Monday, November 28, 2011

Superstar effect and skewed labour markets

I have blogged earlier about "superstar" effect wherein a small number of star players, artists, executives and other professionals command a disproportionately higher compensation premium than everyone else in the market. Superstar effect and its consequent out-sized compensations has been a major contributor to amplifying the attractions of financial sector as a career choice.

Chris Dillow has an excellent post where he points to the cognitive biases that exacerbate the "winner-takes-all" effect of such labour markets. Availability bias preys on our mind and makes us feel that the out-sized success of an always-on-television superstar can be emulated more easily than would be the case in real world. Overconfidence bias makes us over-estimate our own talent and conditions and thereby the probability of success. Probability misperception bias causes us to over-weight smaller probabilities (like winning lotteries).

All this results in misallocation of resources atleast in certain labour markets. Conventional wisdom on the cricket's Indian Premier League (IPL) and the numerous television talent shows is that they provide opportunities for talented youngsters to showcase their abilities and thereby make a livelihood in that field. The sudden emergence of these platforms and the sharp increase in the numbers of people benefiting, amplifies all the aforementioned behavioural biases and makes such vocations even more attractive.

Consider talent shows. A typical parent faces several perception distorting trends - there are a large number of vernacular television channels, most with their own heavily promoted talent shows; the strong memory of the ubiquity and fame of all the winners of the superstar shows (among such talent shows) like the Indian Idol; and the strong possibility of being connected to a relative or neighbour or workplace colleague whose child succeeded in a talent show and knowledge about their initial flush of fame and money. They therefore find the attraction of grooming their child to succeed in such talent shows irresistible.

Apart from all the three aforementioned cognitive baises, the parent faces another bias, the representativeness bias. It makes parents over-estimate the probability of their ward's success using available data as opposed to using an objective Bayesian calculation. Such over-estimation works at four-levels. One, given the large number of participants in each channel, parents fail to appreciate that the probability of their child's success is remote even in the particular show. Two, since their child is participating in a particular show, they over-estimate its importance over and above that of its possibly more influential (among talent hunters) competitors. Three, since any success in a talent show is followed by an initial flush of fame and money, parents over-estimate its importance and tend to see this as an inevitable precursor of things to follow, overlooking the strong probability that such a peak is less likely to be sustained. Finally, an increased number of parents now think the same way and try to encourage their children to participate in talent shows, which in turn significantly increases the market competition and thereby lowers the probability of success.

A Bayesian calculation will bear out the true magnitude of risks associated with attaching the fortunes of their children with such talent shows. However, cognitive biases overcome human beings and they unwisely yoke their children's careers to such professional choices. The same assessment would apply to parents encouraging their children to play cricket in the hope of IPL selection and success.

PS: In fact, a more critical assessment of these markets would reveal that even the argument about the market being able to accommodate an increased number of high paying professionals is questionable. True there is a demand for a greater number of singers and cricketers. But the superstars will always remain few in number. The rest will earn only a moderate amount, and that too only for a smaller shelf-life than with a regular occupation.

No comments: