Fire Alarms, "Gasland," and How Policy Changes

Yesterday morning I watched Josh Fox's documentary Gasland and it got me thinking about something that I've been meaning to write about for a while: how and when Congress changes laws and, in particular, laws affecting things like environmental and financial regulation. The basic ideas of this topic are actually among the least complicated I've discussed here at Nathan Explains Science, so this will be (mercifully?) short.

To motivate this, ask yourself this: what's the main thing elected officials have to do? Yup. Campaign. Also, they're supposed to vote on stuff, which supposedly they've read. (They don't, as I believe John Conyers famously pointed out in, I believe, Fahrenheit 9/11. Do correct me if I've mixed up my members of Congress or Michael Moore films.) The point being, they don't have the time, inclination, and probably even ability to actually find out what's going on out there in the United States, or England, or Namibia, or wherever. So what do they do?

Gasland and a different kind of fire.
This is where Gasland comes in, sort of. Some time ago now, there was a paper, "Congressional Oversight Overlooked: Police Patrols and Fire Alarms," by Mathew McCubbins and Thomas Schwartz (Google scholar will get you a copy), and a book, Agendas and Instability in American Politics, by Frank Baumgartner and Bryan Jones, both of which concern how Congress actually deals with oversight and how laws actually change.

The basic version: they don't.

The slightly less basic version: they do so mainly when someone starts screaming and yelling, which usually happens when something really, really unsettling happens. For example, someone makes a movie about how you can set your tap water on fire if you live close enough to a natural gas extraction site where they've used hydraulic fracturing to free the natural gas.

To be more precise, Baumgartner and Jones's and McCubbins and Schwartz's point is that Congress reacts. In McCubbins and Schwartz's words, they rely on "fire alarms" rather than "police patrols," in which version they would have to do a lot of work themselves. "Fire alarms," on the other hand, put the burden (rightly or wrongly) on others to bring problems to Congress's attention.

Why Congress is nonetheless rational; the consequences, scientific and otherwise
McCubbins and Schwart's paper is really about why this behavior is actually perfectly rational, and you can maybe see what their argument is: it takes a lot of effort for Congress to patrol everybody, and most of the time everything's going to be okay, so better (for them, at least) to rely on others to do the policing. Or fire-alarming.

Baumgartner and Jones's central observation is that this behavior explains some curious things about how funding changes over time. For a long time, people believed that policy changed incrementally—funding just sort of increased a little bit over time, that policy changed slowly in response to problems, and so on. Well, that's not how it really works. Instead what you see is a kind of punctuated equilibrium. Things are incremental for a while, but then Three Mile Island happens, and all of a sudden regulation gets tight and global nuclear reactor construction slows to a near stand-still.

Or, think about the recent financial crisis: there were signs things were about to fall apart, but until Lehman and Fannie and Freddie collapsed, nobody actually did anything.

If this seems a bit discouraging, well, it is. It means that you have to get people's attention—and usually, you don't actually have control over getting people's attention, since people aren't too likely to organize—to change anything in your favor. Think climate change: until we look around and realize we're in typhoon in Indiana, we might not start yelling, and so Congress might not do anything serious.

Too bad drowning people don't actually shout very much.