Good Morning, Economics

There are many ideas floating around for reforming existing institutions so that they produce significantly better outcomes. The three ideas for reforming governing institutions that I know of are Predictocracy, Futarchy, and Professional Voting; I am sure there are more, and I know there are a lot more for reforming other institutions. However, there is a noticeable lack of experimentation with type of idea. Does it make sense not to experiment? Let’s do a few back-of-the-envelope net present value calculations to see if it does or not.

Let’s say that we judge Futarchy to have a 1% chance of working. Let’s also assume that if we experimented with it now it would take 30 years for the benefits to appear, that it would save everyone in the country $100/year (a conservative guesstimate, remember this is contingent on Futarchy “working”), and that the total investment would be about $100 million (surely a massive over guesstimation). Using a discount rate of 10%/year, the Net Present Value of a $100/year stream appearing after 30 years is $57 (I did the calculations). There are 300 million people in the country so the total Net Present Value is ( 1% x $57 x 300 million ) - $100 million = $70 million. So even under these calculations, the expected benefit of researching Futarchy is massively positive. Not experimenting is huge loss.

The thing that makes this investment so good is the fact that there are a lot of people in the country. Finding an institutional reform that makes peoples lives better provides a massive public good, but costs a fixed amount to find. In these calculations I have only included the people in the country, the expected benefit is an of magnitude larger if we consider the benefits to everyone in the whole world.

I thought that Cass Sunstein’s EconTalk podcast on worse case scenarios had an interesting discussion of discount rates (this section starts at about minute 55 and it is about 9 minutes long). He argue’s that there’s a difference between discounting lives and discounting money because money has opportunity cost; you can invest money and make it grow, whereas you cannot do the same thing with lives. Eliezer Yudkowsky has argued similarly that we should not have a discount rate in our pure preferences (Yudkowsky makes this distinction clearer than Sunstein). However, Robin Hanson dissented.

I tend to side side against discount rates in our pure preferences. Even if people act like they have discount rates in their pure preferences (note the lack of charities trying to help the future), as Hanson points out, perhaps they would if they understood how easy it is to help the future.

I have noticed something about myself in the last year or so, I fear being proven wrong. When I see some study on a topic about which I have a prior opinion, I get a light dose of fear. To give a more specific example, after I finish writing a post for this blog, I consider reading it over in order to revise it. My immediate reaction to that suggestion is to fear that my belief that I have written an eloquent post is wrong. Ironically actually discovering errors does not cause me much pain.

This seems like a clear case of confirmation bias, and being biased is bad. Bias leads to costly mistakes, like writing bad posts and making unwise investments. Luckily, I have come to recognize this feeling of fear relatively easily, and recognizing it helps me to identify when I might be prone to confirmation bias and to lookout for possible disconfirming evidence. This feeling of fear is my own special marker for confirmation bias.

When I recognize the feeling, I know that I should think carefully about my choices about what evidence to seek. I make sure not to avoid looking at evidence simply because it might force me to change my beliefs, and I often try to actively seek possible disconfirming evidence.

I am sure I don’t always notice my fear, and overcoming even light fear can be difficult, but I am glad that confirmation bias has a relatively clear marker. I don’t know if other people get this same fear emotion, but I wouldn’t be surprised if it is a general phenomena.

The Commodity Futures Trading Commission is calling for public comments on possible regulation of “Event Contracts.” As I understand it, this is good news for prediction markets; there is a lot of uncertainty about the regulatory status of prediction markets, and this seems like it would clear that up. A clear regulatory framework would pave the way for real-money prediction markets. At least, as long as event contracts are not regulated out of existence. It is my impression that this is a very good thing, so I am very excited.

I will be submitting comments via e-mail (secretary@cftc.gov), and I encourage others to do the same.

The gist of my comments is going to be that I see the potential for prediction markets to help regular people, who don’t have access to a lot of data and analysis, make decisions. For example, prediction markets about the future of renewable resource technologies could help engineering students determine whether they should invest resources in learning about those technologies. Prediction markets about the economic future of specific geographic areas could help young workers decide  where to move to get good jobs. Therefore, I would like to see the prediction markets relatively unlimited.

via Midas Oracle

For the last few months or so, I have been eating lunch on the Ave instead of bringing food from home. The Ave is a long street two blocks over from the University of Washington campus that has a lot of restaurants on it. Many of them are quite good.

Eating on the Ave regularly led me to think about how competitive the lunch food market on the Ave is. There are a lot of restaurants, so it is pretty competitive, but it could probably more competitive. This led me wonder why Seattle is almost devoid of street vendors (there are a few downtown). This article explains it

Back in the 1970s, our fair city decided “clean streets” meant enforcing the stiffest laws in the country regulating street food vendors.

[...]

I recently called the health department to inquire about opening a French-fry cart, the importance of which became apparent to me as a teenager in Amsterdam. The hardened municipal worker on the other end of the phone informed me that if I didn’t see it on the streets, it could not be done. When I decried Seattle’s embarrassing lack of street food variety, she suggested I “move to France, where their food poisoning rate is consequently higher.”

It is unfortunate that selling food on the streets is so heavily regulated. Allowing street vendors should reduce the price of lunch foods by reducing their operating costs. Street vendors have lower operating costs since they do not have to rent expensive storefront space, although they do have to rent or buy licenses. Allowing street vendors would also have the added benefit of reducing the number of storefront restaurants, which would free up storefront space for other things. I would like to see Seattle auction off tradeable street vendor licenses. Auctioned permits would also allow easy to vendor regulation since their licenses could be revoked.

I started reading Overcoming Bias in the last few months, and Eliezer Yudkowsky has more or less convinced me that uncertainty is a property of your mind, not reality. Because of this, I really liked this passage in The Black Swan (p. 198 )

Often, in conferences when they hear me talk about uncertainty and randomness, philosophers, and sometimes mathematicians, bug me about the least relevant point, namely whether the randomness I address is “true randomness” or “deterministic chaos” that masquerades as randomness. A true random system is in fact random and does not have predictable properties. A chaotic system has entirely predictable properties, but they are hard to know.

[...]There is no functional difference in practice between the two since we will never get to make the distinction– the difference is mathematical, not practical. If I see a pregnant woman, the sex of her child is a purely random matter to me (a 50 percent chance for either sex)– but not to her doctor, who might have done an ultrasound. In practice, randomness is fundamentally incomplete information.

[...]Randomness, in the end, is just unknowledge.

I am taking Intermediate Microeconomics at school, and on the first day of class, my professor brought up the pervasiveness of preference uncertainty. I had not given preference uncertainty much thought before, but I have noticed that I am more uncertain about my own preferences than I had realized. I am frequently uncertain about my lunch preferences. Some of this uncertainty may be exaggerated because the situations in which I notice I am uncertain about are those situations where my preferences are not strong. However, I also sometimes realize that deciding to walk elsewhere to buy lunch to save money, that the decision was not even close to worth it. I don’t think I would make that mistake if I had a better grasp on my preferences .

Incidentally, I recently read Predictably Irrational (which was OK), and I noticed that many of the mistakes and biases that the book describes seem to stem, at least partially, from our uncertainty about our preferences. For example, the book discusses our tendency to ‘anchor’ to initial prices. We tend to judge prices relative to the first prices we first observed for a particular product. If we had a good understanding of our own preferences, that wouldn’t happen very frequently.

I am not sure what general implications preference uncertainty has, but it seems useful to keep it in mind when making decisions.

One of the most interesting concepts I learned from Michael Abramowicz’s Predictocracy (which I have mentioned before) is the idea of a “normative prediction market”.

The essential distinction between normal conditional prediction markets (about prediction markets in general) and normative prediction markets is that for a normal prediction market, the contract payout depends on some objective criteria, like whether Barrack Obama wins the Democratic nomination or what the unemployment rate is in 2013, but for a normative prediction market, the contract payouts depend on the the subjective judgment of a person or organization. The subjective judgment could be that acquiring that small startup was a bad idea for Google or that introducing needle exchange programs was a good idea.

The major benefit of using subjective criteria to decide prediction market contract payouts is that the market predictions are more useful because they more precisely target the questions of interest. Normative prediction markets are less biased than normal prediction markets because they do not omit important but hard to quantify effects (which seem commonplace). Normative prediction markets can also do a good job of  aggregating preferences in some cases.

Uses of normative prediction markets

Normative prediction markets can potential help companies make good decisions. A manufacturing company might use a normative prediction market to predict whether some major internal change, like moving one of their manufacturing divisions overseas, would be regarded as a good idea by a judging committee in the future. The company would commit to setting up a committee in the future which would report on whether the change was a good idea. The committee would have a relatively open ended criteria but would likely consist mostly of cost benefit analysis, and prediction market participants would have to predict what factors the committee would think were important.

Normative prediction markets also have a lot of potential in public policy. For example, an executive agency or an interested non-profit could sponsor a normative prediction market to help evaluate whether it would be a good idea to implement some new type of poverty relief program. Prediction market contracts would pay out based on whether a decision judge, a randomly selected agency employee or non-profit board member, would announce that it was a good idea or announce that it was a bad idea to implement the program.

Payout rules

Different rules can be used to select the body responsible for deciding the variables that determine prediction market contract payouts. One method is to determine the payout based on a judgment made by an individual randomly selected from a defined pool. This method would be relatively cheap and the resulting market prediction should be quite stable because prediction market participants must average the expected judgments of the pool of decision makers. This averaging can be useful if one is interested in making decisions partially based on diverse preferences, not just diverse analysis.

