Romance Econometrics

I had a mentor at BYU, Prof. James McDonald, who tried to convince us that

  • Econometrics is Fun.
  • Econometrics is Easy.
  • Econometrics is Your Friend.

One of his classes made a bronze plaque out of it for him. He also tried to convince us that Economics is Romantic because this one guy took a girl to his class on a date and she married him anyway. Because he was one of the economists I’ve tried to model my life after, I’ve always been on the lookout for ways to convince people that econometrics is, in fact, fun, friendly, easy, and romantic.

A while back, Bill Easterly blogged about how marriage search is like development, and in the process talking about how unromantic economists can be:

I recently helped one of my single male graduate students in his search for a spouse.

First, I suggested he conduct a randomized controlled trial of potential mates to identify the one with the best benefit/cost ratio. Unfortunately, all the women randomly selected for the study refused assignment to either the treatment or control groups, using language that does not usually enter academic discourse.

With the “gold standard” methods unavailable, I next recommended an econometric regression approach. He looked for data on a large sample of married women on various inputs (intelligence, beauty, education, family background, did they take a bath every day), as well as on output: marital happiness. Then he ran an econometric regression of output on inputs. Finally, he gathered data on available single women on all the characteristics in the econometric study. He made an out-of-sample prediction of predicted marital happiness. He visited the lucky woman who had the best predicted value in the entire singles sample, explained to her how he calculated her nuptial fitness, and suggested they get married. She called the police.

He goes on from there to describe how he eventually did find a mate and makes a comparison with development and over-reliance on econometric methods. As popular as it is in Libertarian circles to bash on econometrics, I’d like to defend empirics by pointing out that his regression advice was not sound:

1 – The suitor’s regressions ignored the self-selection bias. Regressions only tell us what the ‘average’ effects are, that is the effect for the ‘average’ person. Making the average guy happy is only relevant if he is the average guy. Economists being the strange lot we are, it is likely that it takes a special kind of person to marry one of us. He ought to have found a bunch of guys very similar to himself and examine the qualities that made a difference from among (and this is key) the population of women willing to marry guys like him – the women who self-select themselves into our group. If he then approached a women who was not in that group, no wonder he was rejected! I knew I had my work cut out for me since I was in junior high: a Latter-day Saint economist-in-embryo who read Shakespeare “in the original Klingon”, and who carried a briefcase to school? Small sample sizes indeed!

2 – He ignored endogeneity. Instead of trying to convince her that research showed she would make him happy, he needed to present research that demonstrated he would make her happy, and that’s the other half of the regression: male qualities on marital happiness. No wonder she rejected him: his regressions didn’t answer her question!

Personally, I took more of a Bayesian approach. Bayesians believe that a lot of things in life (like regression coefficients) are random and over time we get better and better signals about where the truth is, but we only ever approach it by degrees. First, by trying to become a friend, I identified if a woman was in the group of people who might marry someone like me. Each interaction gave me more information about the error term and the regression coefficients about fostering a happy, loving friendship that could endure. After any failed relationship, I had a new variable or two to add to my equations and I understood the ‘relationships’ between relationship variables better. That might be about finding out different things I needed (hunh, so her political affiliation isn’t as important as I thought and her willingness to smile at me is vital) or about learning more and better policies over time that I could enact to make her happier (tips for being a better listener or learn to identify her love languages and feed them to her regularly).

One of the most important regression-related romance tips I learned was to control the variables I could control, and leave the residual in God’s hands. I recall a graduate labor economics research seminar where the presenter claimed that the marriage market always cleared. I complained that I was willing to supply a great deal more marriage than had ever been demanded at prevailing prices. I was reassured that the marriage market clears in equilibrium, and I might not have found my equilibrium yet. The presenter’s prediction was, thankfully, prescient: I found a buyer a year later, and last week we celebrated 5250 days of married bliss.

Where is the optimal marriage market?

I have spent the past few weeks playing around with where the optimal marriage market is and thought NoL might want to offer their two cents.

At first my instinct was that a large city like New York or Tokyo would be best. If you have a larger market, your chances of finding a best mate should also increase. This is assuming that transaction costs are minimal though. I have no doubt that larger cities present the possibility of a better match being present in the dating pool.

However it also means that the cost of sorting through the bad ones is harder. There is also the possibility that you have already met your best match, but turned them down in the false belief that someone better was out there. It’s hard to buy a car that we will use for a few years due to the lemon problem. Finding a spouse to spend decades with is infinitely harder.

In comparison a small town information about potential matches is relatively easy to find. If you’re from a small town and have known most people since their school days, you have better information about the type of person they are. What makes someone a fun date is not always the same thing that makes them a golf spouse. You may be constrained in who you have in your market, but you can avoid lemons more easily.

Is the optimal market then a mid sized city like Denver or Kansas City? Large enough to give you a large pool of potential matches, but small enough that you can sort through with minimal costs?

P.S. A friend has pointed out that cities/towns with large student populations or military bases are double edged swords for those looking to marry. On the one hand they supply large numbers of dating age youths. On the other hand, you would not want to marry a 19 year old who is still figuring out what they want to major in.