Global Warming: Why optimization on one axis is bad

A new book by researchers from Victoria University is causing some waves. The title? Time to eat the dog: The real guide to sustainable living. Apparently the book is filled with nuggets of wisdom such as:

In a study published in New Scientist, they calculated a medium dog eats 164 kilograms of meat and 95kg of cereals every year. It takes 43.3 square metres of land to produce 1kg of chicken a year. This means it takes 0.84 hectares to feed Fido.

They compared this with the footprint of a Toyota Land Cruiser, driven 10,000km a year, which uses 55.1 gigajoules (the energy used to build and fuel it). One hectare of land can produce 135 gigajoules a year, which means the vehicle’s eco-footprint is 0.41ha – less than half of the dog’s.The author’s admit they are not actually advocating that we all eat our dogs to save the planet. They go on to–well honestly to ramble on about how we should have chickens as pets because we can eat them. Or something.

My point actually wasn’t to discuss this exercise in academic stupidity and thereby give it any credence. However the work does represent a common mistake I feel people make when designing solutions to global warming. So bear with me for the moment and assume all the hype about global warming is true (something, thankfully, it appears people are less willing to do). The problem that I have is that solutions are generated by optimizing along only one axis–reduce energy usage at all costs. It is the “at all costs” part where these solutions go awry. On the face of it saving energy is a good thing even if you don’t believe the planet faces imminent destruction the day after tomorrow. But when “saving energy” becomes “change your lifestyle to be like X, Y, or Z or you are evil” it becomes too invasive.

Suppose we are trying to save the planet by reducing energy used during daily commutes to and from work. Ordinary cars use a lot of energy. Hybrid cars are more efficient so in the one-axis view are a better choice. But the problem with linear optimization along a single dimension is that there is no stopping point. Motorcycles get much better gas mileage than hybrids. Should we pass a law saying everyone has to drive a motorcycle to work? What about people with kids? How does one drop kids off at day care one a motorcycle? The real optimization problem is how do you lower energy use while still maintaining other needs. There are more dimensions to consider than just energy usage.

I’m actually one of those people that own a dog instead of a chicken. Additionally, I happen to be fascinated by dog training. As a result, I drive home during my lunch hour (10 minutes each way) almost every day so that I can exercise my dog and have a training session. I’ve had people tell me that my actions are inappropriate. I shouldn’t waste the energy by essentially doubling my commute distance every day. Now researchers are telling me that I shouldn’t even have a dog in the first place.

To which I say, “Enough!” Asking me to think about energy conservation is fine. Asking me to change a completely reasonable lifestyle in support of a poorly optimized solution is obnoxious.

Penn Jillette is deeply hurt by his idol Tommy Smothers
Does this make George W. a prophet?