October 28, 2010

A Black Swan Election?

By Peter Landesman
11 Comments

For those who think the upcoming election results are predictable, an enormous surprise may be in store. Unexpected events can occur when experienced prognosticators do not fully grasp the significance of certain variables. Typically, the experts posit a normal distribution (bell curve) of probabilities based on the often-assumed principle that a small variation of parameters results in a small change of what is observed. This assumption is not always valid. Sometimes minute changes in initial conditions produce catastrophic alterations. Edward Lorentz in 1963 found that it is impossible to predict atmospheric conditions due to the non-linearity of the weather equations. Sometimes this is called the butterfly effect.

The study of such unpredictable events in human affairs has led to the development of the Black Swan Theory. This term has become popular due to Nassim Nicholas Teleb's book The Black Swan: The Impact of the Highly Improbable. His criteria for a Black Swan event are as follows: it must come as a surprise to the observer, it must have a major impact on its surroundings, and it is rationalized after the fact as though it was expected. Teleb regards major discoveries such as the internet and the personal computer and historic events such as World War I and the 9/11 attacks as Black Swan events.

The recent meltdown of the world financial markets in the aftermath of the U.S. mortgage crisis generally is considered a Black Swan phenomenon as well. Using normal distributions, money managers base their models on their recent experience. They hedged to protect themselves against financial losses which seemed improbable. Everyone was at first complacent and then surprised as they all lost money. In retrospect, the collapse was perfectly predictable.

The Black Swan is lurking around the corner as we approach Tuesday's midterm elections. Pundits and pollsters believe that small changes in political preference lead to small variations of election results. A normal distribution of current forecasts has a mean of just about the forty seats needed for the Republicans to retake control of the House of Representatives. Among the most extreme predictions by a professional pollster is that of Dick Morris, who postulates a hundred-seat capture (a prediction of about three standard deviations).

An authentic appearance of a Black Swan at this election would result a win of 150 seats in Congress for the Republicans. Few professionals can make such predictions without facing ridicule and possible financial ruin. However, Rush Limbaugh has said that a tsunami is coming, and we all know that he is documented to be correct 98.6 % of the time! His choice of the word "tsunami" is creative because a real ocean-riding tsunami is a prototype of a Black Swan event. How is it possible that so many other experts could be so far off the mark?

It is because the art of polling is new. The pioneers of scientific polling, including George Gallup, started plying their trade only in the 1930s. Therefore, they missed the unusual landslide in the 1894 midterm elections. Also, as pollsters base their forecasts on the correlation between what people say and how people vote, there is a substantial opportunity for unanticipated error, since the behavior of human beings is not completely predictable. In the last few years, pollsters have concentrated their analytical efforts to discover how to find the best sample of voters.

Any one of the following conditions could lead voters to unwittingly repress emotions, potentially causing them to change their minds at the last minute:

The government lacks financial discipline.

Hope and Change have become Despair and the Status Quo.

The promises of open government have succumbed to backroom deals and arm-twisting.

The president's attempt to heal our county has not been successful.
Democratic legislators have seemed to exempt themselves from the consequences of ObamaCare.

A recent example of a polling error occurred in the primary election for the Republican candidate for New York governor, in which Candidate Carl Paladino was polled to be even with Rick Lazio but won by 26%. This is a consequence of the fact that in New York, there is a breeze of financial responsibility blowing across the Hudson River from Governor Christie's New Jersey. Paladino has behaved erratically and may have blown an excellent chance to win the governor's race, but at the last minute, voters could change their minds about him. Senators Schumer and Gillibrand could be knocked off the deck by a squall as New Yorkers throw up their hands and vote for changes.

Apparently somewhere in the Australian outback there really is a black swan; I hope to see one in New York soon.

Peter Landesman (mathmaze@yahoo.com) is a teacher, a mathematician, and an author of the 3D-maze book Spacemazes, with which children can have fun while learning mathematics.

http://www.americanthinker.com/2010/10/ ... ction.html

many links on this post; go to the link above if you wish to view them