Introduction to Neuroeconomics (MOOC)

With the full title Introduction to Neuroeconomics: how the brain makes decisions, this course available on the Coursera platform is at the convergence of economics, psychology, and neuroscience. It is presented by Vasily Klucharev from the Higher School of Economics of Moscow (Russia).

Presentation

The course is divided into four modules of a duration of two or three weeks: how the brain works (brain anatomy and functions), how the brain decides (brain models of decision making and choice, neural representation of the subjective value, basal ganglia and choice value), how the brain feels (affective mechanisms of decision making, dual process theory of decision making, decision making under risk), and society of brains (the social brain, taking an evolutionary perspective with the economic animal).

The whole course moves towards higher levels of complexity, presenting the major factors affecting our decisions from single neuron activity to brain regions, to functions as cognition and emotions, up to society and biosphere.

The content of the each week consists of about five 8- to 10-minute videos and an after-video quiz of 5 questions. During each video the student also has to answer an in-video quiz of one question, which does not count for the final grade. The whole lecture notes can be downloaded in pdf form.

As there is no quiz after each video, the in-video question is a good way to keep the learner focused on the video content, even though its apparition could be quite disruptive while the video plays (but a visual indicator on the video timeline indicates when the in-video quiz is going to happen).

Each lecture (or week) ends with a quiz of about 5 questions. They contribute to 45% of the total grade. The final exam, for which we are only allowed one attempt, accounts for 55%. To pass the course, at least 60% of the total points have to be passed, and 80% to get a Distinction.

