Last week, Russ Roberts (of ever excellent Econtalk) wrote a blog post about economics and engineering. I was surprised by how much I disagreed, though not in the way you might expect.  His point: looking at the economy isn’t anything like an engineering problem. The economy is complex and emergent while engineering problems can be modeled very accurately by simple equations:

“Running an economy is not an engineering problem. There are no simple equations that describe its motion that are akin to the engineering problem of space travel.”

 The gap between economics and engineering is not so large as Russ makes it out to be, but the similarity is in the opposite direction from the normal misconception: rather than economics being captured by simple equations like engineering, engineering is more complex and emergent than most people (even engineers!) acknowledge, like economics.

 Sure, engineering theory can be captured by simple equations (“it’s just physics!”) but then, so can economics. It’s just easier to get to ‘the emperor has no clothes’ point in economics. You can write down the rocket equation, F = MA, do some orbital dynamics and say, “yup, that’s how we get to the moon” but this is what the wiring of the Apollo computer looked like:



Just as in economics, the gap between neat equations and reality is huge. Non-optimal solutions are built on non-optimal solutions because technology has become so complex that you can’t just start from scratch. A huge amount of programming is done using Unix command line commands and the C programming language, both built in the 70’s and flash frozen into the system.

 "Ironically, in an unplanned economy, shopping is usually as straightforward and predicable as space travel. It becomes a engineering problem–what is the shortest route to the grocery–what is my optimal path through the store given my shopping list."

 From my perspective, Russ’ wonder at how well the unplanned economy works, rings just as true for engineering.