So at the end of my last post, I said that although 3 years ago I was gung-ho about arranging marriages, I’m now more hesitant. This post goes into why.
For starters, and to get a bit personal, my parents had one of the worst arranged marriages in history. They separated and divorced when I was very young. A significant fraction of my childhood memories are filled with their marital strife. Trust me when I say it was some ugly stuff.
Yes, anecdotes aren’t data. And the data we have suggest that arranged marriages “succeed” (i.e. divorce less) more than the usual fall-in-love-first marriages.
But then again…there’s no real data on what fraction of traditional arranged marriages are happy and thriving. Especially if the couple is from a culture that ostracizes divorce, we may never know. Again based on anecdata within my extended family, I think many South Asian arranged marriages are not what most would call “good.”
As an aside, it wasn’t until college I got over my discomfort at telling other Desis that my parents were divorced. I had no problem telling white people because, well, all white people get divorced. It’s really not a big deal among the whites. Now, I’m at peace with my it. I even wear a ‘My parents are divorced’ shirt when I go to Indian functions. But I digress.
Anyway…even in my parents generation, people weren’t necessarily “told” who to marry. I believe it was presented as: “Here’s someone we think is a good match for you.” My parent met several times before getting engaged.
There was of course implicit pressure to agree to the proposal. But at least in theory they could have said no. And as I described in my last post, among South Asian Americans there’s definitely a period of dating. It’s common to not go with your first proposed match.
So even for people who grew up around arranged marriages, it’s a big ask to present someone who they haven’t met before and say: “This is your future spouse.” A really big ask. I think it would be even harder for people who are less familiar with it.
And even if enough Americans agree to an arranged marriage to make the endeavor profitable, I wonder if Justin’s strict requirements will lead to happy marriages. Justin laments (as I do) our lost capacity for faith. But I think he’s guilty of too much faith in believing that an algorithm alone, however sophisticated, can predict marital bliss
One of my core beliefs is that people say ‘no’ too easily to potential relationships, and especially when they meet online. They get distracted by potential options and overoptimize. We’re all more compatible with more people than we think.
But it’s also possible to say ‘yes’ too easily. Solid relationships take time and learning from both parties. I’ve been reading Scott Stanley’s blog a lot the past couple days, and his ‘take it slow’ advice sounds like the right move to me.
So how do we square this circle? How do we foster marriage without dictating a spouse while also while not creating too many options? Here’s the You’re Good Enough (YGE) solution:
- Sign people up to to my relationship service. When they sign up, they promise to deactivate all other dating accounts.
- Use data to determine which couples are more or less likely to have a long-term relationship
- This part is quite a bit hand-wavy right now. But I’m instinctively opposed to the machine-learning big data approach used by most sites. I’d lean more towards a fast and frugal tree. This Stanley post delves into some of the research.
- Match two people. You don’t get to see pictures of other people on the site. You don’t need to wink, poke, wave, smile, punch, or message anyone. You get a match.
- That match will be active for 3 months, and both of you agree to go on at least 6 dates over that time.
- If one insists on ending before the 3 months are up, he doesn’t get another match. He has to wait. So there’s an incentive to go on more dates and get to know your match.
- My site would “guide” the dates. Or rather, provide questions to ask / things to look out for to help people decide if they’re with someone who they could marry. Finding a spouse would be the stated end goal.
- After 3 months, the couple can decide to enter into a more committed relationship that (hopefully) leads to marriage, or decide it won’t work out. But they have to stick with it for at least 3 months.
I think some variation of the above would, on average, lead to more successful marriages. The couple has to be given some time to explore and learn for themselves. An algorithm can’t do that.
