paul-glimcher

Paul Glimcher

Thinker

A ‘Brainy’ Approach to Economics and Psychology

Intro

Paul Glimcher is a neuroscientist, economist, psychologist, innovator and entrepreneur. His myriad different fields and interests eventually pushed him to help develop a new field: neuroeconomics. His revolutionary interdisciplinary approach has allowed him to put forth multiple theories about how humans behave and make decisions. Glimcher has become a greatly influential figure in the field of behavioral science as a result of the research he has conducted. He has written a plethora of inspiring articles and books and is perhaps most well-known for his contribution to the textbook Neuroeconomics: Decision Making and the Brain, the first text to examine the science behind economics.1 The textbook is now in its second edition and has become a standard reference for the field of neuroeconomics.

Paul Glimcher believed decisions could not wholly be explained through economic or mathematical models and thus birthed his atypical approach of using neuroscience to better account for the ways humans make decisions.2

A close relationship between the theory of economics, the theory of psychology, and the theory of neuroscience could be forged – but it would have to be forged by a partial reduction… as we adjust the conceptual objects in each discipline to maximize interrelations between disciplinary levels of analysis, we achieve both a formal reduction of two theories and a broadening range of the predictive range of both theories.

– Paul Glimcher in Foundations of Neuroeconomic Analysis

Subscribe to get notified when a new podcast episode is out

The Field of Neuroeconomics

Prior to the emergence of neuroeconomics, decision-making models were not concerned with the brain processes involved in particular behaviors. Instead, they were based on the economic principle of homo economicus, a hypothetical model that assumes humans make their decisions in a purely rational way that maximizes utility. Behavioral economics, which departed from traditional economics, understood that people did not actually act according to perfect logic and instead began to conduct research into the cognitive biases that impact decision-making. Yet still, behavioral economics did not look into the neural processes underlying choice.

For some time, the failure of economics to take into account neurology existed because people didn’t believe it would be possible to study the brain’s intricacies. As technology advanced, neuroscience became a means through which to examine the brain.3 The field of neuroeconomics tried to fill the gaps of both traditional economics and behavioral economics by discovering a relationship between economic decisions, psychology, and neurology.3

As these fields began to intersect, Paul Glimcher took up his position at New York University as a Professor of Neural Science. His 1999 paper, co-authored with psychologist Michael Platt, “Neural correlates of decision variables in parietal cortex” is often credited as the first published paper of neuroeconomics. This paper proposed that based on the eye-movement response of monkeys after they received a reward, the activity of neurons could be predicted when monkeys were making decisions.4 Herein began Glimcher’s career-long interest in reinforcement learning.

Glimcher’s 2003 book, Decisions, Uncertainty and the Brain: The Science of Neuroeconomics was one of the first books to be published with the term “neuroeconomics”. In this book, Glimcher outlined his belief that if we came to understand the neural pathways that control our bodies processes and responses, we would be able to predict our actions.5 He acknowledged the insufficiencies of neuroscience as it stood, suggesting that it did not adequately explain complex behavior. Neuroscience needed to also incorporate psychology and economics to help explain humans’ more complex decision-making processes. He believed that neuroeconomics would be able to replace traditional economic models by determining a mathematical model to predict the optimal course of an action for decision-making.5

Paul Glimcher’s contribution to the field also came from his innovative use of scientific technology in conducting his experiments. He used fMRI brain-scanning technology and other advanced techniques to help create explanatory models which he hopes will evolve into predictive models of real-world behaviour.6

Reinforcement Learning and Dopamine Neurons

As children, we learn largely through reinforcement learning—keeping track of which actions will be rewarded and which actions will be punished allows us to understand the basic social norms of how to behave.

When our actions are rewarded or ‘reinforced’—through positive feedback, good marks in school, or others forms of praise and approval—we tend to continue doing them. In this way, praise and appreciation functions as a form of teaching.

In scientific terms, reinforcement learning is viewed as a machine-like model: through trial and error with an environment, people learn in the forms of rewards or penalties from trying different ‘inputs’ and ‘outputs.’ Based on what it has previously experienced, the agent will predict the level of reward it will obtain for performing a particular action and then compare its prediction to the actual level of reward it obtained in order to adjust its behavior. The margin of difference between the predicted reward and the actual reward is called the ‘reward prediction error’.9 Reinforcement learning can be applied to computers, animals, and humans.

Paul Glimcher wanted to better understand what caused reinforcement learning to work. He was interested in what happens in the brain when rewards or punishments are received. To gain insight into the brain’s reward system, Glimcher conducted an experiment with fellow research Hannah Bayer in 2005.9

Specifically,10 Glimcher wanted to see if dopamine—the ‘pleasure’ hormone—was involved in reinforcement learning, specifically the reward prediction error.10 He believed that when an animal received an unexpected reward, and thus there existed a reward prediction error, a rush of dopamine would occur.9 This makes sense given the special pleasure we feel when presented with an unexpected gift, promotion, or outpouring of love.

