Why We Do the Things We Know Are Bad for Us

There's a particular kind of frustration that comes from knowing exactly what you should do and doing the opposite anyway. I know this frustration well. I've spent entire evenings scrolling through my phone instead of doing the one task that would have made the next morning meaningfully better, fully aware the whole time of the trade I was making, and making it anyway. I used to think this meant something was wrong with my discipline. I've come to think it means something more interesting is going on, and understanding it a little better has made me both more effective and considerably less hard on myself.

The simplest explanation — the one I believed for years — is that willpower is a kind of muscle, and mine was just weaker than it needed to be. Work harder, want it more, and the behavior will follow. I no longer think this is a useful model, mostly because it doesn't survive contact with reality. I have watched myself exercise enormous discipline in one domain — showing up early, pushing through difficult technical problems for hours without complaint — while being completely unable to muster the same discipline for something as small as replying to an email I was dreading. If willpower were a single depletable resource, that inconsistency wouldn't make sense. What made more sense, once I started paying attention, was that the behavior wasn't really about willpower at all. It was about what each task actually represented to me, underneath the surface.

The email I was avoiding wasn't hard in any technical sense. It was hard because it required admitting something I didn't want to admit, to someone whose opinion of me mattered. The difficulty wasn't in the task. It was in what completing the task would force me to confront. Once I started looking at my own procrastination this way — not as laziness, but as avoidance of a specific emotional cost hiding inside an apparently simple task — a lot of my own behavior became legible to me for the first time.

I think this is true more broadly than just procrastination. A huge amount of what looks like irrational or self-sabotaging behavior, in myself and in people I've worked with, makes much more sense once you ask not "why won't they just do the sensible thing" but "what would doing the sensible thing cost them, that has nothing to do with the task itself." The person who keeps missing deadlines might not have a time management problem. They might have a fear, entirely reasonable given their history, that finishing means being judged, and an unfinished thing can't be judged the same way a finished, imperfect thing can.

None of this is an excuse — I still had to send the email, still had to do the task I was avoiding. But understanding the actual mechanism changed how I dealt with it. Instead of trying to white-knuckle my way through with more willpower, which rarely worked, I got better at asking myself what I was actually afraid of underneath the avoidance, and addressing that directly. Sometimes the answer was mundane — I was tired and hadn't admitted it. Sometimes it was more uncomfortable — I was afraid of a specific person's reaction, and no amount of discipline was going to override that fear; I had to actually sit with it.

I want to be careful here, because I'm not a psychologist and I don't want to overstate what is, honestly, just one person's pattern-matching from their own life and the people around them. But the shift from "I have a discipline problem" to "I have a specific, nameable fear I haven't looked at directly" has been one of the more useful reframes I've come across, both for myself and for people I've mentored who were beating themselves up over things that, once we actually talked them through, turned out to be about something else entirely.

The behaviors we know are bad for us and do anyway are rarely stupid, even though they feel that way from the outside. They're usually solving some other problem — avoiding a feeling, protecting an identity, delaying a judgment — more efficiently than the sensible behavior would. If you want to actually change the behavior, in my limited experience, it helps far more to go find the problem it's quietly solving than to just try harder to override it by force.