Friday, September 14, 2007

What We Need Is a Smart, Hungry Beast

Ezra Klein references this David Sirota piece on the Colorado DMV:

While I smiled for the camera, I considered this mundane encounter with state government in economic terms. Between all the people I waited with, about 600 combined hours of economic output was extracted from the state and thrown away. Multiply that over an entire year throughout any given state, and you see how poorly run public services take a severe — and hidden — toll on a state's economy.

Services, of course, do not fail in a vacuum. They fail because budget cuts leave them lacking adequate resources to succeed. While Republican economics teaches that less government spending means a stronger, more efficient economy, my experience at the DMV suggests otherwise, as does this state's overall experience as a test tube for conservatives' budget and tax doctrine.

OK, so Poorly-Run Services Are Bad. But it's a pretty big leap from "the DMV is poorly run" to "the DMV is poorly run because of budget cuts." Is the DMV slow because it needs more staff? More automation and less staff? Some decent process design? Better trained staff? Fewer transactions that require mediation by the DMV? There are lots of possibilities. Not all of them require a budget increase and some of them actually allow a budget decrease. What's the best way to encourage efficiency? Pump money in or restrict it somewhat?

This is not intended as a "starve the beast" argument. But I will assert that there's an optimal band of expenditure. Too small and you're sub-optimal, but too large and you're sub-optimal as well. A vast number of enterprises don't benefit from throwing money at them.

The trick is to design a funding methodology that explores and optimizes the problem space you're addressing. There are lots of techniques to do this. None of them have been applied to budgetary policy. Anybody up for a simulated annealing funding law?

No comments: