Tuesday, April 30, 2013

Choice Architecture

One of the most obviously useful applications of behavioral economics falls under the name choice architecture. In short, choice architecture is the study of crafting options in such a way as to guide deciders to a preferable outcome.  This new sub-discipline has been used to great effect already in a number of nations.

For example, most people would agree that higher organ donation rates are preferable for society as a whole. Yet despite large investment in many European nations organ donation rates remained quite low.  Meanwhile other nations seemed to have near 100% donation rates without much effort. Many studies we conducted on how to best increase public subscription to donation programs but without much result. Cultural, economic and even religious factors were studied but seemingly had little effect. Even nations that were geographically, economically, and culturally quite similar often had wildly different rates (as in the case of Denmark, The Netherlands and Belgium).

Ultimately the answer is likely so simple it was overlooked.  In nations where organ donation was the default option organ donation rates were very high. Nations where organ donation was not the default meanwhile had very low rates.

Not every application of choice architecture is so overt however. A study conducted by Eran Doyan and published in Judgement and Decision Making found that even the most common choices can be easily manipulated.  Doyan discovered that by rearranging items on a menu he could regulate how often certain items were selected. Items placed at the start or end of the menu were selected far more commonly than items placed more towards the middle. It may seem like a trivial choice to try and influence, but to a restaurant owner trying to maximize profits (or a society trying to minimize obesity) it may make a large difference.

The end goal of most choice architecture is to help the public make the right decisions without eliminating their options.  For example, for most people contributing to a 401k plan is a wise decision. However, forcing employee contribution isn't an idea many people would support.  The question then becomes how can we influence people towards contributing to their 401k without removing their choice completely?

Let's examine two scenarios.  Imagine you were told you needed to make your decisions for your retirement savings soon. The human resources specialist visits you and emphasizes the importance and urgency of the decision. You're given a packet of information containing descriptions of a dozen options and how each will influence your eventual retirement or lack thereof. Then, their obligation discharged the human resources employee leaves never to return.  Do you think you're very likely to inform HR about your decision, or even make a decision at all? If you're like a large portion of people the packet of information will be pushed aside for a tomorrow that never arrives.

Contrast with this scenario. The HR specialist comes by your office with a clip board and says, "Hey, do you want to contribute to your 401k?" It's an easy question so you of course respond that you would. The specialist makes a note on their clipboard next to your name and says, "Great, here's a packet of information for your options. You should probably read it and let us know what you'd like to do. If you don't have any particular opinion we'll automatically enroll you in our default contribution program at the end of next week. That program is outlined on the top piece of paper in the packet."

The approach accomplishes a number of things. First, the employee is enrolled in a program that is likely much better than nothing.  Second, there is very little pressure put on the decision. If the employee wants to do something other than the default that's great, but otherwise they know they're more or less taken care of. Finally, by forcing a choice (either take the default or choose your own) the employee is less likely to delay their selection. The employee receives the same information and the same options but by having them agree to a default before they're presented with a complex choice their likelihood of contributing skyrockets.

Choice architecture is a powerful tool for influencing how people make decisions as well as what they choose. When used correctly it can increase efficiency, guide the public to self beneficial decisions, and improve society as a whole. Not only that, it's cheap, easy, and effective when implemented.

More economics next week. Until then stay safe and rationale.

Wednesday, April 24, 2013

Reinhart and Rogoff

The plan, dear reader, was to dedicate this week to an exhaustive economic analysis of raisin production in the United States. I have no doubt that you would have been mesmerized by the flood of dry grape knowledge that would have assaulted you. Alas, current events being what they are we must discuss something all together less compelling.

If you're not an economist or politician you've probably never heard of Reinhart and Rogoff.  In brief they are two Harvard professors who published an influential paper in 2010. The paper dealt primarily with how debt levels are related to economic growth. They concluded that when a nation's debt exceeds 90% of GDP economic growth slows dramatically (and in fact becomes negative generally).  Obviously if you owe nearly as much as you earn annually and you earn less and less each year (negative growth) it becomes very difficult to ever escape the debt trap.

