In a recent post, I wrote:
I think progressives would benefit, alongside the rest of us, if their movement put a higher premium on simplifying government, making it transparent, and avoiding the implementation of structures certain to be captured by special interests at some point, no matter how elegant or appealing their initial design.
Now I’ve got an example! It is owed to Conor Clarke, whose blog is a worthwhile daily read. He writes:
One thing you can say about the new “Cash for Clunkers” legislation — approved by the Senate yesterday and heading to Obama for a signature — is that it’s not as bad as it could have been. And yet the project started with such high hopes! You had Alan Blinder calling it a “remarkable” public policy hat trick: We could stimulate the economy, improve the environment and reduce income inequality all in one go! And all by having the government buy and scrap some crappy old vehicles.
The idea sounded nice, but it was not hard to see the problems. If the government artificially raises the price of crappy old cars, it will create market for fixing up crappy old cars no one drives — hardly an efficient public policy goal. (Last year Steven Levitt hypothesized that the program might actually increase the number of Clunkers on the road.) And if the program lasted for any extended period of time, the effects on income distribution would be tiny. Sure, the immediate beneficiaries would be the current owners (Blinder says most Clunker owners are downscale families). But, after the initial round of sales, the price of used cars would go up to reflect the government subsidy. That wouldn’t exactly be a boon to the driving poor.
But, as near as I can tell, the Bill that passed the Senate avoids the big problems. Will Wilkinson and others have pondered some clever schemes for ripping off the federal government and destroying the environment at the same time, but I haven’t seen one that will work. Why can’t Will sell his fuel efficient Civic, buy a smoggy old truck, trade it to the government for $4,500, and buy a brand new car that pollutes more than the original Civic? Because your clunker must be “continuously insured and registered to the same owner for at least one year” to claim a voucher, and the window for the trade-ins only runs from July 1 to November 1, 2009. (On the other hand, the trade-in policy is so narrow and confusing that I’m not sure there’ll be lines running around the block to take advantage of it.)
The main problem with the bill, I think, is one Ryan Avent latched onto last month: It’s poorly targeted. You can trade in an SUV that gets 17 mpg and buy one that gets 19 mpg with a $3,500 helping hand from the federal government. On the other side of the docket, you can scrap your 19-mpg passenger car, buy a 30-mpg one, and get zilch. (Or maybe a thank you.) That’s because the voucher benefits are all or nothing, depending on whether you cross the 18-mpg line.
Maybe there are some net gains for the environment in there. But the structure looks more like a gift to the auto industry than a gift to mother nature.
As the excerpt illustrates, these failures were all too predictable — when a huge well-connected auto industry stands to benefit from a certain piece of legislation, and the other beneficiaries are diffuse poor people and Mother Nature, which way do you think it’s going to skew? When you have to draw complex rules to keep people from gaming the system, how efficient is the policy ultimately going to be?
There are things to admire about the progressive embrace of wonkery, but this example illustrates a shortcoming — legislative schemes that seem targeted and efficient in a white paper are only enacted if they pass through a political process that inevitably corrupts them.
So what would I ask of progressives? That they factor the most predictable ways legislation changes into their wonkery, and occasionally conclude that though the platonic iteration of their clever idea might work, it is nevertheless a foolish idea to pursue, because the actual outcome is likely to be a costly boondoggle.