To test this theory, Glimcher and Bayer examined the dopamine neurons in primates while they conducted a task that would allow for reinforcement learning and unexpected reward. They found that the firing rate of dopamine neurons post-reward “accurately carries information about positive reward prediction errors but not about negative reward prediction errors” (132).9

Through this experiment, Glimcher and Bayer were able to show that dopamine plays a role in reward, and that reward plays a role in decision-making. This multi-discipline approach was a common method Glimcher used to try and understand the neural events underlying decision-making.

History

Paul Glimcher was born on November 3, 1961 in Boston.11 His father, Arne Glimcher, was the founder of the Pace Gallery, a renowned contemporary art gallery.12 Paul, however, did not have the same artistic thumb as his father and instead was far more interested in science and technology.

The Glimcher family later moved to New York where Paul attended the prestigious Dalton School before going on to complete his B.A. at Princeton University.6 In 1989, he received his PhD in Neuroscience from the University of Pennsylvania. It was the first neuroscience degree the university had ever awarded.13

During his doctorate, Paul Glimcher had the opportunity to work under another distinguished neuroscientist, Randy Gallistel, who also took an interest in the intersection between psychology and neuroscience.6 Gallistel’s work in psychophysics helped forge Glimcher’s path into the field of neuroscience and he chose to do his postdoctoral degree under David Sparks. David Sparks was a neurophysiologist that studied the function of the brain and the nervous system.14

Following his research under Sparks, Glimcher was appointed Assistant Professor of Neural Science at New York University in 1994. During this time, the field of neuroeconomics was beginning to emerge as insights into economics, psychology and neuroscience were intersecting.11 Glimcher went on to become the founder of New York University’s Centre for Neuroeconomics in 2004.13 This was the first academic society solely dedicated to the field of neuroeconomics, making Glimcher a pioneer of the field. In 2008, Glimcher became a Fellow of the American Association for the Advancement of Science.6,15 Fellows of the AAAS are distinguished scientists and innovators, recognized for their achievements in research and excellence in interpreting science to the public.14

Today, New York University’s center is known as the Institute for Study of Decision Making to promote an interdisciplinary approach to studying humans’ decision-making that more closely accounts for the kind of work done in behavioral science.16 Glimcher is still a Professor at New York University, where he continues to focus on his long-term goals of understanding the neural events that underscore behavioral decision-making.17

Glimcher also began another project in 2014 called the HUMAN project. The project is a ‘big data’ longitudinal study of the behavior and biology of thousands of New Yorkers.6 The website is a big human data research platform that uses big data analytics to better understand the complex and interwoven influences on health and behavior.18 Its scope surparsses any similar study attempted in the past and is the latest of Glimcher’s foundational projects.6

Outside of his various academic achievements, Glimcher also likes to put his mathematical brain to practice in yachting. He first started sailing with his father when he was a young boy and grew this passion over the years. He enjoys yachting with his wife, Barbara, and his daughter, Zoe.19

Insights from Paul Glimcher

Glimcher strongly believes that while predictive models of economic behavior would be useful after we gain a better understanding of the ways economic decision-making intersects with psychology and neuroscience, he is also a strong proponent that there is no ability for models to be certain. He said that “doubt is not a pleasant condition, but certainty is an absurd one. 20 In part, this is because departing from the belief that humans are rational decision-makers, Glimcher believes in subjective value: that each individual has different preferences and values and therefore might behave differently than “predicted”.21

It wasn’t just the field of traditional economics that Glimcher understood to have some gaps. He felt as though neuroscience wasn’t being applied to decision-making, which he viewed as a problem. He said that “neurobiological models of the processes that connect sensation and action almost never propose the explicit representation of decision variables by the nervous system”.4

Glimcher also believed that neuroeconomics could be applied to sports. His love for sailing partially grew out of an interest in science and math. He said that “for someone like me who grew up as a scientist soaking up and learning how to analyze data and how to draw conclusions from data, being a navigator is just natural … it’s a question of how you can affect the data that you gather about the yacht, and how you use that to make decisions – how you programme computers to make decisions and help you make decisions… it’s really neat to see that sailboat racing has really come of age as a statistical science.” 19

Another area that Glimcher has more recently become involved in is the technology side of the medical field. He believes that as the field currently stands, “digital health systems are technology-facing, that is, focused on developers and data types. There is very little about the patient side of things … digital health [is] failing to incorporate educational or communication-oriented tools and those systems being deployed were not sensitive to cost”.22 No matter the field, Glimcher never treats humans as computational agents who act rationally.

Instead of completely dismissing economic models, Glimcher suggests that “Neuroeconomics today makes great use of both models that are meant to provide accounts of what should be chosen (normative models) and models that describe what is actually chosen (descriptive models).” 1

Where Can I Learn More?

We’ve already mentioned Paul Glimcher’s textbook, Neuroeconomics: Decision Making and the Brain, which draws together insights from the biggest names in neuroscience, psychology and behavioral economics. Another great read is his first provocative book, Decisions, Uncertainty and the Brain: The Science of Neuroeconomics, published in 2003. This book provides an alternative to the Cartesian model of the brain and behavior (that the mind and body are separate) and was foundational for the field. His later book, published in 2010, Foundation of Economical Analysis, provides a foundational theory of approaching neuroscience, a problem which he identified in his 2003 book, without a concrete answer.

