Designing Federal Programs That Work

Michael Giberson

From Government Executive:

As Washington considers an overhaul of the nation’s health care system affecting roughly one-seventh of our economy, a critical question arises: How effective are we at crafting legislation that can be implemented? To answer this question, we partnered with Government Executive in 2008 to survey members of the Senior Executive Service….

It turns out if you want to get federal executives angry, ask them what they think of the legislative design process. Their responses can be scathing: “Policy design at the federal level is pathetic.” “Policy design too often is done without consideration of implementation challenges.” The consequences of poor design can be severe.

The prime example of a legislative program design disaster is the original California power market design developed in substantial detail by the state’s legislature.  We get the “cartoon version” of what went wrong in California (“The short answer is energy companies like Enron exploited design flaws in the legislation….”*).

The principles advocated are not bad:

  • Think Design, Not Legislation
  • Involve Implementers
  • Evaluate Workability
  • Probe for Weaknesses

But, essentially, these principles invite politicians to reflect upon the practical limits of their ability to change the world, at the same time they are trying to change the world. How practical is it to ask legislators to be practical?  Other the other hand, making suggestions to legislators must work to some degree or another, because lots of organizations spend billions of dollars in the aggregate to explain to legislators just how, practically speaking, laws should be written.  Why not add program implementation lobbyists to the mix?

If we take legislators out of the picture, however, and think about market design outside of the legislative context, then the four principles can be helpful. Government Executive observes, “If someone isn’t looking for the weaknesses during the design phase, rest assured people will be finding weaknesses in the policy after it is launched – with far more serious consequences.”

It is a sort of mantra (i.e., part of the sales pitch) among applied market design specialists that you can pay a little to test designs in the lab, or end up paying much more to test designs in the real world.  One of the benefits of doing applied market design analysis in an economics lab is that is forces the analyst to answer a lot of useful questions about “how this project is actually going to work” that committee discussions, white papers, and theorists rarely consider (until after an implementation goes badly wrong).

*ASIDE: I think Enron gets mentioned in contexts like this due to its bankruptcy for unrelated reasons and because of a few colorful remarks by traders caught on tape, and the unfortunate and mostly misunderstood law firm memo, and the …; well the list is long, but other companies are as deserving of mention as Enron.  The other companies managed to stay out of the headlines, and Enron was too busy falling apart to fight a post-California market meltdown PR battle.

2 thoughts on “Designing Federal Programs That Work

  1. Enron deserves mentioned because of their input (iirc) into California legislation and to national changes slipped into law by Phil Graham and his lobbyist wife.

  2. I’m still not convinced that Enron deserves the share of the blame they get in the press. Is there any suggestion that Enron promoted the design flaws that they later took advantage of?

    (Does anyone know if there is a good history or even in-depth newspaper report on the lobbying and state legislative activity that might link specific failures to ideas promoted to or by members of the legislature? It would be complementary to the ideas discussed in the Government Executive article if a clear case study of legislative market design failure could be developed.)

    I take “national changes slipped into law” to be a reference to the CFMA of 2000, which didn’t figure into the California power market failures in 2000-2001 so far as I know.

Comments are closed.