring debate, and (b) as you’ll see, this post is all about one-person games anyway.

[4] Spoiler: This is Beeminder’s key insight that the more well-known StickK (“put a contract out on yourself”) is missing. Contracts are part of StickK’s DNA. No surprise — one of the founders is a contract lawyer. In fact, the extra K in the name is for the legal shorthand for “contract”. Beeminder, on the other hand, was co-founded by a Messy Matters blogger and core to its DNA is data. That’s what led us to the Akrasia Horizon solution for flexible commitment contracts.

(Perhaps surprisingly, it took a ridiculous number of iterations to get to that point. For the longest time we struggled with different ways to deal with the fact that it’s so often hard to decide what to commit to. We tried many variations of having multiple yellow brick roads for a single goal, so that you could specify an ambitious goal as well as a bare minimum. It was always too messy, or would backfire altogether and be paralyzing. We think the road dial with an akrasia horizon is a big leap forward. And it seems so obvious in retrospect!)

[5] Beeminder does this with a road dial for adjusting the steepness of the path to your goal.

[6] You can also change the goal date to any date in the future that you desire, except within the coming week, or leave it open-ended. Also, if you start with a flat yellow brick road (as Beeminder encourages) then you’re not making any commitment at all until you can do so informed by your data.

[7] For the fine print we’ll mostly piggyback off of a similar meta post on the Beeminder blog. That is, if you catch us off of this yellow brick road, yell “Off the road!” in the comments. The first one to do so gets $800. Messy Matters is on New York time so the post has to be live on messymatters.com at the stroke of midnight on an “emergency blog post day” (when the graph is red) or the $800 is yours. To make sure we don’t keep writing drafts instead of publishing, we have a rule: Once a draft is publicized that’s what we have to actually publish next.

[8] If you catch a typo in a draft you can either hope it survives so you can collect $20 or let us know and we’ll thank you at the end of the post.

Tags: , , , , , , , , , ,