The McWilliams Anomaly

BrendanBurgess

Here is a piece I published in December 2003 on askaboutmoney.com and davidmcwilliams.ie.

The McWilliams Anomaly

Well informed intelligent people tend not to apply their training and intelligence when discussing subjects outside their field of core competence thus reaching bizarre conclusions.

It is named after the respected economist, David McWilliams who wrote an article in the Sunday Business Post (early to mid December 2003)about traffic lights as follows:

Last Thursday at 5.30am, red lights held me up at 27 out of 31 traffic lights. …no one had thought to synchronise the system…there is no sensory device to keep the traffic flowing…This would be a bit too logical for Ireland….nobody is interested in traffic management…even in Ungobungoland, the idea of “joined up” government exists…But not here…joined up thinking was not taught with joined up writing in babies, so it does not happen

If he applied his training as a research economist to the problem of traffic lights in Dublin, he would have approached the issue in a far more rational manner. He would not have relied on one sample – he would have collected data on a variety of routes and at a variety of times. Even with only one observation, he should have realised that a 90% red rate was strongly indicative of some underlying pattern. If the lights were not synchronised, around half of them should have been green. This should have prompted him to call the traffic light engineers to find out what the pattern was. The most likely explanation would be that the lights were synchronised for a particular speed or for a particular set of traffic conditions.

So why do experts in one discipline fail to cross transfer their expertise or methodology to everyday situations outside the area of their core competence? We are not “the rational man” that the economists would have us believe. We all tend towards irrationality. Over the past 20 years, Behavioural Science has made great progress in identifying the many biases in our thinking and decision making. Here are a just a few of the biases which the McWilliams article illustrates:

The Availability Bias.
The tendency for people to make frequency estimates based on the ease with which they come to mind. We notice and remember red traffic lights more than green traffic lights, so we dramatically overestimate their frequency. Which do you think are more common? Words beginning with the letter “L” or words where the third letter is “L”?

The Mental Accounting Bias.
Even when counting honestly, we tend to make mistakes which help to support our biases. So unless we are actually writing down the number of traffic lights, we may miscount them. Investors count their successful investments and often exclude some losses when calculating their overall return.

The Confirmation Bias.
The tendency to notice/remember information which is consistent with one’s prejudices. We believe that most traffic lights are red. We see a red light and this confirms it. We see a green light and we ignore it as an anomaly. Did you know that most black people in Ireland drive white cars? Check it out for yourself.

The Attribution Bias.
The tendency to attribute negative outcomes to external factors rather than to your own behaviour. “These lights are red because nobody is interested in traffic management” is more acceptable psychologically than “These lights are red because I have been exceeding the speed limit”. The corollary is that we tend to attribute positive outcomes to our own behaviour. “I outperformed the market because I am a clever investor, not because I was lucky.”

The Pessimism Bias.
The tendency to believe that others are treated more favourably than yourself e.g. the lights on the other side are always greener.

The Overconfidence Bias.
The tendency to place undue confidence on your own abilities. 82% of respondents in one survey rated their driving skills in the top 30%. Most people feel that they would do a much better job than the traffic engineers in designing traffic light systems or speed limits.

The Mental Rigidity Bias also known as einstellung .
The reluctance to change long-held biases despite overwhelming evidence to the contrary combined with a tendency to avoid data which might change one’s biases. No amount of information or statistics will ever persuade David McWilliams that the traffic lights are well synchronised. Every time he is stopped at a red traffic light, particularly when he is in a hurry, it will confirm his bias. This is deep rooted and psychological and unshiftable. And it is a costly bias – if he took time to understand the traffic lights system, he could adapt his driving behaviour accordingly and increase the frequency of green lights from 10% to 90% and save himself a lot of frustration.

BrendanBurgess

This post kicked off a fairly long thread, virtually all of which supported McWilliams. Unfortunately, the threads have disappeared.

I publicly and privately challenged David McWilliams to a bet. I would meet him at his house at 5.30am on a morning of his choice and drive to the Grand Canal. We would count the traffic lights together. If 50% or more were red, I would donate €1000 to a charity of his choice. If fewer than 50% were red, he would publish a piece with the headline "apology to Dublin Corporation". The only condition was that we had to drive just below the speed limit. He accepted the challenge, but we haven't arranged to carry it out yet.

The responses to the thread were disappointing. People pointed out that in their journey from X to Y, 70% of the traffic lights were red. But these people were crossing traffic, and of course, the lights should be red to give priority to the cars on the main road.

Others complained about the lights at specific locations. I told them that some lights were bound to be faulty or badly programmed, and that they should inform the Corporation. It took Dublin Corporation 3 months to fix the lights at the junction of Shelbourne Road and Lansdowne Road which were causing me problems, but my persistent complaints paid off.

I made the bet before I did any checking of the facts. On Xmas day at around 9pm, I drove from Glasthule to the East Link, which is roughly the same journey as McWilliams takes to work and the timing is probably similar to 5.30 am. There were 41 lights - 3 were red.

williamgrogan

Your comment, “Well informed intelligent people tend not to apply their training and intelligence when discussing subjects outside their field of core competence thus reaching bizarre conclusions”, reminds me of what happened when “Immanuel Velikovsky's Worlds in Collision was published.”

This daft book proposed that many of the stories & miracles from the bible could be attributed to the “fact” that Venus nearly collided with Earth. The “Science” in it was so bad that some Scientists boycotted the publishers.

However, there was some support from Scientists but generally not for the Science from their own discipline. An anecdotal story was told to sum this up.

Two Scientists are discussing the book, a Physicist and a Biologist. The Physicist said that the points made in the book about the effects on the Earth’s biology were very thorough, logical and well thought out but the Physics in it was rubbish. The Biologist disagreed completly about the Biology but pointed out that he was very impressed about the Physics.