Tuesday, October 23, 2012

Doping and Competitive Advantage: Chwang's Argument (Part One)

In one sense, the Lance Armstrong scandal is now over. The USADA report, and accompanying testimony of his numerous teammates and confidants, proves beyond all reasonable doubt (in my mind at least) that Armstrong was guilty of sustained and continual use of banned performance enhancing substances. Within the rules of cycling, he is undoubtedly a “cheat”. A judgment reflected in USADA’s and the UCI’s decision to strip him of his titles.

But in another sense the Armstrong scandal rumbles on. In particular, debates over his legacy can be expected to continue for a long time. Is he still a great athlete, despite his doping? Or is his athletic reputation now in tatters? A common response to these questions is that, despite his doping, Armstrong remains a great athlete (though perhaps not a very nice one). After all, virtually every other top cyclist at the time was guilty of doping and Armstrong was still better than all of them. The playing field was level in terms of doping; what separated Armstrong from the pack was his natural ability and drive, i.e. the attributes we typically value in our elite athletes.

I pass no judgment on the merits of this particular response. I do, however, note that it raises the issue of what the official stance toward the use of performance enhancing drugs (PEDs) should be within elite sport. The likes of Julian Suvalescu, who have long argued for a more liberal approach to PEDs in sport, have wasted no time in using the Armstrong case to support their view. Thus, they argue that, if sport is all about performance enhancement anyway, if athletes are going to do anything they can to enhance their performance, and if testers are always likely to be one-step behind the athletes, why not open up the rules and end the war on PEDs in elite sports?

There’s something to be said for this position. But in this post and the next I want to look at an opposing view. The view is set out in Eric Chwang’s article “Why Athletic Doping should be Banned” and over the next two posts I want to look at that article. In the remainder of this post I want to explain Chwang’s basic argument and the various caveats thereto. In the next post, I’ll see how he deals with the potential criticisms of his position. As we shall see, Chwang’s argument has limited applicability, so much so that it may not work in a sport like cycling, but it still provides interesting food for thought.

1. The Doper’s Dilemma
Chwang’s argument is that, given the right conditions, large athletic institutions (he gives the examples of Major League Baseball and the NCAA) ought to ban PED. Interestingly, he claims that this is for reasons that the athlete’s themselves are likely to agree with, and not due to typical anti-doping rationales such as a concern fairness, or the maintenance of authentic performance and the integrity the sporting enterprise.

His argument works from game theoretic premises. Put simply, his position is that, in very many cases, athletes dope because they are trapped in something like a Prisoner’s Dilemma. In other words, they do not particularly want to dope, and may often recognise risks to their health associated with doping, but they have to do it if they want to be competitive. To adapt the famous line from the Red Queen in Lewis Carroll’s Through the Looking Glass: it takes all the doping they can do just to stay in the same place.

If you’ve spent anytime reading the testimony of Armstrong’s former teammates over the past couple of weeks (see USADA REPORT), or, if you’ve read Paul Kimmage’s excellent (but dated) book Rough Ride, you’ll probably agree that this is a pretty accurate depiction of the situation in professional cycling. Perhaps more so than any other sport. But for the sake of completeness, we can actually try to model the game theoretic interaction imagined here.

Suppose we have two athletes (A, B), playing against each other in a competitive sport, with two choices (Stay Clean, Dope). If they both stay clean, the situation is as normal: what will separate their performances is a combination of natural ability, hard work, training and luck. If one of them dopes and the other doesn’t, the doper will gain a significant competitive advantage, while the clean athlete may well have to look for another line of work. In the parlance of the Prisoner’s Dilemma, this represents a “Sucker’s Payoff” for the clean athlete. Conversely, if they both choose to dope, the competitive advantage is cut away and once again a combination of natural ability, hard work and training will decide the day.

Although the net result of both athletes doping is no different from the situation in which both stay clean, game theoretic axioms assume that the both doping scenario is more likely. This is because doping strictly dominates staying clean: the athlete that dopes does better, no matter what his opponent does. Using figures from Michael Shermer’s Scientific American article from a few years back, the diagram below depicts the Doper’s Dilemma. The figures include the potential costs of getting caught doping, hence the expected payoff of the both doping outcome is slightly less than the expected payoff of the both staying clean outcome. Nevertheless, the logic is the same.