If these books are a bit theory- or science-heavy for you, you might want to check out some podcasts that feature Glimcher. This NobelConference47 podcast includes Glimcher’s speech given at the 47th Nobel Conference, which explores novel collaborations between neuroscientists and other disciplines. The Deciding Mind Podcast also has an episode in which Glimcher provides insights into what the future of neuroeconomics might look like. If you’d like to listen to some of Glimcher’s more recent talks, you can check out the seminar he delivered at the 83rd symposium of the Cold Spring Harbour Laboratory, all about order and disorder in the nervous system.

References

  1. Amazon. (n.d.). Neuroeconomics: Decision making and the brain. Retrieved December 18, 2020, from https://www.amazon.ca/Neuroeconomics-Decision-Paul-W-Glimcher/dp/0123741769
  2. The Information Philosopher. (n.d.). Paul Glimcher. Retrieved December 18, 2020, from https://www.informationphilosopher.com/solutions/scientists/glimcher/
  3. Chen, J. (2019, September 19). Neuroeconomics. Investopedia. https://www.investopedia.com/terms/n/neuroeconomics.asp
  4. Platt, M. L., & Glimcher, P. W. (1999). Neural correlates of decision variables in parietal cortex. Nature, 400(6741), 233-238. https://doi.org/10.1038/22268
  5. Amazon. (n.d.). Decisions, Uncertainty, and the Brain: The Science of Neuroeconomics. Retrieved December 18, 2020, from https://www.amazon.com/gp/product/B08BSYZ1T6/ref=dbs_a_def_rwt_hsch_vapi_tkin_p1_i0
  6. Barham, J. A. (2018, October 15). 25 top behavioral economists. The Best Schools. https://thebestschools.org/features/top-behavioral-economists/
  7. Bhatt, S. (n.d.). 5 Things You Need to Know about Reinforcement Learning. KDnuggets. https://www.kdnuggets.com/2018/03/5-things-reinforcement-learning.html
  8. Osiński, B., & Budek, K. (2018, July 25). What is reinforcement learning? The complete guide. Deep Sense. https://deepsense.ai/what-is-reinforcement-learning-the-complete-guide/
  9. Bayer, H. M., & Glimcher, P. W. (2005). Midbrain dopamine neurons encode a quantitative reward prediction error signal. Neuron, 47(1), 129-141. https://doi.org/10.1016/j.neuron.2005.05.020
  10. WebMD. (2019, June 19). What is dopamine? Retrieved December 18, 2020, from https://www.webmd.com/mental-health/what-is-dopamine#1
  11. Celebrity Age Wiki. (2020, November 11). Paul Glimcher. Retrieved December 18, 2020, from https://www.celebsagewiki.com/paul-glimcher
  12. Robert Rauschenberg Foundation. (2017, July 28). Arne Glimcher. https://www.rauschenbergfoundation.org/artist/oral-history/arne-glimcher
  13. Wayback Machine. (n.d.). Paul W. Glimcher, Ph.D. Retrieved December 18, 2020, from https://web.archive.org/web/20161021004107/kavlihumanproject.org/governance/paul-w-glimcher/
  14. Spader, C. (2020, January 22). Clinical Neurophysiologist: Your Expert in Nervous System Disorders. Healthgrades. https://www.healthgrades.com/right-care/brain-and-nerves/clinical-neurophysiologist-your-expert-in-nervous-system-disorders
  15. American Association for the Advancement of Science. (n.d.). AAAS honorary fellows. Retrieved December 18, 2020, from https://www.aaas.org/fellows
  16. New York University. (n.d.). About ISDM. Institute for the Study of Decision Making (ISDM). Retrieved December 18, 2020, from https://isdm.nyu.edu/about-isdm/
  17. New York University. (n.d.). Paul Glimcher. New York University Arts & Science. Retrieved December 18, 2020, from https://as.nyu.edu/faculty/paul-glimcher.html
  18. The Kavli Foundation. (n.d.). The HUMAN project. Retrieved December 18, 2020, from https://www.kavlifoundation.org/kavli-human-project
  19. Glimcher, P. (2014, May 11). Interview With Paul Glimcher – Owner of Swan 53 Seastar. Interview by Nautor’s Swan. LuvMyBoat. https://www.luvmyboat.com/news/interview-with-paul-glimcher-owner-of-swan-53-seastar/12678/
  20. Montague, P. R. (2003, July 24). Uncertainty rules. Nature. https://www.nature.com/articles/424371a
  21. Encyclopedia. (n.d). Value, subjective. Retrieved December 18, 2020, from https://www.encyclopedia.com/social-sciences/applied-and-social-sciences-magazines/value-subjective
  22. Glimcher, P. (2018, June 4). ResQ is Using Games to Fight Opioid Addiction: Interview with Dr. Paul Glimcher. Interview by M. Batista. MedGadget. https://www.medgadget.com/2018/06/resq-is-using-games-to-fight-opioid-addiction-interview-with-dr-paul-glimcher.html

Read Next