Notes and thoughts

• By studying when a decision is made at the neuron level through electrical activity even before the level of consciousness is reached, many of the experiments illustrating this course ask very profound questions about consciousness itself and free will. The biological and physical substrate of our consciousness makes us nothing more than biological robots governed by algorithms we are trying to understand.
• But beyond the philosophical questions that arise, we can better understand through the use of fMRI for example how we can make better interfaces to meet the needs of our users (or when neurosciences meets user experience). That branch of neuromarketing is thriving of possibilities, like we can see on Marc Van Rymenant’s blog for instance, where A/B testing is pushed to its limits (as long as ethical considerations are taken into account).
• To go further in the marketing domain, here is some information to think about. The dopamine system, or reward system, is related to the predicted utility. It releases more dopamine, the well-being hormone, hence gives more value to brands associated with wealth and social dominance. And it can be trained by associating brands with specific situations to strengthen this relationship. This is why product placement is so pervasive in movies. Think about the latest James Bond movie: Audi, Beetle, Heineken, Jaguar, Macallan whisky, Omega watch, Sony Vaio… They all want a piece of our brain value system cake, even those not (yet?) associated with luxury values (Heineken, really?). Do not forget that the dopamine system computes anticipated and remembered utility, so if the product behind a brand does not add up to its brand value, the whole brand value is damaged.
• Usually, fast and optimal decisions can be made by following our emotions: emotions are heuristics.
• We can rewire our brain by consciously linking an aspect of the reward token to something calm and unrelated like the ocean, or the breeze in a forest. This self-regulation breaks down the stimulus powered by the nucleus accumbens (which computes the expected gain) by giving the hand to the dorsolateral prefrontal cortex (DLPFC) to counteract the power of emotions. Furthermore, the more the DLPFC is active, the less the ventromedial portion of the orbital frontal cortex (vmOFC) is.
• Race condition in the brain. Two systems are running in parallel to reach a decision: the intuition (fast, parallel, automatic, effortless, associative, slow-learning, emotional) and the reasoning (slow, serial, controlled, effortful, rule-governed, flexible, neutral). Curiously, decisions taken using the intuition path lead to a better decision (in the sense they bring more satisfaction over a certain length of time). We take our decision emotionally, then we add an argumentative layer to prove ourselves we took the right one. If we think too much about a decision, the bypassed emotions will bite us back as we will not find ourselves satisfied about it.
• But the intuition system prefers smaller immediate outcomes, whereas the reasoning system prefers bigger outcomes, even if they come later (temporal discounting question: do you prefer 10 euros today or 11 tomorrow?, work after college or go to university? save for retirement or spend your money today?).
• Prospect theory: we overestimate small probabilities and underestimate large probabilities, which is (partly) why we keep on playing lotteries.
• The value function is asymmetrical, as we feel less joy for gaining a certain amount of money than we feel pain for losing the same amount of money.
• Framing effect: framing options change the subjects behavior. We are risk averse in the gain domain, and risk seeking in the loss domain. In other words, we choose the sure option in the gain frame, and the gamble option in the loss frame. When proposing two alternatives, the one you want to be chosen needs to be presented as a positive one (the number or the probability of people saved in the Asian Disease problem).
• The insular cortex (anterior insula) is involved in the emotional risk processing, the dorsal cingulate cortex (dACC) in the cognitive risk processing. The ventral striatum (NAcc, nucleus accumbens) contributes to the approach to risk (risk seeking), and the amygdala modulates the risk-taking behavior by implementing the framing effect (risk avoidance). All these information affect the DLPFC which finalizes the decision risk processing.
• Game theory lesson: the Tit for Tat strategy outperforms any other strategy in a group (Prisoner dilemma). First cooperates, then do what the other player did on the last round (either cooperate or defect). A group of cooperators always outperforms a group of defectors, even though an individual defector outperforms any other individual in the group. And do not forget the shadow of the future parameter: the higher the probability for the players to meet again in the future, the better they should cooperate.
• Mutual cooperation is associated with consistent activation of brain areas linked with subjective values (nucleus accumbens, orbitofrontal cortex) and we feel better when we cooperate, while unreciprocated cooperation activates the insular cortex and is aversive (an effect that can be tricked by using oxytocin, a neuropeptide that shortcuts the amygdala and increases trust between individuals). In other words, we are wired for cooperation.
• How to detect the intentions of other players? The mirror neurons mechanism runs an internal simulation of the perceived emotion of others.
• Observed fair players are empathically preferred to unfair ones (more agreeable, pleasant, likeable and attractive). But unfair players elicit pleasure in their punishment (observed simulated pain), especially for men. As a single individual defector can break the cooperation in a group, they have to be punished to sustain the cooperation within the group (altruistic punishment: the punishment is costly for the punisher who gains nothing in it). Our brain is designed by natural selection and evolution to optimize our decisions in a very complex social context. That is why the behavioral activities of the key brain regions involved into the decision making is heavily modulated by this specific social context.
• Cooperation has an ontogenetic and evolutionary origin: human toddlers, chimpanzees and other great apes cooperate naturally. But chimpanzees prefer the option of solo actions, whereas 3-year-old children choose the cooperative option.
• Biological market theory: cleaner fishes at their station and their clients, either roaming (they can change station) or residents (they can’t), either predatory (if cheated upon, they can eat their cleaner) or not (no retaliation possible). Roaming clients are served first, and predators are not cheated upon: the cooperation in nature matches the market theory expectations.
• Law of supply and demand in baboons society: females with no children groom females with infants to interact with their baby (they clean their fur of their parasites). The supply is infant time, while the price is grooming time. The less infant in the group, the longest grooming time is required to get access to them (market theory predicted animal behaviors). In other words, the exchange of commodities in primate groups is a trading on a market with exchange rates fluctuating from day to day depending on supply and demand.
• Capuchin monkeys behave like humans in monetary markets when exchanging monetary tokens for goods. They also share the same fundamental biases that humans display, like the framing effect (gambles are evaluated in terms of arbitrary reference points) or the endowment effect (overvaluation of objects we own other objects we do not own). They are also averse of unfairness as they reject unequal pay. They branched out from humans on the evolutionary tree about 35 million years ago, but we share this evolutionary heritage and are programmed by evolution.

To sum it up

I took this course out of curiosity to better understand how the brain works, and to get some neuromarketing insights. But this course went way further than any direct usefulness as it raised deep questions about how evolution and biology govern our strategies, deeply rooted in our neuronal circuitry. It is thus very interesting to follow, with a lot of scientific experiments and papers to dive into, and its multidisciplinarity brings many lights from many directions, each of them very promising.

Suggested reading

Neuroeconomics, Judgment, and Decision Making, Psychology Press (2014)

Neuroeconomics: Decision Making and the Brain, Academic Press (2013)

Decisions, Uncertainty, and the Brain: The Science of Neuroeconomics, by Paul W. Glimcher, The MIT Press (2003)

Introduction à la neuroéconomie (MOOC) (in French)
Introducción a la neuroeconomía (MOOC) (in Spanish)
Introdução à neuroeconomia (MOOC) (in Portuguese)