Sunday, May 6, 2012

The Biological Basis of Economic Preferences and Behavior

The Becker Friedman Institute sponsored an interdisciplinary conference Friday and Saturday titled "The Biological Basis of Economic Preferences and Behavior."

I went to all the talks and some were better than others. Here are some of the highlights and some things I learned (listed by presenter):

  • Coren Apicella (Harvard): Coren's research revolves around the Hadza tribe in Tanzania (a hunter-gatherer society). She had three pieces to her presentation: (1) The endowment effect (the effect where people tend to value things more if they own them) was only present in Hadza clans that had high exposure to outside (western) cultures. She speculated that the endowment effect may come from exposure to market systems where it may give a bargaining advantage. An alternative hypothesis is that strong norms of sharing in Hadza communities may overwhelm the endowment effect (if it is in fact an innate human universal) in isolated clans. There were concerns about selection effects in this research she glossed over since all results were just correlations. (2) Using a public goods game to get a measure of personal generosity, Apicella found assortative matching in the sense that when asked who they would like to live with (and were limited to listing 10 people), Hadza listed people of similar generosity levels. (3) Apicella had confirming evidence that averageness is attractive. (When presented pictures which were composite images of real people, the composite image was rated as more attractive than the individual pictures). The new twist here was that since Hadza were not familiar with what Europeans look like on average, they did not prefer the composite image until they were primed to know what Europeans looked like (they did show the bias right away when shown composite Hadza pictures). So while genes may lead us to prefer average looking people, we need experience to know what average is. 
  • David Cesarini (NYU): Encouraged the use of genetic data in economics as latent/control variables, as a basis for biologically microfounded economics, and to better understand how the environment affects people with different genes differently. He also talked about problems in the medical genetics literature (see previous post). I learned the difference between behavioral genetics and molecular genetics. I also learned what a SNP (pronounced: "snip") is. 
  • Colin Camerer (CIT): Presented results from an experiment on chimps and humans playing various 2x2 matching-pennies-style games. Trained chimps did better than untrained people. Cute result, but there were serious problems with interpretation.
  • Juan Carrillo (USC): He thought about how to model resource allocation in the brain as a principal-agent problem. I learned that when a person multitasks on two consciously difficult tasks (e.g. rotating 3D objects in their head and spelling 7 letter words at the same time), an area of the brain known as the Central Executive System (CES) is activated which manages how scarce resources (oxygen and glucose) are allocated to different systems in the brain. The CES is actually observable in MRI/fMRI scans. In his model the CES was the principal and the systems were the agents, but I did not understand why the systems were selfish (would misreport to the CES how many resources they needed to complete the task). Why would evolutionary forces work at the system level since systems live or die with the brain? I didn't get this modeling choice.
  • Garbriella Conti (UChicago): This was one of the most interesting talks of the conference. She (with Heckman) looked at data from an experiment run on monkeys which changed their social environment when they were young (three environments tested: baby raised by their mother (control) , their peers, or a water bottle (alone)), keeping the physical environment constant. She found that SOCIAL environment affected gene EXPRESSION which affected outcomes such as health and these effects are persistent over time. The economics question they are interested in is: To what extent can the effects of adverse conditions when young (e.g. being orphaned) be reversed when older, possibly by social policy? What blew me away is that one can actually measure gene expression (whether a gene is "up-regulated" or "down-regulated"). Their measure was based on photons of light reflected, so I gather the genes actually glowed differently under some types of light?? I am hazy on how this is done, but that it can be done is amazing! Their results were incredibly clear-cut, too. 
  • Jeff Ely (Northwestern): He tried to use an evolutionary model to describe recent declines in worldwide fertility and why income and fertility are negatively correlated. The big take-away here for me was that evolutionary tools are probably not right for the job (traditional price theory tools do MUCH better) and any explanation that ignores the invention of effective contraception is not going to stand up to scrutiny.
  • Luis Rayo (U. of Utah): He developed a principal agent model of natural selection which was interesting. This model gives a good explanation for why utility functions wouldn't just care about fitness (basically, sometimes it's better to have people care about things related to fitness rather than fitness directly).
  • Arthur Robson (Simon Fraser):  The last slide was the best. He has a picture of the Rocky Mountains from 38k feet. He said that up close individual mountains each have their own unique character and identity, but from a respectable distance away, patterns begin to emerge. This is the real goal of economic modeling: not to get the individual modeled exactly right, but try to get insights into patterns of behavior. I really liked this analogy.

Here are some of my general thoughts after attending the conference:

1. I am surprised how many papers had more than three authors. Non-economists really know how to collaborate.

2. There was nothing at the conference to change my opinion that economics talks are poorly done -- not that the ideas are generally bad, but the way they are presented is bad: cluttered or unorganized slides, bad presentation skills, and incomplete or unmotivated models. I cringe when I think about what the non-economists in the room must have been thinking after a couple of the talks if this was their first introduction to research in the field.

3. I loved that the conference had an interdisciplinary feel. I think there are a lot of unexploited gains from trade between fields.

No comments:

Post a Comment