Irrational Addiction

Gary Becker and Kevin Murphy turned the rational choice eye on addiction in a widely-cited 1988 JPE piece "A Theory of Rational Addiction." Not only is it widely cited, but judging by some of the empirical pieces I've had come across my desk, it's occasionally misunderstood. Here's the relevant quote: "[t]he basic definition of addiction at the foundation of our analysis is that a person is potentially addicted to c [the consumption good] if an increase in his current consumption of c increases his future consumption of c. This occurs if and only if his behavior displays adjacent complementarity." They rigorously define what that means, but it's pretty much what it says on the box: consumption now implies desire later.

If that doesn't sound like straightforward Chicago school economics, I'm not sure what else would. It's a positive theory of behavior, and if there are any normative claims to be found within, you'd need Coke-bottle Straussian spectacles to find them.

These sorts of rational choice models don't tend to match up with what behavioral economists find in the lab. Subjects tend to display odd aversions to things like loss, risk, ambiguity, and they tend to react to surprising events and pay attention to irrelevant options in ways that rational choice theorists didn't expressly model. Kahneman and Tversky refined basic behavioral models to account for these vagaries. The flagship paper in this literature remains the 1979 Econometrica piece found here. I encourage you to read it, but if you don't have the time, the very short version is that people have asymmetrical preferences between gains and losses rather than simply caring about an end measure of utility. K&T allow for what psychologists call "hedonic adaptation", an empirical phenomenon where people tend to get used to their surroundings. It's this same phenomenon that accounts for people griping about having lousy wi-fi on airplanes or having to stand in line for a Diet Coke at the ATL Delta Skyclub*.

So what's the salient difference? Becker and Miller say that heroin today implies heroin tomorrow. It's a theory of why addicts will continue to consume their substance of choice. Kahneman and Tversky offer a framework to explain why someone would even decide to try heroin in the first place. Prospect theory says that the potential drug user stands at the cusp of a decision, weighs all the potential upsides and downsides adjusted by all the outcome probabilities (and yes, this means going all the way down that slippery slope) and weighs them with whatever heuristics the agent thinks appropriate. Only then does she decide. It's a version of what Buchanan wrote in Cost and Choice, only with a few extra pieces of explicit psychological theory included.

It's the last bit there in italics that concern me from the perspective of euvoluntary exchange. How are heuristics formed? In particular, do programs like D.A.R.E. keep kids off drugs (hint: no)? If not, why?

I submit to you that these sorts of programs (not all that unlike abstinence-only sex education) suffer fatal credibility problems. In a Prospect Theory framework, the downsides of unapproved activity are oversold and the upsides are either never mentioned, are treated dismissively, or are conflated with the costs. Those of you with prior military service will undoubtedly remember the vivid images of sexually transmitted infections more than almost anything else you witnessed in boot camp. Some of you haven't eaten cauliflower since. Do any of you believe that sex isn't fun? Or that recreational drugs can't be enjoyable?

So how should kids react to courses like this? With skepticism, of course. Humans have very finely honed bullshit detection skills, even (especially?) children. If the intent of these programs is to alter behavior using non-credible talk, you might imagine a reversing equilibrium, where some potential users could become more likely to try drugs or unprotected sex or whatever because they knew the prior beliefs they'd been given were bogus.

Okay, so thankfully the meta-study I linked to up there says it's not that bad. D.A.R.E. is "just" a useless waste rather than an active source of harm as measured by addiction rates. But imagine how prospect theory might help folks build a more euvoluntary world by providing accurate, honest information and perhaps trusting that meta-rational people are capable of conducting economic calculus under conditions of if not perfect information, then at least marginally more complete information. Heck, it might even be worth it to help overcome the Holden Caulfield complaint. Yes Holden, you're right. Adults who tell you drugs are nothing but bad are in fact part of a cabal of phonies.




Becker, G. S. & K. M. Murphy. "A Theory of Rational Addiction." The Journal of Political Economy, Vol. 96, No. 4 (1988): 675-700.

Kahneman, D. & A. Tversky. "Prospect Theory: An Analysis of Decision Under Risk." Econometrica: Journal of the Econometric Society (1979): 263-291.


*a useful phenomenon, one that refused to let humanity squat over its laurels after the discovery of fire