Another method is to base the payout on a judgment made by a committee. Committees would generally give payouts with less variance than randomly selected individuals, which would make participating in the market more attractive because it would lower the risk involved. In some cases committees might also produce higher quality judgments; for example, when back and forth argumentation is important for producing good judgments. Of course, committees would also be more expensive since they involve more people and have other potential problems.

Potential problems with normative prediction markets

Shirking by decision judges is an important potential problem for normative prediction markets, because decision judges do not directly influence the final decision (since it has already been made). Thus, they will have lower incentives to make informed and well thought through decisions than traditional decision makers. This would reduce the quality of the predictions made by prediction markets.

Norms about not shirking and monitoring of decision judges could reduce shirking. However, decision judges should in general be harder than monitoring traditional decision makers because market judges do not directly change real outcomes, which eliminates one method of assessing decision quality.

One normative prediction market feature that could be either harmful or beneficial is that decision judges are likely to give much higher weight to higher order principles and ideology than traditional decision makers. Because decision judges who determine the contract payouts do not determine the decision that was actually made, if they care about other prediction market outcomes in the future, they have an incentive influence how market participants perceive the average decision judgment in order to shift that average in the direction they favor. This may include giving more more extreme judgments than they really prefer.

Consider a situation where the government uses a normative prediction market to decide how much it should spend on a new social programs. The prediction market decision judges are randomly drawn from body of liberals and conservatives. For a particular decision, if a liberal decision judge (a judge who’s spending judgments are generally higher than the mean of the group’s judgments) is selected, then they have an incentive to try to move the group’s mean judgment upward as much as they can by announcing a very high spending amount, even if that amount is higher than the amount they would choose if they were the actual decision maker. The reverse would be true when conservative decision judges are selected.

Strong norms pushing people towards giving genuine preferences would mitigate this effect somewhat.

The beneficial aspect potential of this effect is that decisions judges would generally be less tempted than traditional decision makers to trade off reduced rule of law or other high order principles for better immediate outcomes.

Abramowicz emphasizes this second potential, but on net, I think this effect would be more harmful than beneficial.

Over all, I think normative prediction markets are a very cool idea. They definitely deserve experimentation, and I suspect that they will eventually be used to make many types of decisions.

Links: Abramowiczs description of normative prediction markets

Bryan Caplan asks whether statistical discrimination prevents good products from being adopted by retailers

[I]magine this hypothetical. Suppose you have a genuinely new and improved t.v. which would be profitable to manufacture if you had a serious order from Best Buy. What could you do to get Best Buy to start carrying it? How would you even get Best Buy’s buyer to take your calls? Could statistical discrimination (most people like you are too useless even to talk to) keep a good idea off the market for good?

The solution to this problem seems simple: retailers should charge a submission fee to review products for which there is little information. This would weed out people who are not confident in their own products and cover the cost of reviewing products that are not eventually accepted.

In the case of books (which Bryan mentions), a $300 submission fee would certainly cover the cost of having a competent person research whether the book should go in the store. This function could probably even be outsourced to a separate company. I am surprised that booksellers do not already do similar things. Am I missing something? Or do retailers already do this?

Will Wilkinson has a great bloggingheads.tv interview with Dan Ariely, author of Predictably Irrational: The Hidden Forces That Shape Our Decisions, which is about behavioral economics and the effects of human irrationality. Listening to the interview convinced me to read the book. Ariely also has a blog, but I have not found it very interesting.

One segment (”Throwing away the keys to your own chastity belt” starting around minute 42) I found particularly interesting was the section where Ariely discusses procrastination, and specifically one mechanism to get students to procrastinate less. Ariely mentions a teacher for a class which has three papers which will not be graded until the end of the quarter. However, the teacher offers the students the ability to pre-commit to turning the papers in on dates of their choosing. Students who turn in their papers after their dates loose points for each day late they turn their paper in. Students who are completely ‘rational’ would choose dates at the very end of the quarter, but work on their papers throughout the quarter none-the-less. Unlike these students, real students, who are well aware of their inclination to procrastinate, would choose dates which are somewhat spaced out, in order to motivate themselves to work on their papers in a timely manner and thus produce higher quality work and expose themselves to less stress.

I think this is a wonderful idea. I struggle with procrastination, and I know a lot of bright students who also struggle with procrastination to some extent. Thus, I appreciate teachers who give weekly quizzes and mandatory homework, because they help me time my review of class material appropriately. This scheme would do something similar.

One potential problem is see with this scheme is that it might be a bit time consuming to administer. Keeping track of students’ individual due dates and when they actually turned them in could consume a many hours of a teacher’s time if they had to do it all manually. Luckily, it also seems like it would be fairly easy to automate. A software program could take students due dates and their actual turn in dates, and compute the loss of points for all students. The teacher would simply need to pass around a form in the beginning of the class for students to assign themselves due dates, and enter those into the program. Then as the students turn in their assignments, the teacher would enter in when each paper was received, and the program would generate point losses for every student at the end of the quarter. The system could also be set up to send out personalized e-mail reminders to the students.