In the comment threads for the “Hey! I didn’t say Science Says!” post, there were a couple attempts to come up with metrics to check the impact of changes to marriage law. Some of them are definitely quantifiable (suicide rate, reported domestic abuse) but don’t do a good job capturing the impact of a given marriage schema on the whole population. Those kinds of outcomes are only tracking couples who hit a certain threshhold of dysfunction. It’s like measuring the effectiveness of a vaccine by just tracking the incidence of Guillain–Barré syndrome. We care about the data, and we might tweak our intervention as a result, but it doesn’t give us the whole picture. The purpose of marriage isn’t simply ‘lower the suicide rate’ after all.
There are ways to gather data for less extreme successes or failures. One way to check data about subjective feelings is to give subject a beeper and have them answer some questions each time the beeper goes off (“How happy do you feel right now? How tired? How calm” etc). The trouble is that some of the studies of this type have turned up some counter-intuitive results (childless couples are happier than parents, etc), and we have to decide what we trust more: our expectations of the right answer or this methodology’s ability to reflect reality.
You can pull out a Kahneman-type solution and draw a distinction between your experiencing self and your remembering self, and choose which one to privilege. But another solution is just not to care much about happiness as an outcome variable. If contentedness was our highest goal, we’d never find ourselves saying “I do not understand what I do. For what I want to do I do not do, but what I hate I do.” We’d just try to want what our actions told us we wanted instead.
That’s the difference between happiness and rightly-ordered happiness, but we can’t tell the difference between these two by asking ourselves how happy a choice makes us. And that’s when we turn to teleology and ontology. If we can’t tell higher pleasures from lower ones subjectively (or, at least, there’s an inconvenient time lag for epiphanies), then we try to reverse engineer ourselves. Figure out what we are and we might be able to suss out what we need.
But that means that a fair amount of the time our needs are going to not track our wants or our happiness. Parents might well be unhappier than non-parent couples, but happiness isn’t the result we’re most interested in. And when enough of the world and culture have been optimized for happiness or contentedness, going along the other route will be more jarring.
More further reading: Timothy Burke has a good essay on the difficulty of balancing even obviously valuable utils.