The complexity of mowing the lawn
20/12/2011 § 8 Comments
We sometimes underestimate the complexity of decisions that include intertemporal components and stochastic elements. That was my thought as I forced my push mower through lank, moist grass this weekend. I couldn’t be bothered to mow the lawn two weeks ago when the weather was fine, and I was paying the price for my laziness.
It shouldn’t be hard to optimise my intertemporal utility with respect to mowing. I want to minimise the effort to keep the lawn respectable-ish and usable. Effort is a combination of number of mowing occasions and difficulty of mowing. It’s just min(effort) = min(f(occasions, difficulty)).
The problem is the stochastic nature of both occasions and difficulty.
- Occasions is a function of my schedule and my family’s schedule. I can’t always be at home when the right mowing occasion arises. It is also weather-dependent: I can’t mow in the rain or shortly afterward. That means I’m trying to minimise a more complex function: min(f(occasions|schedules, weather; difficulty).
- Difficulty is also variable. Partly, it is a function of the recent weather and my mowing technology. I bought a push mower several years ago because my lawn, my garage, and my children were small. With a power mower, the difficulty function would be different. I could more easily mow wet grass, and I could let the grass get longer. Difficulty is also a cumulative function. It isn’t just rain yesterday or today that increases the difficulty. It is the cumulative moisture and warmth that encourages growth. The function is closer to min(f(occasions|schedules, weather; difficulty|weather, technology)).
Of course, the same weather that shrinks the number of available occasions increases the difficulty, so there’s a non-linear, reinforcing effect. Add to that the role of expectations. I have poorly anchored expectations for Wellington weather, having lived here only a year. My priors from other climates are reasonably poor for predicting rain patterns and grass growth here. I end up with Shacklian surprise.
The good news is that I didn’t feel like I’d made the wrong decision two weeks ago. That is, even with perfect foresight I still would have chosen to be lazy. However, my total utility was lower than expected.
If I were really risk-averse, I could participate in the lawn mowing insurance market and reduce the variability of my intertemporal utility. I could just hire a gardener.