2. Chwang’s Argument
You might be wondering: where is the anti-doping argument in all of this? After all, the Doper’s Dilemma is, in a sense, neutral as regards the pro or anti-doping positions. It is simply a model of rational behaviour that suggests, if there is a genuine competitive advantage to be gained from doping (not an insignificant assumption), we can expect the majority of athletes to do it, thereby erasing whatever competitive advantage it might have provided.

To get an anti-doping conclusion out of all this we need to add another crucial premise. Specifically, we need to assume that, all things considered, taking PEDs is harmful. That is to say, that there are certain risks associated with taking PEDs which, unless there is a competitive advantage to be won from using them, make it undesirable to use them. The harm does not have to be particularly great, but has to be there nonetheless.

As it happens, I’m not sure how harmful PEDs really are. Certainly, one hears horror stories about athletes, particularly young athletes, having serious health problems arising from their use. But often this is because they have to administer the drug clandestinely and without proper medical guidance. If PEDs were permitted, some of these risks would disappear. Nevertheless, the assumption that PEDs will be, at least minimally, harmful is an important one for Chwang’s argument, and I’m willing to run with it just to see where it’s all going.

Once we have the premise that PEDs deliver no competitive advantage, and we also have the premise that PEDs are minimally harmful, we have the skeleton we need for Chwang’s argument. The flesh can then be added to the skeleton in the following manner:

  • (1) Doping is, at least minimally, harmful to dopers. 
  • (2) Because of the rational incentives faced by elite athletes, doping is unlikely to deliver any competitive advantage (because of the Doper’s Dilemma). 
  • (3) An athlete would prefer, all things considered, to avoid doing something that is harmful unless it delivers a competitive advantage. 
  • (4) Therefore, athletes would prefer it if no one dopes. (from 1, 2, 3) 
  • (5) We should not enact any policy that gives some athletes permission to dope but not others. (Impartiality Premise) 
  • (6) Within some constraints, we should give athletes what they want (Beneficence Premise) 
  • (7) Therefore, we should ban doping. (from 4, 5 and 6)

With some modifications, this is how Chwang phrases his argument (pg. 35-6 of his article). Is it any good? Chwang suggests that one thing it has going for it is that it is more robust than the typical anti-doping arguments. For instance, fairness-based arguments, as has long been noted, don’t clearly support anti-doping policies, and integrity/authenticity-based arguments rely on opaque concepts of dubious normative relevance.

3. Some General Caveats
But still, the strength of Chwang’s argument can only really be judged after we look at his attempt to fend off certain obvious criticisms. In particular, criticisms suggesting that the argument is both over and under-inclusive. We will give those criticisms the space they deserve in the next post. For now, a few general caveats and warnings about the argument should be borne in mind.

The first caveat is that, as Chwang himself notes, the argument isn’t perfectly valid. Indeed, while my formulation is slightly more valid than the formulation offered by Chwang, there is still a considerable logical gap between the conclusion and the premises: the need for a ban does not follow from 4, 5 and 6. However, this gap can be plugged fairly easily. The reason why a ban is mandated, as opposed to some other policy, is that, following the logic of the Doper’s Dilemma, the athletes will not be able to enforce their own preferences among themselves. Top-down intervention is needed.

A second caveat is that the argument assumes that comparative results are more important than absolute results. In other words, it assumes that what really matters in sport is how the athletes rank relative to one another, not how they perform relative to some absolute standard (e.g. time). This may be true in some sports, but not in others. For example, it may be true that, in soccer (Football) what we care about is which team is better, not how many goals they score; but, conversely, it may be true that, in 100m sprinting say, what we care about is how fast the winning time is, not how the sprinters rank relative to one another. In sports in which the absolute outcome is more important than the relative one, Chwang’s argument may not apply.

A third caveat, and a fairly major one at that, is that the “within some constraints” clause of premise (6) is left unspecified. Athletes may want lots of things that it would be unreasonable to grant them, how do we determine where the borderline lies between what is reasonable and what is not? Chwang tries to address this problem when responding to the over-inclusiveness criticism of his argument. We’ll be looking at that, along with some other criticisms, the next day.

No comments:

Post a Comment