site banner
Jump in the discussion.

No email address required.

Here’s another graph where the benefit of other people caps out at 50% of them going outside. The situation is much better! Your couch can be pretty great before the activity starts seriously dying off. But again, just like before, once it reaches that critical threshold, average happiness gets worse for a bit before it gets better again in the world of only shut-ins.

[Insert Graph that looks barely at all like the first one]

So the moral of the story is, to avoid isolation and depression, move to a big city, pick popular hobbies, and if someone asks you to go caroling this Christmas, go even if you don’t want to, because it will make the experience better for everyone else.

I was with you up until this conclusion. How on Earth do you look at that second graph and draw that moral from it. The existence of a tiny window where happiness is a decreasing function while being increasing in the vast majority of space suggests that the moral for the second set of parameters is "we should prioritize making indoor activities more fun because in almost all cases it increases happiness. Also we shouldn't pressure introverts to participate in social activities because they're clearly enjoying themselves more via their revealed preferences."

We get very different conclusions for the first and second set of parameters if we're being honest instead of using motivated reasoning based on a pre-supposed moral. The actual moral for the overall model should probably be something like "this deadweight loss phenomenon appears under some model parameters but not others, so the applicability of this to the real world depends a lot on the circumstances and we need to measure and investigate further to determine which world we live in, but it's an interesting possibility especially in less crowded spaces such as rural areas."

Yeah I think that's fair. I spelled out the situations I think it's realistic for in the middle a bit, though probably should have made that more front and center instead of presenting it as a totalizing view of society.

For now assume that this is a niche interest or a small community where the number of people doing it still has incremental returns, things like “Are there kids playing in the local park” or “Does your block have a block party” or “Do my friends of friends host house parties that I’m invited to.” This also models things that require a very large amount of effort to organize that few people are willing to do, so you need a large amount of people choosing to invest in your community to get just a few willing to put in the hours, things like “Does your grade school have an active cub scout troop” or “Are there community choirs nearby.” At minimum, the density of people willing to do the thing determines how far you have to drive to do the thing, and how much of a pain it is to do.

Sure. But then the second graph demonstrates the sensitivity of the model to this assumption. This turns out to be a critical assumption that heavily influences the results, at least the part of the results we care about. Therefore any conclusions/morals/takeaways need to emphasize this caveat.