Update: I try to synthesize the comments from this post here.
Warning: this post is long, but hopefully thoughtful and at least somewhat original. Also, if you’re new to the blog, subscribe in the upper right hand corner via RSS. Check out my recent posts on epistemic closure and fiscal sustainability.
At an off-campus discussion toward the end of my senior year of college, the topic of behavioral economics came up. Leading the discussion was a professor of mine, David Ruccio- whose blog I link to regularly- who argued that to really move forward with these iconoclast ideas, we still have to get rid of the max u thing- it’s holding everything back. I didn’t really agree with him at the time, or I just didn’t know, but a recent panel I attended helped clarify why Ruccio, and other heterodox economists before him, are right, even if the panelists themselves don’t want to see it it or admit.
The topic of the panel, hosted by the Brookings Institution, was “Happiness in a Time of Uncertainty.” The panel was striking in it’s premise- 4 mainstream economists discussing the topic of subjective well-being. A number of interesting points were raised. For instance, it was pointed out that people seem to be very adaptive with regards to happiness, such that they can settle into what one might consider a “bad” equilibrium but regain subjective happiness, and thus show little incentive to leave.
CAROL GRAHAM: But there is one major complication or a fly in the ointment so to speak, and that’s adaptation. People seem to be able to adapt to high levels of adversity, poor health and all kinds of things and retain their natural cheerfulness or their natural happiness…People really can adapt to adversity. They also adapt to prosperity. As I thought about this around the world and how it aggregated up across societies I thought it’s probably really good from an individual psychological perspective that people are able to adapt to adversity and maintain their natural cheerfulness, but it may also result in collective tolerance for bad equilibrium.
The comments from panelists also touched on what behavioural economics has been saying for a while- people do a lot of things that don’t seem to be maximizing anything; there are myriad inconsistencies across time and across situations.
During Q&A, I raised the issue that these points seem to undermine the whole utility maximization assumption, and that we may need to completely overhaul the so-called microfoundations of neoclassical models (EJ Dionne, the moderator, called it an “attack on old-fashioned economics). Here’s the exchange, copied from the event transcript (pdf):
Q: A lot of the comments that all of you have made point to the idea that people don’t really seem to be maximizing happiness or utility at least not in measurably consistent way. Thinking about theoretical economics, how do we continue to go forward with neoclassical models that are based on a maximizing utility assumption?
ALAN KRUEGER: Behavioral economics is an important and growing branch of economics. It’s been recognized. Daniel Kahneman won a Nobel Prize for his work on the limits of people’s decision making. I think the real challenge for economics is to know when to apply lessons we’ve learned from the psychologists and when to return to revealed preference. I’d also caution on how we define utility because I don’t think happiness is utility. Maybe it’s an element in the utility function along with some of the other emotions that I mentioned before like spending time in a good mood or not being angry all day long and so on. I think we need to be careful not to equate our measures of subjective well-being with utility, and at the same time I think we have to be very careful to define utility in a way that’s not tautological because often the way some of our colleagues in economics define utility, it’s impossible to reject and what I think the work of Kahneman and others has shown is that people will make inconsistent choices depending upon the way that they’re framed. You can change the parameters and they’ll make choices which were not consistent with choices that were their interests before, so I think it’s possible to demonstrate that people don’t always behave in correspondence to the axioms that are necessary for utility maximization. But at the same time I think our measurement of subjective well-being is at a relatively early stage and I would be very cautious about going too far in the direction of equating well-being with utility.
KAREN DYNAN: I’ll just jump in and say I think Alan did a nice job of describing both the potential good that’s going to come out the behavioral research and the complications. I’m not a person who believes we need to throw out everything that we’ve been working with. I think if we could have incorporated cognitive limitations and information costs and biases into our standard models we could have gone a long way toward avoiding what we’ve been through in the last couple of years.
