ON PHISHING, PHOOLS AND IRRATIONAL BEHAVIOR
April 29, 2013 § Leave a comment
I had the immense pleasure last week of listening to George Akerlof give his lecture Phishing for Phools, as part of Duke University’s conference on behavioral matters. The organizers were from the D-CIDES interdisciplinary program at Duke. Aside from the unique experience of listening to a Nobel Laureate, I had a personal reason for the gratification. In my career as an engineering researcher and manager I have been strongly influenced by two non-engineer/scientists. They are George Akerlof and Adrian Slywotzky (the concept of Value Migration). I used their work directly in commercial endeavor, although likely in somewhat amateurish fashion. However, there was nothing amateurish about the bottom line financial results from so doing. A third influencer, Richard Thaler, I never had the opportunity to reduce to practice, but affects my thinking today in matters such as encouraging energy efficient behavior.
Akerlof was awarded the Nobel Prize for Economics in 2001 (together with Michael Spence and Joseph Stiglitz) essentially for the concept of Information Asymmetry. It goes something like this. If I am selling you a used car and provide all information regarding maintenance records and so forth, you will have a sense of the price you are prepared to pay. If I provide you nothing at all, you are likely to think I am trying to pass off a lemon and will devalue it. Essentially, the asymmetry of information (seller knows more than the buyer) devalues it in the eyes of the buyer.
Akerlof is generally credited with coining the term ”lemon” for a defective car in his 1970 paper that first described the concept above. The work is also credited with an impetus to make more information available to car buyers, including the use of Vehicle Identification Numbers (VIN). Also of note is that this paper, which essentially produced the Nobel Prize, was reputedly rejected by the first three journals to which it was sent! This is not unlike the reception that Stanley Prusiner got for his paper on prions, infectious proteins that now are the accepted cause of Mad Cow Disease and other cross species infection. In his case it went beyond rejection to overt denigration, when it did get published. It too resulted in the Nobel Prize, in this case for Medicine. This is the curious instance of a Nobel being awarded for the right mechanism for a disease, after another was given for the wrong explanation (Carlton Gadjucek, Nobel Prize for Medicine, 1976) much earlier!
Akerlof defined a phool as a person who is an informed person but still makes an error in judgment. He explains the recent recession as occurring because phools were misled by highly rated derivatives. The ratings agencies could not possibly evaluate the mortgages underlying the derivatives but gave them high ratings nevertheless. He ascribes this to greed. One wonders whether there was an element of sheer arrogance: their sterling reputations demanded that they have the expertise to do so, and so they did. He reminded us that economists rely on the belief that people are rational and always act in their own best interests. But phools may believe they are acting rationally and in fact are not. He believes that there are folks out there phishing for these phools. Phishing here is used in the broader context of profiting from the gullible. Akerlof has a book in the writing entitled Phishing for Phools.
One other paper at the D-CIDES conference was very interesting. It was by Rick Larrick and co-authors and soon to be published in the Proceedings of the National Academy of Sciences. We have discussed his other work previously in this blog. It demonstrated that consumer choice was affected by the labeling information, in this case on compact fluorescent bulbs. In a laboratory study with real money and technology, conservative and liberal subjects purchased CFLs at the same rate when the economic benefits alone were emphasized, but if a “Protect the Environment” label also appeared with the CFL, moderates and conservatives became significantly less likely to purchase the CFL. I will post the paper when it is published. You may remember the Cialdini experiment from a previous blog post of mine. One wonders whether a segmentation of the subjects may have yielded different results.
The original work by Richard Thaler and Laureate Daniel Kahneman is now being built upon by a number of people, chipping away at this original belief by economists, that people act rationally and in their own best interests. Duke’s Dan Ariely has a body of work here. The social science of why people make the decisions they do will be crucial to the adoption of environmentally responsible practices. This will run the gamut from using less energy for the same level of gratification to the substitution of oil derivatives with more benign alternatives.
Gallons Per Mile: Its Time May Have Come
July 6, 2011 § 1 Comment
Three years ago Richard Larrick and co-workers at Duke University proposed the logic of using gallons per mile instead of miles per gallon (mpg). In their paper they pointed out that mpg was not linear and hence not intuitive in making decisions. Such a decision may be for example when considering replacing one of your two vehicles with more fuel efficient ones. Vehicle X is a land cruiser with 10 mpg and Vehicle Y is a compact with 25 mpg. One would assume that replacing Y with a 40 mpg vehicle would save more than replacing X with a 20 mpg vehicle. Not so. Not even close.
Assume that you drive 200 miles a week. Vehicle X would consume 20 gallons of fuel and the replacement would consume 10 gallons. At $4 a gallon, the saving would be $40.
Vehicle Y would consume 8 gallons and the replacement would consume 5 gallons. The saving would be $12.
This underlines the character of mpg: it is not intuitive for decision making. Gallons-per-100-miles is completely linear. Using that metric it is abundantly clear that the Vehicle A replacement results in more savings.
Three years ago that paper did not catch people’s fancy. One possible reason is that we have difficulty with a smaller number being more desirable. Also, gallons per mile would be a fraction; not wanting to deal in fractions forced the use of gallons per 100 miles or some such. The biggest reason was likely inertia and not wanting to replace the familiar.
But now it’s time may have come. Electric vehicles (EV’s) could force the issue. The easiest way to compute the “fuel” cost in EV’s is to determine the electricity used per mile and multiply by the cost of electricity. EV’s are likely to be very similar to each other in the charge used per mile. This is a huge departure from conventional vehicles, where the fuel efficiency can be dramatically different. EV differences will primarily be premised upon body weight. Other than that the variability will largely be in the type of driving: stop and go city driving will be more economical than distance driving, largely due to the regenerative braking systems. This is of course the opposite of conventional cars and the EPA will have to consider this in their testing.
The capacity of an EV will be measured in kilowatt hours (kWh). The Nissan Leaf has a capacity of 24 kWh and uses 0.25 kWh per mile, so it has a range of 100 miles for average driving on a full charge. There you have it. The kWh per mile is your gallons per mile analog. For EV’s nothing else makes more sense as a metric because the consumer knows what she pays for electricity at home, and easily computes fuel cost. In fact the ease of computation is going to be a key to driving the right battery charging behavior. We expect utilities to incent night time charging by heavily discounting rates at night. Absent that, significant day time charging could create serious problems for the grid.
Now, one could argue for EV’s to most usefully employ kWh per mile but for conventional cars to remain as before. Sure, but when being responsive to federal fuel efficiency mandates, the mixed fleet will need to be assessed. In so doing, the gallons per 100 miles, taken together with kWh per 100 miles for EV’s will allow direct comparison. This is because we know that a gallon of gasoline has 34 kWh of energy. So gallons per 100 miles could convert to kWh per 100. Somehow I doubt the populace will go for this wholesale switch to kWh. But changing from mpg to gallons per 100 miles will allow an easy computation of fuel efficiency of a mixed fleet and allow direct comparison with EV’s for consumer choice.