Neuroeconomics and Innovation: Blogging the Conference
As promised, I’ve written up some of my reactions to papers presented at the Law, Economics and Neuroscience Conference: Implications for Innovation, sponsored by The Southern California Innovation Project, Theoretical Research in Neuroeconomic Decision-making (TREND) and The Center for Communication Law & Policy. There were many terrific paper presentations, but I’m going to focus only on those that seemed particularly applicable to legal academic research.
First, Daniel Houser presented his co-authored paper, Neural Responses to Sanction Threats in a Two-Party Economic Exchange. As Houser explained, recent experimental work and field studies have tended to show that monetary incentives intended to produce cooperation sometimes seem to reduce it. In one well-known example, Israeli day cares instituted a fine regime to deter parents picking up children late from day-care. Surprisingly, the fines reduced, instead of promoted, compliance: researchers theorized that the parents had changed the way they framed the decision to pick up their children. While before, the decision was considered as a part of a social experience, the fine changed it into a mere economic tax. Houser’s experiment was designed to examine the neural correlates of this kind of framing-shift.
The experiment was a modified trust game, and subjects were to decide how much to reciprocate. In some conditions, their cooperation was promoted by a fine for breach-of-trust, and in others, it was not. Houser found that the sanction condition reduced activity in the parts of the brain that involve social relationships, while increasing activity in the parietal cortex, identified as a seat of calculated reasoning.
Similarly, Mat McCubbins presented a paper on the neural correlates of trust, The Effect of Institutions on Behavior and Brain Activity. In the experiment, subjects in a trust game were hooked up to EEGs and then tested on whether they trusted an unknown reporter to accurately report the outcome of a coin toss. Subjects trusted reporters whose incentives were aligned with theirs at equivalent rates to those who were punished when they failed to be trustworthy. However, the EEG results were divergent, suggesting that equivalent decisions to cooperate with others may have significantly different biological roots.
As McCubbins pointed out, EEGs are significantly cheaper and simpler to use as mechanisms for evaluating the brain. I commented on the paper, and noted that political scientists’ use of the term trust seemed to vary significantly from that used by law professors, making it somewhat difficult to see immediate applications for this paper to work on trust and corporate law. McCubbins et al. suggested as an implication of different EEG findings that we are motivated to see as uniquely informative subjects whose interests are perceived to be aligned with ours (as opposed to those who will be punished when they do not cooperate with us). This finding would seem to have useful implications for marketing research.
Drazen Prelec presented Self-delusions, neuroeconomic modeling and fMRI evidence. The experiment was great. Subjects were shown modified characters from the Korean alphabet and asked whether they seemed more male or more feminine. Once each subject had a baseline, they were scanned in an fMRI. In each trial, the subject predicted what the next symbol would be, and was told that they would be paid extra if they guessed “right.” They were then were shown the stimulus, and asked its gender.
Thus, each subject would see a symbol that they were predisposed to think was, say, male. But they had just committed to a guess that it was female, and would be rewarded more if they conformed their view to their guess. The question is what happens in the brain when subjects are lying to themselves. Prelec et al. found that the dopaminergic reward system was activated during these episodes of self-deception/deception. That is, your brain seems to have a small region rewarding you for successfully engaging in self-deceit.
Alan Sanfrey presented his group’s paper, Functional Imaging of Decision Conflict, recently published in the Journal of Neuroscience. In the paper, male subjects were shown pairs of female faces, and asked which face most men would find more attractive. (Apparently men have pretty convergent views of attractiveness, making the experiment possible. Women subjects’ views of male beauty are too divergent to permit this kind of test on them.) The faces had been previously classified as similarly attractive, similarly unattractive, or mixed (one attractive and one unattractive). Not surprisingly, where the faces were similarly attractive, subjects took longer to make decisions. At those moments, subjects’ anterior cingulate cortext (ACC) was activated at high rates. Unlike previous studies of this type, subjects were not asked to vocalize their choices, or implement them by pushing a button, reducing the risk that ACC activation was linked to motor-neuron activation. The study provided additional evidence that the ACC is the locus of decision conflict problems in the brain.
Finally, Colin Camerer presented a series of fantastic experiments about curiosity (The Hunger for Knowledge: Neural Correlates of Curiosity.) Camerer and his co-authors hypothesized that curiosity would be located in the brain in those areas linked with anticipated reward. Testing Caltech students for their knowledge of trivia, the research team found that answers about which subjects were highly curious (such as, for example, which kind of animal excrement is a delicacy) produced activation in caudate area, which has been well-established to be linked to our reward system. (This activity was not linked to correct answers, merely the presentation of the trivia question itself.)
The team followed this finding up by asking whether questions in which subjects had exhibited high-curiosity and had gotten wrong would be linked with greater levels of recall after a decent time interval. They found that there was a strong relationship between curiosity, wrong answers, and recall.
In the plenary session, discussants and presenters talked about the kinds of applications such experiments might have for legal studies, and, in particular, innovation. Many present were somewhat skeptical that neuro-imaging has immediate applications for legal policy, either because (1) the neural bases for behavior rarely matters for law; (2) the research is simply uncooked, and not generalizable, and much imaging work feels like basic science; and (3) law professors, economists, and neuroscientists lack a common vocabulary and a set of cross-disciplinary research questions.
Others, more sanguine, suggested that innovation-promotion in today’s economy rests on expanding networks between people, a topic that neuroscience can inform. Similarly, work on how to effectively promote trust would seem to be one that law has a strong interest in, and which neuroscience would inform. Another scholar, who had presented research on the neurological roots of accounting institutions, suggested that one message of neurological research is that many of our present politico-legal institutions are determined by biological imperatives, thus, institutional change must be carefully considered. Finally, one participant suggested that Neurolaw was at the stage where Neuroeconomics was five years ago: enterprising researchers who want to engage in the work will do so, despite skeptics, and will eventually produce useful results.
I found the conference very illuminating, not least because it exposed me to how hard it is to do this kind of research, and the depth and talent of the teams that are necessary to accomplish it. This suggested to me an immediate barrier and opportunity for legal academics seeking to learn more about the brain. Many law schools today are skeptical about co-authored research, halving the “count” of co-authored articles in promotion and tenure decisions, requiring professors to seek statements from co-authors about their value to projects, and continuing to discount joint authored articles after tenure. (Shucks, even the citation rules are stacked against co-authorship.) But it’s obvious that neurolaw research would require co-authorship, usually in teams of five to seven people. If law schools are seriously committed to this kind of interdisciplinary research, they will have to change their tune and think about new ways to evaluate and reward articles that their faculty had only partial roles in producing.