Thread by @lymanstoneky

The only #NBERday paper I'll cover today:

Q: What is college for?

A: Getting married.
#NBERday
nber.org/papers/w28688

This is a cool paper. They use Norwegian data to track people before, during, and after college attendance. And they exploit a quirk of Norway's admission system: there's a lot of hard cutoffs and quasi-random variation in admission. #NBERday

There's even hard cutoffs and quasi-random variation in what *field of study* a person can enroll in. Everybody applies to a centralized system and is allocated out to schools. You apply to a field and a school simultaneously. #NBERday

This means we get EXOGENOUS VARIATION WHOOOO BREAK OUT THE PARTY HATS IT'S A NATURAL EXPERIMENT #NBERday

Being just above/below the cutoff for getting into one's preferred program doesn't impact odds of marrying, but it radically alters odds of marrying someone *in your preferred program*. #NBERday

Note that 100% of the effect of "economists marrying economists" in this data is ACTUALLY driven by "UK-Econ grad marries UK-Econ grad," once you control for specific program, field doesn't predict homogamy! #NBERday

So people are not matching on *similar interests generally*. And their formal model controlling for some background characteristics validates this more extensively.

People match on *sharing a program*. Same-institution-different-field boosts match odds a lot too! #NBERday

It's not *interests* that lead to marriage it's *incidence*! #NBERday

This effect is bigger for women than for men once you control for women's over-representation in programs. i.e. Norwegian women more aggressively use their immediate collegiate circle for mate-finding than Norwegian men do. #NBERday

This truly is an *institutional effect*. The effect of institution on institution-level matches is WAY bigger than the effect of field on field-matches, or even program on program-matches. #NBERday

And what group of people marry each other most!

Lawyers.

lol.
#NBERday

Mere physical proximity, i.e. at the same institution at the same time, is THE BIG DRIVER of marriage formation. #NBERday

Field and program stuff matters, it's not *nothing*, but field/program basically matters *within the institution.* And institution matters independently of field. And this is controlling for a bunch of background characteristics. #NBERday

Folks.

Going to college may alter your odds of marriage or what job you get.

*Which* college you go to does not alter those things much.

But it has a HUGE effect on WHO you marry. #NBERday

The correct way to choose a college is when you visit it to request that the college host a speed-dating program where a bunch of current freshman/sophomores and admitted students participate together. #NBERday

I'm joking but only 75%. #NBERday

A good proxy for this btw is, when visiting a college, skip the admissions stuff and go to every student extracurricular you can. That's what I did and wildly enough I ended up marrying one of the three people I remember meeting from one such visit! #NBERday

Also she minored in economics so that's some institution-field-program homogamy right there! #NBERday

Other fun stuff in here too. They find that the sex ratio matters: being the minority sex boosts odds of a same-institution marriage. But market size does not matter: going to a bigger school does NOT increase odds of homogamy! #NBERday

That's some strong evidence for human-scale frictions being a key part of the matching process! #NBERday

This paper argues, I think convincingly, that marriage markets are intensely local, mostly driven by physical proximity, and way less highly dependent on shared characteristics besides proximity than you might think. #NBERday

This also highlights why it's so hard to form relationships in today's atomized environment! College is an intensely social-capital-building period; you see a person in numerous different contexts, not just coupling. #NBERday

As physically proxemic socialization in groups for purposes not-explicitly-about-coupling has declined, we've seen people have a harder time form stable relationships lasting until marriage. Because *it's the proximity stupid*. #NBERday

Good question. They don't address it. But since the advent of Tinder has NOT led to earlier matching, we should conclude that Tinder creates the *illusion* of interest-based matching, but in fact proximity is still king. #NBERday

https://twitter.com/NickJHiggins/status/1384178563977416710

That a large share of unions being formed were initiated online is not a contradiction to this since that is a growing share of unions ***at a time when unions are coming ever-later***. #NBERday

The rise of internet-matching is temporally associated with *lower* overall age-specific matching rates. That's the key intuition I would argue for. #NBERday

By the way, lots of people making Mrs. degree jokes. But that's not what this paper found. Women are more aggressive about pursuing institutional homogamy, but that's merely because of the sex ratio issue mentioned. #NBERday

And while women are marginally more likely, the effect is extremely strong for men too.

I would suggest that this paper shows a lot of men re looking for their Mr. degree. #NBERday

which SIDENOTE, returning to a take from a while ago, we should really have different salutations for married and unmarried men as we do for married and unmarried women! It's crazy we don't do this! #NBERday

I need a term I can use to subtly indicate to my single dude friends that they need to hurry up and join the world of adults and settle down with a stable partner. #NBERday