The politics of complexity in program implementation

This post will wrap up my series on complexity and program implementation, though I’m sure I’ll return to the topic. In When Things Break I presented the idea of asking  what happens when one part of a system breaks. In The Broken ATM I discussed experiences during fieldwork in rural Brazil that got me thinking along these lines. Then in Wiring Government Programs in Series or Parallel I summarized Pressman and Wildavsky’s argument that the more “clearances” a program needs the more likely it is to fail.  Criticism of the complexity argument and Criticism of the complexity argument, continued summarized some responses to this approach. This post looks what practical knowledge we can take away from all this.

There’s an old joke about every academic title needing  a colon (my titles generally conform to this convention…). Not to be outdone, Pressman and Wildavsky’s title has a semi-colon, too:

Implementation: How Great Expectations in Washington Are Dashed in Oakland; Or Why It’s Amazing that Federal Programs Work at All, This being a Saga of the Economic Development Administration as Told by Two Sympathetic Observers Who Seek to Build Morals on a Foundation of Ruined Hopes 

This hints at some of the political implications of their work. To understand this we need to look at what was going on around the time they published the book and how their argument could be used to support the anti-government backlash of the 1980’s. In spite of this, I think we can still take something away from their insight about complexity. Once we take into account criticisms like those surveyed earlier, I’d argue that instead of conceding that “It’s Amazing that Federal Programs Work at All,” we are left with one consideration for how to make government programs work better.

Pressman and Wildavsky first published their book in 1973. The program it looked at in Oakland, was part of the Johnson administration’s War on Poverty. The New Deal and World War II had both dramatically expanded the federal government. The momentum in the post-war period was for a continued expansion, including The Food Stamp Act (1964), The Civil Rights Act (1964), The Voting Rights Act (1965), Medicaid (1965), Medicare (1966), and the creation of the Environmental Protection Agency (1970).

The 1970’s is when the right-wing attempt to roll this back really got underway, and one of the things it needed was ideas. This account of horrible failure by a well-intentioned program to provide opportunity for impoverished urban African-Americans is clearly useful for someone trying to argue that the federal government just messes up everything it touches. Some discussions of the book rightly frame it in terms of this trend. However,  we’ve already seen that this isn’t the only way to read it.

Here we need to distinguish between describing events and prescribing policy changes. We can accept the description that complexity undermines success without reaching Reagan/Thatcher conclusions about needing the federal government (which–side note–had recently put a man on the moon…) usually mucking things up.

In my dissertation, I propose the following take on  the complexity argument:

…we can restate the claim in terms of relative, rather than specific, probability. Namely, given a particular implementation scenario, adding an additional clearance is one more place for things to go wrong, and thus decreases the chance of success.

So, where does that leave us? First of all, a lot of complexity is necessary. That is, the program design most likely to work is still going to have many components. That does not mean it can’t work, since as we saw there are plenty of examples of complex systems that usually work, but that it may be challenging. Paul Krugman recently wrote that “Obamacare looks the way it does because it has to.” So if something has to be complex we should probably expect unexpected delays as one or more components need to be fixed and expect this to take more time and resources.

Second, sometimes complexity is optional, yet policymakers still choose it. I liken this to an attraction to making up “combo platters” instead going “a la carte.” I’ve already discussed Josh Schultz’s work on health exchanges in Oregon and Washington. In one piece he writes:

Washington defined its scope in the beginning and stuck to “Version 1.0”

Josh and I have been having a great conversation about this implementation literature and similarities between his research on health reform in the U.S. and my research on Brazil’s Zero Hunger. One of the arguments in my dissertation is that unnecessary complexity severely undermined the program in its first year. Even though there were already three existing cash transfer programs, Special Minister for Food Security and the Fight Against Hunger José Graziano insisted on adding a fourth program modeled after the U.S. Food Stamp. Brand new committees with representatives from civil society were responsible for selecting beneficiaries. The beneficiaries were required to provide receipts proving they’d spent their money on food, even though most open-air markets do not offer receipts and they are easy to counterfeit. I could go on, but I’ll stop there.

(Before people get depressed–there’s a happy ending. Within a year President Lula switched to a cash transfer program called Bolsa Família that works quite well. I argue this is in part because it’s design is relatively simple.)

So, at first the lesson of “don’t make things more complicated than they need to be” may seem too obvious to even bother writing about. However, enough policymakers ignore it that I think this is a conversation worth having.