Forced to Defend Rational Choice Theory

I’m finally responding to Eli Thorkelson, who asked for my comments on an article by feminist economist Julie Nelson.  The article is a critique of rational choice theory (RCT) and I think it has some omissions and misleading claims.

I regularly come across particular instances of rational choice theorizing that I dislike, but non-economists, including sociologists, often dismiss rational choice theory without understanding it, so when the topic comes up among non-economists, I almost inevitably find myself defending it.  My claim is that rational choice theory, broadly construed, is an important, though certainly not the only, useful framework for understanding human behavior.  This should be considered an utterly boring claim.  What is interesting is how any social scientist could deny it…

The primary problem with Nelson’s article is that she assumes an essential piece of what she should be trying to prove, namely, that RCT is inaccurate.  Yes, RCT posits an inaccurate, unrealistic psychology, and all else equal, this counts against the theory.  The reason it isn’t a fatal blow, is that the theory is relatively more accurate when applied to many social phenomena.  To argue that RCT rarely led to accurate predictions is to dismiss essentially all of microeconomics and hundreds of studies in sociology and political science as well as other fields.  e.g. What happens to car theft when a new anti-theft device is introduced? What happens to group solidarity as group size increases? To me, these examples prove that it is possible to do accurate work within RCT.

Though I’m sure there are exceptions, everyone I know who respects RCT believes it has some weaknesses.  Some mainstream economists even argue that an overdependence on the assumption of rationality has led large segments of modern macroeconomics astray.  This is a debate worth having.  As Nelson notes, even in economics, the discipline most closely tied to RCT, many mainstream scholars are relaxing the rationality postulate.  One might think Nelson would enthusiastically support such work, but she expresses skepticism that their revisions will ever be radical enough.

In a section entitled “How RCT has Done Economics Wrong,” Nelson claims, on the basis of cherry-picked quotes, that social problems such as “…the problem of child poverty—the physical suffering and stunted development of those in our society who are most manifestly not autonomous—is erased by the application of “rigorous” RCT economics.”  But this ignores the fact that hundreds of scholars working within the rational choice framework, are very concerned about childhood poverty, and hope that their work will help us learn how to ameliorate it.  Nelson, who was trained as an economist at the highly-regarded University of Wisconsin at Madison, must know this, and it is misleading of her not to mention it.

Nelson could have written a subtler critique of RCT, after all, I agree that alternative approaches are too often overlooked in economics.  But even if we limited our view of RCT to economics, we will find a great deal of diversity in how it is used.  I highly recommend those unfamiliar with this diversity see Tyler Cowen’s article, How Do Economists Think About Rationality?


13 Responses to Forced to Defend Rational Choice Theory

  1. wet says:

    I think a succint way of summarizing your article, as relevant to economics, is don’t throw out the baby with the bathwater. Sure there are some problems with rational choice theory, but the problems cited by academics are often very pointed. We shouldn’t advocate a ‘Popperian’ (if that is a word) mentality and assume the entire framework is defunct because of a few criticisms.

    The biggest problem, as I see it, that economics has with this framework is problems with matching empirical data with the intertemporal optimization solutions suggested by rational expectations. I notice that rational expectations and rational choice theory are not exactly the same thing, but I’m wired on caffeine. For example, there is ample evidence that Euler consumption equations do not perfectly account for consumer behavior, at least for those in their early twenties. But, that does not mean that the entire framework is ruined; it just means it must be readjusted to account for this buffer-stock behavior (see chris carroll, JHU).

  2. eli says:

    Hi Mike, interesting post. As I think about it, I don’t know that I have a huge stake in this debate (to the extent that there is an active debate and not just a set of dormant interdisciplinary prejudices). I glanced over the paper on rationality and the one on car theft. Can you give me a quick explanation of why the latter paper is a good example of rational choice-based research in action? On my admittedly hasty reading, it seems to be primarily a statistical analysis of how car theft rates shift in the presence of lojack, it says very little about the ‘rationality’ of any party’s action (though the insurance companies obviously profit substantially from the deploying of the lojack system), and it employs, in fact, a notion of “social benefits” that seems somewhat at odds with a completely individualist theory of costs and benefits. Obviously they assume that car buyers would be more likely to buy lojacks if the price was better subsidized, since individual benefits from lojack are currently somewhat low even though positive ‘externalities’ are high, so I guess on this level they presuppose that car buyers are rational consumers — is that where it comes in? This isn’t a critique, I’m just trying to get my bearings in this kind of research.

