(Ir)rationality is in the Eye of the Beholder

What does it mean for a behavior or choice to be irrational?

A simple example: a person is given the choice between receiving one or two dollars and chooses to receive one dollar. Was her choice irrational? Common sense dictates that it was. Now a more difficult one: An overweight person has decided that he wants to lose weight and you see him eating a chocolate croissant. Is this behavior on that person’s part irrational?

This second example is much more nuanced than the first one. For starters, it’s not clinical. In the first example you know all there is to know about that person’s situation, that she has two possibilities with one being objectively better than the other and yet not choosing it. In the second example, you only know that that person wants to lose weight but the context behind his eating the croissant is lost on you. I imagine most people would say that that person’s behavior is irrational since it goes against his stated goal. You just don’t eat chocolate croissants if you want to lose weight. So let’s put a tentative answer: Eating a chocolate croissant if you want to lose weight is irrational.

Now lets introduce introduce some context. Let’s call that person Doug. Doug has been on a new diet for three weeks. The nitty gritty details about his diet are irrelevant, what matters is that the diet represents a big change in his eating behaviors and every day Doug must be mindful that he cannot slip back into his previous habits. His goal truly is to lose weight. Yet, on Sundays, he eats chocolate croissants for breakfast. You see, Doug allows himself to cheat on Sundays. It’s all part of his plan to lose weight. The idea behind his decision is that by allowing himself to eat chocolate croissants on Sundays he won’t be craving them throughout the week, or at least he’ll know that come Sunday his cravings will be fulfilled. And you know what? It’s working. He really is sticking to his diet, not only on Sundays but throughout the week. It might be that the croissants are just an excuse to pig out on unhealthy food, but they don’t seem to be hindering his weight loss journey.

Now, let’s ask the question again: Is Doug’s eating of that croissant irrational on his part?

Context is everything, and it doesn’t come readily for those looking outside-in. I was motivated to think on this upon reading the following quote from Meir Statman’s Finance or Normal People: How Investors and Markets Behave:

Consider a person with debt balances on two credit cards. The debt on the first is large and its interest rate is high, whereas the debt on the second is small and its interest rate is low. Now suppose that the person has cash sufficient to fully extinguish the small debt on the low-interest card but not sufficient to fully extinguish the large debt on the high-interest card. Wants for utilitarian benefits direct that person to reduce the debt balance on the high-interest card. Indeed, we might regard a decision to extinguish the debt on the low-interest card as a cognitive or emotional error. Yet it is not, because sacrificing short-term utilitarian benefits by extinguishing the debt of the low-interest card might enhance long-term utilitarian benefits. The expressive and emotional benefits of the small victory that comes with fully extinguishing the small low-interest debt balance might encourage a person to persevere in extinguishing the large high-interest debt.

Again, context is everything. Classifying another person’s behavior as rational or irrational is contingent on us having all relevant information that underlies that behavior, and it’s not obvious that we’re even able to get that information most of the time.

For starters, that same information might be opaque for the person who does the behavior. Grandma says that she goes to church to look upon the face of God but the true reason might be that all her friends are there on Sundays and she enjoys their presence and the routine. Less charitable people might insist that her going to church is irrational on her part because God doesn’t exist so she’s only losing her time, but spending time with her friends is reason enough to go to church.

Curse of knowledge problems also arise when assessing context. Consider two persons going on a diet but one believes a low-fat diet is the best tool for the job while the other believes a low-carb one is where its at. Which one of these is the irrational belief? Even if one is indeed correct while the other is not, I’m not sure acting on one’s beliefs, even if they turn out to be factually wrong, is irrational, but a zealous low-carb dieter might believe the low-fat dieter to be irrational in his beliefs and vice-versa.

This does not mean that there is no such thing as irrational behavior, as the first example of choosing one dollar versus two dollars shows. The point is that in assessing another person’s behavior as irrational or rational, we are prone to committing errors in thinking by assuming that what we know is all there is to know. Doug wasn’t being irrational, quite the contrary, but upon seeing Doug we couldn’t know the context behind that decision so we default on the explanation that that decision is not a rational decision on his part. It’s possible that we will be correct in our assessment most of the time since it’s unlikely that every Doug out there is eating chocolate croissants because it’s their cheat day (or maybe cheat day is every day), but this just means that we must think probabilistically, that while we may be correct in our assessment, it’s entirely possible that we’re wrong.