More Art than Science

Today, during my annual physical, my doctor, who is well known and widely quoted in the Washington region, asked me what I did for a living.

I said "Oh, I work in international development." And then, after a pause, I said, "We are trying to become more scientific, by doing more randomized controlled trials like you do in the medical field."

He looked up at me sharply and laughed.  "Are you crazy?" he replied.  "Only a very small proportion of what I do is based on randomized trials.  In fact, most medical treatments are not based on rigorous evidence at all.Whatever anyone might tell you, people are just too complex and too different to do lots of good randomized trials."

"Well, then, why do people say you are such a good doctor?" I asked him.

"I guess I may be better than other doctors at trial and error.  I just get people to try things, and if they don't work, then I get them to try other things.  Being a doctor is really much more of an art than a science."

World Bank as Convener 1st, Lender 2nd?

"Are you sure we should do this in the World Bank atrium?"

"Yes, sir," I said.  "Why not?"

I tried to act confident, but I was terrified.  It was February 1998, and on the phone was a managing director, second only in power to the president of the World Bank.  He was known for asking questions that were actually orders. Arguing rather than obeying was considered ill-advised.

"Well," he said, "we may have outside visitors that day.   Do you think they should see some of those crazy ideas that staff may come up with?  What if we relocated the event to the basement, out of the way?"

A small innovation team, with encouragement from Jim Wolfensohn, the Bank's president, had decided to launch the first-ever Innovation Marketplace, a one-day event when any Bank staff member (regardless of title or seniority) could propose an idea for helping the Bank better fight poverty.  We would award $5 million to help fund the start-up of the most promising ideas. We had decided to hold the event in the atrium of the Bank, a beautiful space that soars thirteen stories.

The only problem was that the atrium was almost never used for public events; it was always eerily quiet and deserted, even sterile.  You never wanted to linger there - you always hurried through. The culture of the Bank at that time was such that everyone worked behind closed doors, beavering away on in-depth country or sector studies designed to find out the right answers to various development challenges.  Staff emerged only for the occasional review meetings, or to go on mission to the countries they worked on.

My, oh, my how the world - and the World Bank - have changed since then.  A couple of months ago, I went back to my old stomping grounds to witness an Apps for Development competition. Not only was the Bank's atrium a beehive of activity, but there was even techno music pumping up the mood of the crowd before the announcement of the event's winners.

Stephanie Strom recently wrote a nice article in the NY Times about how the Bank is opening its "treasure chest of data" for researchers and others around the world to use.  The articles has good insights into the benefits of making data public; and the fact that the World Bank (previously among the most secretive of aid institutions) is now making such huge strides is sure to encourage other agencies to do the same.  (The UNDP also created a very impressive open data portal for some of its projects.)

Sharing data is a big leap, but there is now a window of opportunity for the Bank to do something much, much bigger.  Back in 1998, with my heart pounding, I stood my ground with the Managing Director, and the Innovation Marketplace was a huge success; by cutting through the usual layers of bureaucracy and giving everyone an equal voice, the event allowed all sorts of good ideas to bubble up, any many of them soon became major strategic.  In some sense, we were democratizing the World Bank, even if for only one day.

Building on that success, in early 2000 we went on to launch the Development Marketplace, which allowed anyone in the world (not just the Bank) to propose an idea for funding.  The planning of this event also generated a lot of controversy (and not only from senior managers this time), but it was a big success, too.  And though I left the World Bank shortly afterwards to co-found GlobalGiving, the Bank went on to replicate the Development Marketplace many times, including in some seventy countries around the world over the past decade.  Country directors would sometime report that it was the first time they had been able to get civil society groups and government officials in one room together talking about ideas and solutions rather than just arguing.

Nonetheless, despite their success, these marketplaces have remained peripheral to World Bank's main business.  In the decade since my departure, many former colleagues complained to me that the Bank was failing to innovate in ways that would keep it relevant for a changing world.

That may be about to change.  Bob Zoellick, the Bank's current president, recently talked in a speech about the "democratization" of development, where the Bank and other aid agencies no longer pretend to have a monopoly on understanding problems and devising solutions. Bank experts would still have a great deal of technical expertise, but the role of Bank staff would shift.  Instead of trying to find and hire the elusive "best expert in the world on subject X," the Bank would hire very good experts who are capable of leading a conversation among other experts, government officials, and regular citizens about the most pressing problems and the most viable potential solutions.  Bank staff would in a sense become "hosts" of conversations about what to do, and then would have the ability to financially support the initiatives and approaches that arise from these conversations.