EDUARDO LORA: My only reaction to that question about the value of standard economics given all these developments I think is very much along the lines of what Alan said. I don’t think that we should throw away what we had. What these new developments lead us to is to understand phenomena that were important for people’s well-being that could not be related with standard economics and that’s precisely why I see that this area of public goods is so important and so amenable to this approach because certainly the way that we economics were trying to approach many of these problems of public goods didn’t go anywhere, they were just approaches that were too convoluted, while this view provides an approach that is very simple and really takes you very far. So I think that the toolkit of economics is a very powerful one and I don’t think that it’s to be discarded with these developments. I think it’s going to be complemented with these developments and I think that there is a strong consensus in the profession about that.CAROL GRAHAM: Quickly let me echo what everybody said that you can’t throw out the baby with the bathwater on this one particularly because even though we found a great way to think about things differently and to identify quirks in the way people make choices and all kinds of irrationality, we still I don’t think have a lot of questions answered for example about the definition of happiness, how does happiness relate to subjective well-being and all the domains that Alan mentioned, which ones matter more, what definition of happiness do we care about in the policy domain. So it’s a new tool, it’s great, but it’s not ready to replace everything and I think it will always be an important complement and may make us rethink some of our traditional tools.
The point of my question was not that happiness should replace utility. It was that, if as these panelists had noted, adaptiveness in terms of happiness is natural, it’s likely that adaptiveness occurs in other elements of the utility function. Utility maximizing behavior in one time period may give way to a more indifferent, “just get by” strategy, in another. My Game Theory prof once defined rational as “strategic”- however, people aren’t necessarily strategic. Instead of maxing, they may just satisfy- get just enough. How do we model this? To stick with reasonable premises, I think we have to throw away that whole thing. Looking at how individuals actually behave, with which both the happiness economics and behavorial economics literature helps us, it becomes hard to argue that individuals are maximizing anything. If we do cling to this argument, it’s more likely that we are merely defining utility in a way that Krueger appropriately terms “tautological” (and warns against).
The panelists’ responses are striking in another way too- they see hope for behavioral economics as a way to tweak the inputs into utility maximization function. It’s not that surprising, though, that mainstream economists can be a defensive bunch- they’ve spent many years learning the maths and plying the trade, and each in this group has achieved significant status by using these sorts of tools (although Krueger was nearly rendered apostate by challenging the notion that a minimum wage will reduce employment, and Graham spent years trying to make the concept of studying happiness credible). Each of their responses referred to these sets of tools, and if we abandon the assumption of utility maximization, it makes the standard models useless, as it introduces untold degrees of heterogenity.
This sort of denial is frustrating to me, because behavioral economics is becoming more accepted in the mainstream and it makes clear that people do not behave in rational or predictable ways. However, making that next leap is difficult because there is an inconvenient consequence of throwing away max u (not just the loss of the “toolkit”). The degree of complexity that removal would introduce would chip away at economics’ drive to become more like physics. A recent paper (h/t Economic Logic) discusses this phenomena, pointing out how entire theories can blossom from any one unrealistic assumption:
And much of the economics and finance literature since Foundations has followed Samuelson’s lead in attempting to deduce implications from certain postulates such as utility maximization, the absence of arbitrage, or the equalization of supply and demand. In fact, one of the most recent mile- stones in economics—rational expectations—is founded on a single postulate, around which a large and still-growing literature has developed.
The authors rightly criticize the excessive mathematization of economics, and they argue that a large degree of uncertainty should be incorporated into models.
By acknowledging that financial challenges cannot always be resolved with more sophisticated mathematics, and incorporating fear and greed into models and risk-management protocols explicitly rather than assuming them away, we believe that the financial models of the future will be considerably more successful, even if less mathematically elegant and tractable.
So, however troubling or intractable it may be, I think economists need to pull themselves away from the max u assumption. It may result in a complete trashing of their neoclassical model. It may even result in a situation where microfoundations are no longer the basis for our macroeconomic thought. However, being wedded to unrealistic assumptions, falling prey to methodological individualism and methodological utilitarianism doesn’t get us anywhere. It preserves the status quo of poor models with poor conclusions that ultimately harm people. There are a number of ways in which economics does this, but let’s start by doing away with the most basic methodological one.
Let me hear it- what am I missing here?