The Broken ATM

In my post When Things Break I mentioned thinking a lot about a broken ATM in Piauí, Brazil. I should back up. I was in Brazil the 2010-2011 academic year doing research for my dissertation on Brazil’s Zero Hunger. During that time I lived in Brasília, where I improved my Portuguese and interviewed people working for the federal government and NGOs.


Photo credit: The Express Tribune.

When Brazil’s President Lula (2003-2010) first launched Zero Hunger in 2003, the Ministry for Food Security and the Fight Against Hunger (MESA) selected two pilot municipalities in Piauí. The idea was to use these as policy laboratories and then scale programs across the country. I knew I couldn’t just get the perspective of people in the capital; I needed to talk to people in the Northeast and hear about their experiences during Lula’s government. So, in June 2011, with my Portuguese much improved from living with three Brazilians, I headed to Piauí.

I packed a lot of interviews into a few days, and they spanned a lot of topics, but one key one was Bolsa Família. This is a conditional cash transfer (CCT) that gives small amounts of money to very poor families. The conditional part is that they’re required to send their kids to school and to take part in preventive health programs, hopefully leaving the next generation better equipped to escape poverty. I’ll be writing plenty more about it over the next few months as I try to continue to break my research into easily digestible pieces.

One of the things I was interested in, though it wasn’t one of my central research questions, was if they felt like it was stimulating the local economy (a “virtuous cycle” in Portuguese, a “multiplier effect” in Economic English). Some people told me that sometimes the ATM where people could withdraw their money would break and people would have to go to a neighboring, larger municipality to get their money. Since they were already there, they would do much of their shopping there.

I don’t know how big of a problem this actually was and how much money it took out of circulation in the municipality. What made it interesting to me is that, in all my years of reading, writing, and thinking about this program I’d never thought of something like a broken ATM as being relevant. 

Political science has a concept of “veto players”: people with the power to stop something from happening, either individually or collectively. The more places it’s possible to kill a proposal the harder it will be to pass. What if we thought of the broken ATM along similar lines. Now, an ATM obviously has no opinions on policy (that was an odd sentence to write…). It’s not going to deliberately stop something from moving forward, but it can still have that effect.

I wondered if we can call this a “micro veto player.” It doesn’t kill the entire policy, but it undermines it. It is also limited in geographical scope (one municipality) and time (until it gets fixed). Note also that we’re looking at undermining the implementation of something that has been approved, rather than undermining the initial passage. Breaking down doesn’t stop people from getting their money, but makes them expend more effort to get it and undermines the multiplier effect part of the program.

While riding on the bus I journaled about this. It seemed to me that the more pieces something has, be it a physical object or a government program, the more places it can break. Thus, all else begin equal (which thing never are), a program with more pieces is more likely to break.

Later that month Federal Senator Cristovam Buarque put it to me this way: “The program with many components very rarely succeeds.” (“O programa com muitos componentes pouco vezes da certo”).

This seemed a simple enough idea that there had to be work already making a similar argument. Indeed, as I mentioned in my last post, people there had been a discussion along these lines for some time. I just hadn’t read that stuff yet, so this seemed new to me, though. Next time I’ll highlight some key points from that literature.

Next post in series: Wiring Government Programs in Series or Parallel.

One thought on “The Broken ATM

  1. Pingback: How Cover Oregon fits into my discussion of complexity | Russell Terry

Comments are closed.