    Anyway, I did tend to agree with the comment — I think in Cowen’s paper — that RCT research per se (and the rationality postulate in particular) is lumped together with larger ideological debates about the role of loosely rational choice-inspired policy and governance. It strikes me that I’d like to see a more thorough exploration of the historically evolving relationship between RCT, market-oriented social science, and market-based ideology or government policy — are these social phenomena tightly coupled together? Loosely coupled? What about the rumors that economics students tend to change their behavior to conform to their theory of rational action? A lot of “looping effects” (ie, people changing to conform to the theory) seem to exist in this domain, and would merit further exploration…

  3. Michael Bishop says:

    I should have said a little more about the lojack paper, excuse me while I spell it out in even more detail than you need. Many people think that a lot of research in economics is shaky because it depends on an assumption of rational individual behavior. The lojack paper is motivated by this assumption in a manner that is very common:

    1. Lojack makes it harder to get away with stealing cars, therefore potential criminals rationally choose to steal fewer cars.
    2. People buy lojack rationally, based on the private benefits. But non-buyers also benefit from more deterrence with each lojack sold. In other words, there is a positive externality, the social benefits exceed private (buyer and seller) benefits.

    Then, very importantly, there is careful statistical work to estimate precisely what effects lojack had (and yes, they seem to be consistent with the theory).

    The point is, this is a fairly typical paper in economics, which teaches us something important which could inform public policy. What does RCT look like in practice? Well, its not monolithic, but often it looks like this paper, which seems really hard to complain about.

  4. Michael Bishop says:

    I think there is something to the idea that learning RCT causes one to act in a more self-interested rational way, but I think that the research on this has been overblown. 1) Econ students were probably more that way before they took the class. 2) More importantly, the experiments I’ve seen were quite artificial, and don’t necessarily represent how those people would act in the real world. 3) In real jobs people become more economically rational over time, this is partly due to learning, and partly due to the more successful people staying on while the less successful ones move to another job.

  5. Michael Bishop says:

    @WET, yes, I think “Don’t throw the baby out with the bathwater” is a fair summary. I appreciate your comment even if I detect the subtlest trace of irony.

  6. jimi adams says:

    Personally, i’ve never been huge on RCT, however, the “unreality” critique has never been one that’s bothered me much, despite hearing it a lot. But, i’ve two other long lingering lingering questions about RCT. In fairness, my work actually rarely thinks about individuals qua individuals, so RCT rarely makes its way into my active thinking. As such, these questions may be lingering for me, but have already been sufficiently answered.

    The first concerns the importance of the analytic unit. If macro (which is more of what i’ve encountered, see above), i’ve occasionally puzzled on whether it is possible for a population where micro-level decisions are determined entirely irrationally to generate aggregate level “rationality.”

    The second is more simple: “whose rationality?” While the question here is seemingly simple, unpacking it leads to a world of other critiques of RCT, some of which i am aware of, others likely not so much.

  7. Josh says:

    I never got the critiques of RCT. For the Austrians, it’s a simple tautology which we take for granted, so I don’t see the big fuss. How can action not be rational?

    Michael, could you give me a quick rundown on why people don’t buy it?

  8. Michael Bishop says:

    Jimi, I’m not sure this has much to do with whether RCT is a useful theoretical framework, but I’ll give you a toy example of irrational individuals creating a rational i.e. efficient, outcome. Imagine people come in only two types, the kind that overestimates the future price of iron and those who underestimate it. Everyone is irrationally clinging to their beliefs and making stock purchases and career decisions on this basis. It is definitely bad that every individual was wrong, but the silver lining is that people are wrong in opposite directions. On average people may have been right. This means that in aggregate, the correct investments were made, even if every individual judges themselves to have over or under invested. There seem to be a lot of areas where people’s errors do cancel out, guessing the weight of a horse, or the performance of a stock. This is related to a weak version of the efficient markets hypothesis – the only version anyone really believes.

  9. Michael Bishop says:

    Josh, I suppose the rationality of action depends on our definition of rational. Do you want to offer a definition and then argue that all action is rational? I’d rather just say, “I have this model, I think it captures enough important things about the world to justify thinking about it, but it leaves some important things out as well.”

  10. Brian Pitt says:

    I am sure that this is a “dead thread” by now, but I agree with Michael Bishop here Josh (as I am sure that you do). “Flexible” rational choice simply asks that we identify the institutional context and assume “purposiveness.”
    Thus, I am quite comfortable with results that reveal that those who, e.g., are subject to comply with the nebulous statutes of “work-first” programs rarely obtain jobs.

  11. […] on rational choice theory.  It is framed as a critique, but I consider it wholly compatible with my defense of RCT. Human behavior is complex and different aspects of it will be best understood with different […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: