Thursday, March 13, 2014

dubiously rational preferences

I suppose first I should announce, for context, that this August I will be joining the University of Queensland School of Economics as an assistant professor.* I'm very very excited!

That's not the dubiously rational decision I'm referring to, of course :)

During the job market I was in the strange position of having to pick a single option from a set lacking any objective measure of overall quality and in a very high-stakes situation. That means, unlike most (ambiguous) decisions that I make out of habit or simple guiding principle or quick gut instinct or based on a simple measure like "price" or "amazon rating", I actually had to carefully figure out what my preferences were in order to maximize utility. And in doing so, I identified a couple of systematic irrationalities - or at least, behavioral-econonomicy aspects - to my decision making.

First of all, and most clearly, my preferences are definitely over choice sets in addition to outcomes. In at least two separate cases, I prefer school A from the set {A,B}, prefer school C from the set {A,B,C}, but prefer choice set {A,B} to {A,B,C}.

Now, to a non-economist, that might seem pretty straightforwardly irrational, or perhaps perfectly straightforwardly rational if my first instinct is a product of thinking like a classical economist for too long.** The technical definition of rationality is "complete and transitive preferences" (completeness being trivially satisfied in any real-world scenario I can think of). And preferences over contingent choices, i.e. over choices in combination with the choice set from which they were selected, can be transitive just as easily as preferences over outcomes alone. But that's certainly not something classical economics is concerned with.

It's reminiscent of Gul and Pesendorfers model of self-control, actually, but in a slightly twisted way. They say that self-control is demonstrated by manipulating your choice set to remove the tempting option. Seems reasonable. But in some sense, my reasons for preferring smaller choice sets are the exact opposite - I wanted my choice set to exclude options I should take, for some definition of "should", so I could choose what I wanted to choose freely. Now, this definition of "should" in the self-control model is a feeling that the individual unambiguously agrees with - the value of going to the gym, for example. For me, I knew that others would think I should unambiguously go with choice C, and I myself had very conflicting feelings about it, and preferred not to have to deal with the cognitive dissonance at all.

I guess it amounted to some combination of 1) wanting to avoid succumbing to peer pressure, 2) wanting to avoid being forced to update my preferences based on the information contained in everyone else's opinion, essentially changing my mind from thinking McDonalds is guilt-freely great to reluctantly agreeing I should go to the gym instead, or 3) avoiding the mental effort of the whole process. I'm not sure. I suspect 2 is the most correct, because I'm historically very immune (often to the extent of intentionally going the other way, so my mom would probably say) to peer pressure and certainly was already putting plenty of mental effort into the decision. But if that's the case, that's pretty unambiguously/uncomfortably irrational (in the colloquial sense of the word, again.)

Hmm, this might be the basis of a decent axiomatic economic definition of stubbornness...

Anyway, I prefer to ignore these uncomfortable inclinations and tell myself I just have abnormally high utility from weather/nature and abnormally low marginal utility of wealth :)

OK this blog post is getting too long, so I'll write about technically-irrational preferences (non-transitivity) separately...


*Actually they call it a "lecturer" position in Australia, but it's the equivalent of a tenure-track assistant professorship in American vernacular.

**You know, like when economists are surprised by things like dictator game sharing. I swear it has nothing to do with economists being greedy people, either - they just see a dictator game, immediately model it with their handy classical economics tools in order to solve the problem of what they should do, and then forget that they used to be a normal human with with behavioral complications like "fairness norms". A better example might be the economists who choose 0 in the beauty contest game, because that's the obvious "answer".


Anonymous said...

1. Actually, you are probably the only one correctly forecasting that the marginal utility of wealth was going to be pretty low whichever job you went to, and that the utility from good weather and gorgeous countryside will be high.
2. I also found it hard to figure out my own preferences once I had gathered all the information. It's a difficult thing to do for these rarely-made choices, especially for an un-self-aware person like me.
3. For what it's worth, I'm pretty sure C would have have been welfare-diminishing for you. I suspect you would have still chosen as you did, but it would have been a pain.

Anonymous said...

Oops, sorry, that was me up there.

Anonymous said...

When you take into account superannuation (retirement contributions) and top-ups (bonuses for publications, grants, and good teaching) the pay differential is not huge, and can be in your favor if you publish well.

In hindsight I think the most important thing is a department with friendly, stimulating colleagues. People I actively look forward to seeing when I head over in the morning. Also, living somewhere your family is happy to call home. :)

Congratulations on your offer and your decision!

Anonymous said...

Coolest economics school IN THE WORLD ! (okay, so I studied there)

Vera L. te Velde said...

Oh hmm, yeah you're probably right, I might be comparing a couple of numbers that aren't representative of overall opportunities in each country.

And as for the rest, absolutely! That makes ALL the difference in the long run.

Vera L. te Velde said...

awesome :)

Vera L. te Velde said...

1) hi Gautam! hehe thanks, at least on the marginal utility of wealth front, I have research on my side, but on the weather side, I don't :) (although I maintain the research is flawed. or rather, the conclusions. obviously, any time the research clashes with my experience, it must be wrong.)
2) yeah I know what you mean! it's really agonizing.
3) Well there were a couple of school C's, and every time I think about it again I come to a different conclusion about what would've happened. glad it worked out this way!