Ed.X Behavioral Economics

Notes from Ed.X course on Behavioral Economics

6.3 Discussion on End User License Agreement (EULA)
The way in which something is disclosed is also important. If the disclosure is written by lawyers with the intent to not be clear and to cover themselves in all possibilities, then clearly consumers will not have the patience to read lengthy disclosures or the capability to understand them. The software End User License Agreement (EULA) is the classic case of forcing an active choice, but giving totally incomprehensible information. Perhaps we will have future digital assistants that will translate and summarize this stuff for us humans. I hope those programs don't come with EULAs.

6.2 Decision Aids

Wisdom of Crowds
I think there is a third alternative to expert versus model, and that is applying the fascinating principle of the wisdom of crowds. It is applying the same technique as averaging the model and the expert to get the best performance. This seems to be a dominant strategy, ranging from choosing a product based on Amazon customer ratings and reviews, choosing a restaurant based on customer ratings, or choosing an index fund in your 401k.


6.1 Some recent nudging interventions

Flu Shot
If employers provide free flu shots, it might be possible to do away with the standard economic incentive of $50 and use an entirely nudge-based approach where calendar invites are sent (which I have found hold more significance than emails) so the default opt-in makes it as easy as possible for the employee to receive the flu shot.


5.4 Nudge Challenge

Applied learning
This is a great idea to make sure we apply the learning from this course into practice.


5.3 A Practitioner's guide to Nudging.

Nudging for charitable contributions
Some companies allow employees to make charitable contributions to be deducted from their paychecks. This is a great nudge that makes it easy as possible to do the right thing. Providing charts showing participation of different parts of the organization helps add an element of fun with friendly competition.


5.2 The Nudger's Toolkit

Hedonomics - the case for Apple
It seems like the case of Apple is already taking advantage of Hedonomics. By blending marketing to make Apple the cool choice with attractive, stylish products and intuitive and attractive interfaces, Steve Jobs and Apple changed the conversation around computers and phones from technical specs and price to subjective decisions on visual appeal and ease of use. In fact the cognitive strategy of getting people to think differently played a prominent role in one of their marketing campaigns.


Nudging 2.0

Cost Effective Approach
Whether it is getting kids to make healthier choices by presenting healthy foods as the easiest option, by putting unhealthy options last or more hidden, to efforts to minimize global warming with low cost solutions, like California's orbs for homes which turn green when energy usage is low or red when above average, nudging has proved in practice to be quite a cost effective approach. In this era of austerity caused by the financial collapse, I would think people from all political persuasions could support something like the UK's behavioral intervention team. Business would do well to adapt and hire/contract behavioral teams as well to give them a competitive advantage.


Debate 4

Experimentation - Theory interaction
Human behavior is influenced by cultural factors, context, individual history, and so many other factors including chance, that it is folly to try to produce one overarching simplified theory. That said, one of the participants made a good point that theory helps guide experiments, and I suppose the reverse could happen where unexpected experimental results might drive the creation of new theories. This back-and-forth interaction between the two is what has allowed for progress (not perfection) in the social sciences, which until recently were all theory with no experimentation or flawed approaches, such as no randomized control group.


4.3 Consumption Vocabulary

Online Shopping tools
Seems like an opportunity for retailers to provide attribute based information for their products and perhaps even asking consumers to rate certain attributes as part of a recommendation engine, which would then run a regression. The common algorithm used today is based on whole purchases or whole ratings and not ratings of parts. For example, people who bought this item also bought this other item or people who like this movie also rated this other movie high. Just not sure if people could be enticed to answer lengthy ratings (like 150 quilt patterns), or take the time to learn terminology. Imagine for each movie they rated 4-5 stars on Netflix, to be further questioned on details such as acting, dialogue, setting, costumes, plot, etc. This approach would definitely work though with free wine!



4.2 Understanding Intuition and Judgment
Applicability to Hiring?
This seems like it would be good to run a regression for the job interview process so that one could identify over history, what cues could be used to identify those hired who became stars and those who eventually left the company, and whose judgment of the interview panel turned out to be most accurate and what attributes did they weight the most.


4.1 Issues in Experimentation

RCTs - The Scientific method finally reaches government policy
The Behavioral Intervention Team in the UK is leading the way in finally applying the benefits of the scientific method in evaluating the effectiveness of government programs versus the previous way of philosophical debate. The illustration of the belief of even medical scientists that steroids would reduce inflammation of head injuries, which was proved incorrect by a Randomized Controlled Trial (RCT) in 2005 is proof that even very smart people can make assumptions, which may turn out to be incorrect and even harmful if not properly tested in an RCT. The fact that many people vehemently argued against such an RCT is further cautionary tale that change will meet with resistance, even if well intentioned.


Debate 3

Web field & lab
With commerce more and more conducted on the web, field studies will become easier to conduct given web technology to randomly assign a percentage of traffic to different versions of web pages, called A/B testing. With the click stream data gathered, Google even has software which will analyze the data and give recommendation on whether to go with the new page or the control page. This is even better than lab studies because participants in the study don't even realize they are part of a study. Lab studies on university campuses have to overcome the problem of homogeneous participation and the observation effect where the participants know they are being observed and that might affect behavior.


3.5 Field Experiments

Roll-over quotas?

Did the cab drivers have a daily quota imposed on them? Did those who owned their own cabs, and thus able to set their own quotas, behave any differently? I wonder if behavior can be influenced by changing the incentives. Sort of like some phone companies let you roll over unused minutes to the next month, would behavior change if above quota performance can be applied to the next day's quota? If the going is easy when it rains, then I think that would nudge drivers to keep working past their quota.
I think the same thing happens in real estate, where there are boom and bust periods. Real estate agents have monthly quotas. I wonder if the data would show similar behavior in a boom as the rainy taxi days?






3.4 The Spectrum from Lab to Field

A/B web based testing
Most top e-commerce web sites do what is called A/B testing. A certain percentage of visitors will be directed to the normal site (control) and a small percentage will be directed to an experiment to test out effect on consumer behavior, mainly did purchase rates go up.




3.3 Laboratory Experiments

Checking balance
Back when we paid with checks, we had to balance our checkbooks by writing every purchase down in the register. This prevented us from bouncing checks (overdrawn). The credit card has removed that feature and credit card companies don't have an incentive to give feedback to consumers. To check one's balance in the US, we have to call the credit card company and navigate a lengthy phone decision tree, remember our PIN, and finally have a computer read the recent transactions. Seems like we need some system of automated feedback of balances and warnings like text messages when discretionary budgets are at a threshold.






3.2 Analysis and Interpretation of Results

I'm regressing a bit
This is my first exposure to such statistical analysis, so it is going to take me while to remember ANOVA stands for analysis of variation. Regression is apparently taken from tendency to regress down towards the mean. Why no acronym for regression, like LR for linear regression ;) Isn't there software for this stuff? Plug in your data and it will perform statistical analysis for you!






3.1 The Basics of Experimentation

Lab experiments outside of the lab?

Pennies a day appears to be another form of disrupting metering. I wonder where the threshold is exceeded where it backfires?
For experiments, I wonder if it would be better to conduct them outside of the university in rooms that are not labeled as "behavioral lab"? I'm also curious how people are recruited for these experiments. Are they mostly students and university personnel and does that skew the results? When experiments are replicated, are these done in other labs or locations across the country and world?




Debate 2

Convenience always wins
Credit cards are a boon to convenience. As Amazon.com has demonstrated, if you make things easier, then people will spend more. Although merchants have to pay the hidden 3% fee, they may make it up in increased revenue from people who wouldn't be able to buy their product if they didn't have enough cash. Also, online purchases would be virtually impossible without some form of non-cash system. This hidden fee enables credit card companies to not charge an annual fee to consumers who always pay off their card. An option for larger merchants is to sponsor their own credit cards. One thing that should be allowed is for all merchants to offer a 3% discount to those paying cash. I think the credit card companies force merchants to disallow this to carry their card.






2.4 Spending Uncash

points

The most opaque system is when prices or money are converted to a different system, which then requires calculation, similar to a foreign currency transaction.
Arcades where you buy cards with points on them and all games cost points which do not easily match up to quarters or dollars.
Loans where you purchases points to reduce the interest rate, but it is difficult to calculate the overall cost of the loan, or that information is provided to you at closing among many papers to sign as recounted in the book "Nudge".
Casinos where you purchase chips to stand in for money, giving you an incentive to spend them all or hold on to them for next time so you don't have the inconvenience of exchanging them back.






2.3 Self-control

gym strategy
One way I consistently make it to the gym is to go before work and wear my exercise clothes. I couldn't show up to work unshaven and in my gym clothes, so that acts to lock in going to the gym. I am lucky that I don't have any early meetings so this works in my case and has become a habit that I don't even think about it.