Things I’ve been thinking about: May 1st, 2014

1. Can Elinor Ostrom’s work help us understand what to do about antibacterial resistance? I’ve been reading a lot of Ostrom in an attempt to develop an opinion, but it’s tough going for me.  Econ was never my strong suit. Anyone smarter than me want to chime in and explain it?

2. How do we balance the importance of innovation and learning from our mistakes with the importance of not wasting our finite? Leaving room for failure is important, but people get hurt when we fail. It’s a confusing and ugly dilemma and it’s much less clear-cut than anyone wants to admit. Here’s one interesting take on failure.

3. Worms. The good and the bad. I am now the proud owner of a worm box for composting, which brings me a genuinely embarrassing amount of joy. I have also been thinking about helminthes – the intestinal worms that can lead to malnutrition or even death. I suspect we’re going to find out they matter a lot more for children’s health than anyone ever realized.

4. The new David Roodman / GiveWell report that finds that reducing child mortality does not lead parents to have fewer children. When I was writing What’s Killing Us, I looked for research linking reduced child mortality to a decrease in fertility rate. I assumed it was true. It seems true. Heck, it seems obvious. But I couldn’t any definitive research indicating it was true, so I left the claim out. Interesting to see that backed but by someone who did a serious study.

(photo credit: me)

circus, circus

Last week I took my son to the circus. Specifically, a traveling troupe of Chinese acrobats. It was quite clearly the troupe that plays Dushanbe, not the troupe that plays Moscow. They attempted the big stunts, but they didn’t always make it. Spinning plates got dropped, a human pyramid crashed, and one tumbler tumbled right off.

This is what interested me: it didn’t affect the show. They were ready for failure. They had spare plates standing by for quick replacement after droppage. The ribbon twirlers had fresh ribbon at hand in case of tangling. The air acrobatics had truly fantastic spotters. Everyone who fell had at least one person gracefully rush up to soften their fall. They responded to errors so quickly and smoothly that it was like a dance.

Ever since I saw the show, I’ve been wondering how we can build that kind of resilience into development interventions. How can we make sure our errors don’t wreck our work? One thought: maybe ongoing monitoring is the equivalent of those dedicated spotters who saved the falling acrobats. Collecting implementation data will let you know if your human pyramid is going askew, or keep the guy on the springboard from bouncing onto hard ground. Another: you have to be profoundly humble and honest to prepare for failure that way. You have to admit, up front, that mistakes are possible. If your spotters are hiding in the back room, they won’t catch the tumbler in time. You can’t seamlessly replace a knotted ribbon if the new one isn’t right next to you.

It’s a beautiful analogy. Would it be allowed in real life? True, some people do call this industry a circus. But do our donors actually want us to be honest and humble? Would people think we were just incompetent if we visibly prepared for failure? And what, exactly, would preparing for failure look like?


Photo credit: Don Fulano

Now picture that top girl falling, and landing in the arms of a costumed spotter

What we can’t do, part I

Thdepressing picture of a muddy streetere have been an awful lot of people I haven’t been able to help. My career feels, sometimes, like a long list of things I haven’t been able to do, punctuated by the occasional success.

I know that isn’t unusual. When you live in a poor country, you are constantly assaulted by the terrible need of the people around you. Our ability to respond is limited by so many things – program scope, funding, human capacity and host country conditions – just to start. There is never enough money to do everything, or you need to branch out into some new area you know nothing about. Sometimes the problem is caused by destructive traditional practices or bad government regulations.

At times, you can’t help people because you failed. Your program just got it wrong. You trained doctors but they didn’t change their behavior afterwards. You wasted your money and their time and no patients benefited. Or the broiler chickens turned out to cost more to raise than they earned when you sold them. Or your families sold the vegetables from their kitchen gardens and used the money to buy sugar and children’s nutrition actually got worse.

You can make bad choices with the best of intentions, you can discover your every choice has unintended consequences, and you can just be flat out stupid. Luckily, we’re not houseflies. We have the capacity for learning. And if we’re willing to genuinely examine our failures, we can avoid making the same mistake twice. It’s hard, but it’s possible.

