Recently, a new viral video made the cover of People magazine. Six cheerleadering high school girls trapped a seventh girl at their house and beat her up on tape. I can't find the original youtube, but I suspect it was hugely popular for it to make People's cover. The girls were charged with a string of offenses, including kidnapping, which had a maximum punishment of life in prison.
I've been wondering this for a while, but what is the right response for intentionally harmful viral videos? First of all, to the defense of makers of viral videos. They do not control what becomes viral. They may intentionally be trying to humiliate someone, but they cannot by themselves make people watch the videos. Nevertheless, viral videos cause from what I can see tremendous damage to victims. Both of these points have to do with the probability function that characterize the distribution of viral videos. I can't prove this, but I suspect that like websites in general, viral videos are distributed by a power law function. This is another way of saying that they are unpredictable, partly because a small portion of all videos end up becoming viral. But, when they do become viral, their effect is huge. Here's an example of a power law distribution from a study done by a police department working with social services:
DEPT.OF SOCIAL SERVICES about solutions to problems like homelessness, violent police officers and high-polluting cars. All three problems follow a power-law distribution when plotted statistically on a graph… Writer tells about Murray Barr, a homeless alcoholic man in Reno, Nevada. Two local police officers, Steve Johns and Patrick O'Bryan, tracked chronic inebriates for six months and found that just one of them ran up a bill of a hundred thousand dollars at a single hospital.My question is what is the appropriate punishment for viral videos ex ante? That is, how should we compensate victims? If you think that the damages to the victim grows with the size of the viewings, then very quickly the exponential growth in the video's popularity will mean damages will grow in proportion to the number of viewings. That means that it may be optimal to set fines equal to some scale of the phenomenon. If having the video viewed publicly creates harm of $1 per viewing, then already the girls probably owe the victim $1-2 million, and it provides an incentive for perpetrators to work hard to get the video offline as soon as possible.
With the ubiquity of mobile phone video cameras, and distribution sites like youtube being so efficient, the cost of producing a harmful video has fallen to almost nothing. Setting high penalties could be effective at raising the expected cost of making these potentially viral videos. Even though the probability of making a viral video is statistically at zero (if it follows something like a power law distribution), a high enough penalty will make the expected cost positive. But that necessarily means that when one of these viral videos are brought to law, it will receive a seemingly draconian punishment that will feel harsh beyond words. This falls right out of Gary Becker's seminal economics of crime paper, though: Crime and Punishment: An Economic Approach. In that paper, the supply of offenses to the market is a function of `price' wherein price is itself a combination of the probability of arrest and the penalty conditional on arrest/conviction. If probability is low, because of scarce resources, then to deter, penalties must be high, and vice versa. The optimal punishment for viral videos is, I suspect, analogous to this problem, requiring the setting of very high penalties to offset the way in which the power law makes probabilities of success on the creation of viral videos astronomically low.
No comments:
Post a Comment