(Reading time: About four minutes)

What is behavioral economics, AGAIN?

In a February 20th Forbes article, Peter Ubel posted an email exchange he had with Richard Thaler, one of the founders of the field of behavioral economics. In it, Thaler remarks:

I certainly agree that the definition of behavioral economics has been abused.

Thaler's gripe comes down to a few things:

  1. A nudge is not always an example of behavioral economics (and vice versa)
  2. People often mistake examples of social psychology for behavioral economics
  3. People without formal training (a PhD in economics) are labeling themselves as behavioral economists

I agree with Thaler that all three of his points are common, and that they're all problems. But I wonder if the "abused" issue goes a bit further? The definition of behavioral economics may have been abused partly because it doesn't seem like it's ever been that well-defined. And if a term isn't well-defined, won't people be more likely to (unintentionally) abuse it?

To be clear, there are many definitions of behavioral economics out there. But my criticism stems from a lack of specificity within most of these definitions - most only go so far as to define the field as either an explanation of "irrational behavior," or that it's simply a modification of economics with principles of psychology.

Knowing that behavioral economics is an attempt to make economic models more accurate with insights from psychology isn't very descriptive, though (and this doesn't get rid of the blank look on the faces of my friends when I explain what I do...). Can we develop a framework from this? Based off of all of the research to-date, what's the more descriptive definition?


let's take a look at Some definitions OF B.E. over the past fifteen years

2000: Richard Thaler and Sendhil Mullainathan published 'Behavioral Economics,' where they partly defined it as: 

"...three important ways in which humans deviate from the standard economic model. Bounded rationality... Bounded willpower... Bounded self-interest..."

2002: Colin Camerer and George Loewenstein released 'Behavioral Economics: Past, Present, and Future,' where they roughly echoed this partial definition: 

"Other assumptions simply acknowledge human limits on computational power, willpower, and self-interest."

2006: Erik Angner and George Loewenstein wrote in their own 'Behavioral Economics:' 

"...behavioral economics emerged in reaction to the notion that entities like cognitive and affective states cannot be directly observed... [it] emerged in reaction to neoclassical economics..."

2008: Two groups that specialize in the use of behavioral economics and social psychology were founded in 2008 - The Behavioural Insights Team and ideas42. Here's the BIT's take on what they do:

"We coined the term ‘behavioural insights’ in 2010 to help bring together ideas from a range of inter-related academic disciplines (behavioural economics, psychology, and social anthropology). These fields seek to understand how individuals take decisions in practice and how they are likely to respond to options."

2013: The ideas42 folks authored a short paper with CGD on what behavioral economics isn't:

"Behavioral economics is not about controlling behavior... is not liberal (or conservative)... is not about irrationality. ...Behavioral economists study how the context of decisions interacts with our expanding understanding of human psychology."


So where does this leave us?

I made this point in December, and Colin Camerer repeated the definition from his 2002 paper:

I agree with Colin (especially since he has 1,000x the credibility as me). This is really the only "framework" I've stumbled across - and it seems to originate from Dick and Sendhil's earlier-referenced 2000 paper 'Behavioral Economics.' That is, behavioral economics helps us to understand that, under situations of economic decision-making, there are significant limits to:

  • Cognitive function / computation
  • Willpower
  • Self-interest

But under a 2015 lens, this framework holds up in some places, and might be somewhat lacking in others. Limits to "rationality" and "cognitive function / computation" translate to heuristic-based decision-making and the various cognitive biases that come from it (think most of Kahneman's work). So that fits nicely.

"Bounded willpower" covers the concept of the intention-action gap (Ex: I have the "intention" to go to the gym five days a week, but do I actually consistently do those "actions?"), and is a hot topic in many domains (see Ben Castleman & Lindsay Page's fantastic work in education, or Katy Milkman's work in various topics as stellar examples). But I'd argue that this piece of the framework could be better described as general limitations on self-control, not just willpower (willpower assumes an intention, self-control doesn't necessarily). 

Further, "bounded self-interest" stems mainly from findings in behavioral game theory (h/t Matt Darling), and doesn't translate very well to the expansive work that's been done with respect to integrating social psychology into economics. For example, where does something like affective decision-making fall?


So that leaves us with?

  • We often use simple rules of thumb when making decisions (heuristics), which can elicit cognitive biases
  • There are limits to our own self-control
  • We are often significantly influenced by various social contexts

I actually feel pretty comfortable with this as a general, basic framework, but I'd be very curious to hear about other frameworks people use or have seen. Has anyone else come across other definitions? Is a framework even appropriate?


*UPDATE: Erik Angner remarked that his definition of behavioral economics in his and Loewenstein's 2004 paper is actually this:

These days, as it is typically employed, “behavioral economics” refers to the attempt to increase the explanatory and predictive power of economic theory by providing it with more psychologically plausible foundations. By "psychologically plausible" we mean consistent with the best available psychology.

I think this is the definition, but as I mentioned above, what I'm pushing for is a framework to better capture what behavioral economics specifically tells us.