I’m not saying that failure is a good thing. No one wants to waste their limited resources – time, money, and community commitment. And most of the time failure isn’t failing well – it’s just an ugly mess. But you can learn to fail well, and over time most of us learn how to.

For me, at least, it’s not failure that devastates. It’s the sheer scope of the problems we face and the tininess of our ability to help. Even the most holistic project has its limits. You simply can’t tackle everything all at once. But as you live your life, everything all at once is what you see.

When I was living in Uzbekistan in 2005, there was an incident in a conservative city, Andijon, in the Ferghana valley. A protest got out of hand, leading to a break-in at the city jail and a massive demonstration in the main square. When the police got involved, it ended in violence. Somewhere between 169 and 700 people were killed. The Uzbek government holds that those who died were terrorists; NGOs in the country report deaths among innocent civilians, including women and children. It’s been a source of a lot of controversy.

Nobody, however, denies that it was bloody, terrible, and heartbreaking. The deaths in Andijon left the whole country stunned. My office manager came to me in tears; he was thinking of quitting his job. What is the point, he wanted to know, of running a health project when there were so many other things going wrong in his country? Training pediatricians struck him tiny and useless.

He had a point. Most of what we do is tiny and pointless in the grand scheme of things. One average-size project isn’t going to have much impact on an entire country. That is brought home to us every day, all the time, as we live and work in the developing world.

If you’re working for an HIV project, helping people access anti-retrovirals, you know you’re saving lives. If you visit a clinic that is giving out the drugs, you can actually watch people get healthier over time. But what about all the people who don’t have AIDS? What about your neighbor, whose mother has cancer and there is no treatment available in the country for it? What about your friend’s son, who has no way to pay for university? What about the woman down the street, who always has bruises and you can hear the shouting in her house? And the children begging in the street, or the local school which has no windows or books?

(photo credit: me)

On process

A friend of mine recently attended a meeting that was intended to develop a process to guide the preparatory meetings for the coordination meetings with the Ministry of Health. And the thing is, when you’re in the thick of it, these meetings make sense. You do need a unified message before you talk to your host government, and without some ground rules, the prep meetings to develop that message can get genuinely ugly.

All of this led me to think about process, and its sibling, bureaucracy. I’ve always had a pretty unpopular belief in the value of bureaucracy. Bureaucracy, to me, is the core of an organization. It’s what keeps an organization functioning when its staff changes. Without the forms and the regulations, you don’t have an organization. You have a cult. The structures are what makes it about more than just whoever works there at the moment. Bureaucracy puts the “organize” into organization.

That does not mean, however that bureaucracy should rule your work. It’s supposed to help the work get done. The work does not come second. And everyone accepts that. In fact, most non-me people hate bureaucracy.

Unfortunately, this is not the case with process. Plenty of people will sing the praises of “process.” Having a well-thought-out process means that you are Doing a Good Thing. If your intervention fails – the village women don’t feed their children more beans, or the Ministry of Education refuses to adopt your snazzy new curriculum – well, at least your process was good. Everyone benefits from being part of it.

I call shenanigans. Process is a jargon word that we use to obscure what’s going on. If your process is a series of meetings (and it almost always is), say so. And a good process is a process that achieves your goals. No more and no less. Nobody benefits from your stakeholder interviews if their input never turns into anything.

Lastly, some food for thought. A project I was connected to wanted to solve a problem they were seeing in a lot of rural clinics. The clinics would just use up all of their medicines, and then request more from the central supply. Since new drugs didn’t arrive instantly, there could be stock-out periods of a week or more while they waited for the new drugs to come.

To fix this, the project wanted to implement a pharmaceutical logistics system. They brought in a consultant from Europe, who worked with a group of clinic managers and Ministry of Health staff to estimate ongoing demand from different kinds of drugs. Based on these estimates, they then set re-order points for drugs. So, if you distributed, say, 10 IUDs a week, you would reorder IUDs when you were down to 15 of them, giving you a week and a half of time until the new ones came. The consultant turned these plans and estimates into a training system, and the project went around training rural clinics to use the new method.

