TED Talk: We all cheat a little...


For the past few years I have really enjoyed TED talks. The ideas presented there challenge me as much as anything else I have found. TED simply started as an annual conference (now popping up in  satellite cities – Columbia even has one) which bring together speakers who are passionate about the three areas behind the name: Technology, Engineering and Design - though recently they seem to be heavy on social science and psychology (which I enjoy the most). Talks are usually only about 15 minutes and usually feature a nice corresponding presentation. Tt the moment there are over 1400 talks and I thought about making a resolution to watch one every day (or at a minimum at least a few per week) and then make a small write up about each one.
Here goes….


Dan Ariely, author of Predictably Irrational has been challenging the basic assumptions of classical economics for the past decade or so. Primarily that we are logical, profit-maximizing machines. I have already seen a few of his videos online and would recommend anyone check them out.
Here is a good example from his blog.
______
Dear Dan,
I live in a quasi-urban area near Washington, D.C., don’t own a car and take the metro to work. Near my home is a fleet of Zipcars (a car-sharing system starting at $8 an hour, including gas, insurance and up to 180 miles of driving in a day). If I bought a car, the monthly costs alone (insurance, parking) would amount to about $200; then there’s the purchase of the car, gas and tolls. For that money I could regularly rent Zipcars.
So why don’t I? I could go to different restaurants and entertainment. But each time I think of doing this, I ask myself whether I want to spend the extra money to rent the car and usually decide against it.
This issue comes up the most with groceries. There’s a fantastic supermarket a quick drive away that sells much better and cheaper produce than my local store. In the end, I feel like I’m choosing between (1) overpaying at my local store and feeling cheated and (2) going to the better store but also feeling cheated because I spent $30 on a Zipcar to save that same amount on groceries. What do you suggest?
—Michal

What you’re experiencing is a conflict between your enjoyment of a better supermarket and your cost-benefit analysis. What’s interesting is that if you bought a car, you’d spend much more money overall, but on any given week you wouldn’t feel the pain of paying to get to the supermarket. Because a car can be used for so many different purposes, no single one will feel like the reason for the car, and you’d only focus on the marginal cost of driving a few extra miles, despite the car’s overall expense and inefficiency.
Instead, you could try calling Zipcar and offering to pay them in advance for three hours of car use four times a month for a year. This way you wouldn’t undergo a cost-benefit calculus for every visit to the supermarket.
And if you can’t convince Zipcar to do this, how about putting the money you’re saving by not having a car into a “Zipcar” bank account, and linking the Zipcar use to the money you’re saving? And to make sure you use this money for the Zipcar, commit to giving whatever’s left in that account at the end of the year to a charity you hate.
______
In this video he talks about how we all cheat a little - our own personal fudge factor. Essentially we cheat as much as we can get away with. He also indicated his research suggests we tend to not cheat more or less when the stakes increase or decrease. I simple timed experiment he asked participants to answer as many simple but time consuming math problems they could in the allotted time – for each correct answer they would get $1. Typically the number was around 5 when they were verified by the researchers. When they were asked to self-report the scores, this number rose to 7. When the price per correct answer rose, cheating did not measurably increase over the self-reported scores with a lower price per answer – essentially increased reward did not correspond to increased risk of dishonesty. From there the experiments only got more interesting. He found we also tend to cheat less when reminded of morals (ie when we are asked to swear on the bible [which even worked to keep atheists honest], sign an honor code. People tended to be honest when they self-reported the number of Ten Commandments they could remember relative to the math problems listed above.  
He also suggested that we cheat more when we see others cheating (at least if we view them as part of our group). The most interesting experiment used paid actors in a research setting to clearly cheat. The experiment pre-paid participants with an envelope of money and asked participants to complete a series of questionnaires. After a few minutes the paid actor stood up, declared he was finished and left. Interestingly (the study was conducted at Carnegie Mellon) if he was wearing a CM sweatshirt cheating went up. However, if he was wearing a sweatshirt from nearby Univ. of Pittsburgh, cheating actually went down (at least the time devoted to completing the task).  
Lastly he pointed out a huge disconnect between money and money substitutes. His first experiment measured how long a pack of coke would last in a common area refrigerator. He then placed six dollar bills on a plate and left them in the fridge. None of the bills disappeared, all the cokes vanished.  In another math problem test he found that when people self-reported significantly more correct answers when they were paid in tokens. These short experiments have obvious implications for theft and financial markets.  
I appreciate how simple these experiments are. I do wonder how “scientific” they are. How well the controls were designed? How much variability is there between the different groups?  How well do intelligent college students represent society as a whole? I can guarantee those dollar bills wouldn’t last in the homeless shelter down the street. What about other countries? In The Righteous Mind by Haidt in his book, talked about how different other cultural groups are from those found here in the US. It seems rather obvious to me that morality would vary by culture – I wonder how it would alter the findings. It would also be interesting to see how age and development would alter these findings. How honest would children be in these experiments?
 It would also be interesting to conduct these experiments in real life settings. How much does forced interaction with the researchers alter behavior? How would they different if there was no fear of getting caught. Haidt also talked about how people don’t really care about being truthful but they care very deeply about the perception of their truthfulness. Ariely didn’t dive into this very much.
Also, given the relatively meaningless amount of money (a few dollars to a college student) diminish the value of the experiments. How would the results change if the values represented an entire paycheck or someone who desperately needed the money?
All of the questions aside, his results do tend to mesh very well with my own intuition and I enjoyed the video.