As it turns out a simple spreadsheet error resulted in Reinhart and Rogoff's conclusions being quite different than reality.  Although debt levels are weakly negatively correlated with growth, there is no evident tipping point at a 90% debt to GDP ratio.

Now the real question is, "Why should anyone care?" After all there are certainly dozens if not hundreds of other economic papers out there with similar (though hopefully unintentional) errors. This one may be more prominent than the average but it hardly seems worthy of a great deal of attention.

To be quite honest, the error in Reinhart and Rogoff's work is not terribly significant. It is however very symbolic.  As everyone knows the last several years have been difficult for the global economy.  The US housing and credit crisis gave way to problems with government debt in the eurozone. Uncertainty in China is growing and governments have been trying to figure out the best way to forge through their difficulties.  In the United States we opted to stimulate the economy by easing credit markets and stimulus spending. In Europe many nations decided to sharply reduce government spending in order to bring debt levels down.

The decision of many of the European nations was based upon the idea that high debt levels would halt economic growth.  The exact same conclusion that Reinhart and Rogoff came to in their paper. Were the decisions of the Europeans governments based upon the data found in Reinhart and Rogoff's work? Probably at least in part, though hopefully not completely. Now that the paper has been shown to be incorrect doubt is cast upon austerity measures as a whole. Or so it would seem.

The fact is that most American economists already believed that austerity measures in Europe were not effective. There's been a gradual trickle of data from European economies showing that large reductions in government spending haven't improved much economically speaking and have often left the public worse off. Thus a paper that supports austerity tactics turning out to be inaccurate isn't really much of a surprise.  After all, we can see real world data that indicates austerity doesn't seem to improve economic outlook. Compared to that data an error in a pro-austerity paper is relatively inconsequential.

However, Reinhart and Rogoff's error is an easily digestible fact for the general public. Anti-austerity economists can crow and point the finger at how wrong austerity is without having to result to time consuming activities such as modeling and data collection. Armchair politicians can knowingly cluck their tongues at dinner parties and tell their longtime right wing nemesis "I told you so." And everyone can feel very well informed because they know who Reinhart and Rogoff are now.

Ultimately not very much has changed from a week ago.  Two professors are a bit embarrassed, one graduate student has earned a bit of fame, the European austerity experiments proceed apace and no one really knows any more than they did before a spreadsheet error was discovered. While the signs don't seem positive for austerity proponents, only time, not spreadsheets will tell who was right in the end.

More economics next week. Until then stay safe and rational.

Tuesday, April 16, 2013

On Terrorism and Knee Jerk Reactions

Yesterday Americans were stunned by explosions in Boston that killed at least three and wounded dozens more. It goes without saying that my sympathies are with the victims and their families. I wish the survivors a speedy recovery and give my condolences to the families of the lost.

At this time law enforcement and government officials are scrambling to determine what happened and who is responsible. Soon enough politicians will turn their minds to what can be done to prevent such a catastrophe from happening again.  I'm not going to offer any opinions on this most recent tragedy. The wound for many is still too fresh to be a safe topic. Instead I want to touch upon terrorism (or any other similar preventable event) and why I feel that as a nation we handle it poorly.

Economists are often asked to quantify things, such as the value of a human life, that seem to defy measurement.  Many people react to such a notion with horror. Life is sacred and can't be reduced to a simple number on a sheet of paper. Trying to transform a person into a dollar amount is objectification at it's worst.  And maybe they're right.

However, here are some numbers that maybe show why I believe such analysis is important.

In 2012 the Transportation Security Administration had a budget of approximately $7.5 billion. Very nearly all of that was spent to make sure people didn't die in airplanes. Before the creation of the TSA, approximately one hundred people died annually due to commercial plane crashes in the United States. Actual numbers vary quite a bit but one hundred per year is a good average for the 90's. Since then that number has dropped rather significantly. There is a compelling case to be made that this reduction in fatalities is due to technological improvements rather than the TSA, but for now we'll assume that the TSA is saving approximately a hundred lives per year. That's $75 million dollars per life saved.

That's a college education at a state university for almost eight thousand students per life.

