A few thoughts about O.R. and dishonesty

Growing up in rural England, I didn't have many encounters with blatant dishonesty.  This was in the days before television brought wickedness into everyone's front rooms.  As children we were taught the standards of the Ten Commandments so didn't steal, didn't tell too many lies (little ones were "fibs" - does anyone use that word today?) etc. 

Then my brother had a temporary job working for a mail order company, and some of his work was in the returns department.  He regaled us with stories about the number of suits which were returned for a total refund as being unsuitable and "only worn to try them out for size", and which came back with theatre stubs or confetti in the pockets - clearly worn for a special occasion.  Cheating the mail order company was cheaper than hiring a suit for that occasion.  So, if you are going to tell a lie, make sure that there is no evidence to contradict your story.  Obviously, I have encountered many more examples of blatant dishonesty since then.

So I was intrigued by an article which describes this behaviour of "Order and return for a refund" with the name "Wardrobing".  (Not in the dictionary, yet.)  Guangzhi Shang, Bikram Ghosh and Michael Galbreth published "Optimal Retail Return Policies with Wardrobing" in Production and Operations Management (v26, n7 (Jul2017), pp1315-1332) and studied how the behaviour can be used to the benefit of the retailer!  Their results "provide new insights into how retailers can set prices and refund policies to effectively manage opportunistic behaviour by consumers".

So can O.R. help the dishonest person?  Well, yes it clearly can.  If the consumer knows the policy, they can play the game against the retailer! 

Long ago, one of the exercises we studied in a course on scheduling concerned a team or robbers who had two options (explosives or thermic lance) for their attempts to rob a bank, and they wouldn't know which option would work for that bank until they had made the initial assault.  The time to complete the tasks in the critical path diagram depended on the size of the team.  And there was a time constraint - the robbery had to be completed within a given time (how long before the intruders had been detected).  The question asked was "How big should the team be?"  There was a catch in the way that the problem had been set.  If the objective was to complete the robbery within the time schedule, then it needed a large team.  If the objective was to maximise the expected takings for each member of the team, then a small team was needed, because the robbers would come prepared for one option and abandon their attack if that was not appropriate.  Most students assumed the first objective. 

Fortunately, police forces use O.R. extensively, so can thwart many of the ploys of dishonest uses of O.R.


Popular Posts