Nobody ever did. Despite the training, and the eminent logic of the system, nobody ever did. Rather than try to determine why, the project wrote off the exercise as a failed pilot project and carried on. (My own suspicion is that clinics ordered their drugs when they knew that central supply had them, and were afraid that if they ordered according to some system, their orders would go unfilled.)

One of the project staff, when describing the whole fiasco to me, said something I’ve always remembered. “We paid thirty thousand dollars for the consultant, the curriculum, and the trainings,” he said. “If we’d given that thirty thousand to the government in return for a promise to improve their ordering system, every clinic in the country would be using it by now.”

(photo credit: markhillary)
Chosen because that’s exactly like many of the meetings I attend.

Things I believe in #17 – recognizing and learning from failure

There are a whole lot of ways you can screw up a development program. I’ve talked about some of them in the past. You can make bad choices with the best of intentions, you can discover your every choice has unintended consequences, and you can just be flat out stupid. Luckily, we’re not houseflies. We have the capacity for learning. And if we’re willing to genuinely examine our failures, we can avoid making the same mistake twice. It’s hard, but it’s possible.

Determining what went wrong is actually the easy part. There are so many moving parts in any development project that many, or most of them will not turn out how you expected. Making a list of what could have gone better will be distressingly easy.

The hard part comes next. You have to sort through your list to figure out which factors were actually the deal-breakers, and which were annoyances. Then you identify what you can actually avoid or compensate for next time. You have to figure out the right lesson to learn.

For example, if your agriculture program failed because of insufficient rainfall, you need to design a program flexible enough to adapt to multiple weather projections. Knowing that you need rain is not a useful lesson to learn. Knowing you need to have different options for different weather conditions is a useful lesson. Vasco Pyjama has a great entry about learning from something her organization failed at.

Being able to recognize and accept our failures is what makes future success possible.
To take this discussion from the theoretical to the concrete, here are two of my own programmatic failures, and what I learned from them:

1. Nutrition pamphlet for women in Turkmenistan. Literacy is high in Turkmenistan, and people seem to genuinely love brochures and pamphlets. When we focus-group tested health pamphlets, we got responses like “this should be longer.” When a Peace Corps volunteer applied for a small grant to produce a brochure with nutrition information and healthy recipes, it seemed like a good fit.

This is what I learned: Turkmen women don’t cook from recipes. They think the whole idea is weird. They feed their families by buying what’s cheap and in season from the bazaar, and then cooking it using traditional techniques. The nutrition information was useful, but women had no way to act on it. Our next attempt at nutrition education focused on cooking methods to make Turkmen dishes higher in nutritional value. That one was much more successful.

2. Community website for Baghdad. This was a proposal I helped to design and write. We were going to set up a community website for Baghdad, kind of combination between Craigslist and Yelp. It would allow for violence mapping, to help people avoid dangerous areas, and offer ways to discuss things like whether a particular doctor or lawyer would take clients across sectarian lines. We were going to combine that with increased wifi access using repeaters, and providing cheap laptops to large households. In an environment too dangerous for community-building in person, it was meant to help give people a unified identity as Baghdadis once again. Just about everyone who heard about the idea loved it, but I never found a donor that would pay for it.

This is what I learned: no donor will fund a program that gives people to the chance to say anything they want. As one person told me, “The first time the New York Times runs an article about Islamic militants posting to a US/UN/UK (insert donor here)-funded website, we’re all screwed.” I still love the idea of the Baghdad site, but it’s going to take some very innovative funding to make it happen.

Failure feels really, really bad. It’s a blow to your ego and you feel like you’ve let down the people who rely on your organization. If you can get past that, though, it’s the best teacher you’ll find.

(photo credit: Quod)
Chosen because I think that tiny apple on the right represents hope.