That's enough to essentially eliminate hunger in America. source

That's enough to provide blood pressure medication to one hundred twenty five thousand at risk adults per life (at $600 for medication annually).

It's hard to imagine that any of these three alternatives wouldn't save more lives than the TSA does annually. Yet in the wake of the September 11th tragedy we created a large government organization that will very likely be a drain on our resources for decades to come.

The fact is it's impossible to prevent every conceivable disaster. Making decisions regarding the allocation of scarce resources when emotions are high simply leads to situations where you spend a lot and accomplish very little. When a bomb goes off we scramble for more security screenings that generally won't find bombs. When a school shooting occurs we try to put more restrictions on guns that don't stop school shootings. When a hurricane hits we demand better preparation that won't prevent hurricanes. The reality is the marginal benefit from more and more preparation or security declines very rapidly. Reasonable precautions should be taken and enforced, but beyond that resources are better used elsewhere.

The most effective preventative when it comes to terrorism isn't something you can buy.  It's an aware and vigilant public that doesn't simply ignore the people around them. Unfortunately our culture has grown in such a way that pretending everyone else doesn't exist has become a social norm.

I hope that the public servants and politicians of Boston learn what they can from this tragedy and then make wise and prudent decisions moving forward. I hope that they choose to institute policies which protect the public without wasting their time and tax dollars. I hope that they allow time for the emotion of the moment to subside before reacting with knee jerk legislation that will be difficult to retract.  Unfortunately through experience I've found these hopes are likely in vain.

More economics next week. Until then stay safe and rational.

Tuesday, April 2, 2013

Matters of the Heart

What do you imagine when you think of your perfect partner? Are they rich? Good looking? Funny and outgoing?  Would you be surprised if you were told none of those traits mattered as much as you thought they did?

A great deal of research has gone into determining what men look for in women and vice versa. Of course everyone wants someone who is attractive, engaging and reasonably self sufficient. But how important different characteristics are to each gender has seemingly varied quite a bit. Men for example generally claim that looks are more important than earning potential while women find potential wealth more compelling. Both men and women find warmth or friendliness approximately as important as the other gender.

However, these preferences are based on self reported data provided by study participants who are not actually interacting with potential partners. Most studies are conducted via a simple survey or by asking participants to rate pictures and profiles of potential mates. When participants rate real people they've actually met their preferences are quite different.

Recent studies have used a clever method to reveal participants true preferences. Participants were divided into two dating groups. Each group read profiles and saw pictures of the other group and then rated them for  a variety of factors including looks, earning potential and warmth. This provided a baseline rating for each individual without personal engagement. Groups then attended speed dating sessions within their own group (who they had not seen profiles for). After the speed dating sessions they rated their own group on the same metrics as the other group as well as listing which dates they preferred to see again.

If everyone's stated preferences held true we would expect that men would heavily favor more attractive women and women would heavily favor wealthier men. However, the results did not support this hypothesis.  Instead men and women did not show substantial differences between their preferences for attractiveness, wealth or warmth.  In other words although men and women both preferred attractive counterparts neither seemed to prefer attractiveness more than the other gender. The same was true for earning potential, friendliness and a variety of other factors.

The interesting part about this phenomenon is that both the stated preferences and revealed preferences still have significant influence on the mate selection process. However, each preference set is influential at a different time. Stated preferences dominate during the initial "weeding out" phase when individuals must narrow the field from hundreds to a few dozen. Later, revealed preferences are impactful when settling on a certain individual.

Imagine a man or woman attempting to find a potential spouse on an online dating website. During the initial stages when he or she is deciding who to contact for a date their stated preferences are dominant. Men will select women based on looks more than women. Meanwhile women will select men based more on potential future earnings than men. However, once they actually meet their counterpart both genders will weigh such things as looks and wealth equally as compared to the other gender.

So what's the lesson to be learned? Most importantly it would seem that as individuals we often don't know what we really want. In a more practical sense it would seem that exaggerating a bit on a dating profile (or using a flattering picture) might be worth it in order to get past the initial selection stage as long as you don't get categorized as being deceptive.

More economics next week. Until then stay safe and rational.