Hosting effective conversations is hard, and it may be the most in-demand skill at the Bank in the decade ahead.  If the Bank can make progress in this area, however, the payoff for the institution could be large.  In addition, by modeling openness and an ability to listen, the Bank would be putting indirect pressure on governments around the world to do the same.  If regular people are able to comment on and even contribute to the design of World Bank projects, they are surely going to begin demanding the same treatment from their own governments.  The resulting increase in citizen voice and government responsiveness could end up having a far more important impact than any particular Bank project(s).

Zoellick has assembled a solid team to help the Bank remake itself.  Sanjay Pradhan, Randi Ryterman, Aleem Walji, and others are helping lead a conversation within the Bank on ways to cultivate new thinking and approaches. As the NY Times article notes, significant culture change is going to be required, and that is  always tough.  And change will also demand hard thinking about the World Bank's business model, which currently relies on generating a spread on its loans and other financial instruments.  Incentives within the institution remain tied to the ability of staff to make loans and help the institution generate the income that it needs to operate.

The good news is that the best staff at the World Bank are leading the way.  A while back, some of my former colleagues hosted an informal all-day Saturday session for health officials in a Latin American country.  No ties were allowed, and there was no rigid agenda.  When I told one of the Bank conveners that I was sorry he had to work on a Saturday, he told me that it was one of the most productive days of his career. "For once, we did not give them a long lecture," he told me.  "We just served them pizza and kept the conversation headed in the right direction, injecting bits of information but not pressing my own views too hard."

As a result, the health officials swapped stories about what was working and what wasn't in their country.  They were able to be candid about their failures as well as their successes.  The conversation was not about What is the RIGHT ANSWER? Rather, it was about What are some reasonable things to try in our country, and how can we best evaluate the results? The participants liked the experience so much that one of the officials present hosted a similar session, with Bank help, for other countries in the region. The effort led to a great deal of social capital that allowed successes and failures to be honestly shared, thereby speeding experimentation and improving feedback loops.

My friend told me he felt he had helped advance reform more on that single day than in months of formal meetings and expert presentations.  Instead of being a Bank expert pushing for the RIGHT policy, he was helping a country's own experts iterate toward an approach that would work in their context. Furthermore, the trust and social capital that emanated from these meetings led the countries involved to seek several hundred million dollars of funding from the Bank to support their reform agendas.  The revenue from these loans in turn enable the Bank staff to provide additional intellectual support, including hosting more conversations.

Got to Admit it's Getting Better

 Over the last one hundred years, the physical well being of the world's population has improved far more than in all of the previous natural history of humankind.

Charles Kenny's recent book Getting Better is an antidote to the pessimism many of us feel about the state of the world.  Those of us in international development f are frustrated we have not been able to find a toolkit of approaches that reliably increases growth in poor countries.  Kenny acknowledges this frustration, but shows that many indicators of wellbeing have improved rapidly even in the absence of consistent economic growth:

Global average life expectancy increased from around thirty-one years in 1900 to sixty-six by 2000.

In terms of GDP, there is an increasing divergence among countries in the world, with many of the poorer falling further and further behind.  But with respect to other indicators of wellbeing, there is a striking convergence.  Infant mortality has plunged by over half in eighty percent of the world's poor countries in the last fifty years.  Literacy rates in sub-Saharan Africa rose from 28 to 61 percent from 1970 to 2000.  Some very poor countries such as Vietnam have achieved 95% literacy.  Democratic ideas and practices are spreading rapidly (notwithstanding big setbacks and uneven progress), and violence is way down: global homicide levels, for example, are about one-third of what they were in England in the Middle Ages.

How can this be?  It turns out that it is not very expensive to improve these aspects of wellbeing. This is because of tremendous innovation in technologies and ideas in some domains, ranging from vaccines to understanding germs to oral rehydration therapy to increased agricultural productivity.  In the old days, even the wealthiest and most powerful of kings died early for lack of the cheap modern vaccines that keep the most vulnerable and poor people alive today.

The state of Kerala in India, with a per capita income of $300 per year, has been able to achieve a life expectancy of 72 and a literacy rate of 91%.  By comparison, the US has a per capita income of $29,000 and has a life expectancy only five years greater than Kerala. Lest we think that achieving those extra five years of life expectancy requires huge resources, take the case of Costa Rica.  Life expectancy in Costa Rica is two years greater than the US, but income per capita is only $6,500, and Costa Rica spends only $305 per capita per year on health, compared to $5,711 in the US.

Getting Better is full of facts and figures such as these.  Kenny neither minimizes the remaining problems (poor people do have fewer opportunities and poverty does equal misery for millions of people), but he does want to zoom out and take the long view so we can see we are making progress.  He does note that the spread of even the cheap and effectives innovations he discusses has been uneven - and thus there is great scope for further progress.

Progress must often come on the demand side rather than the supply side, Kenny argues.  In many (not all) cases, there is enough money in the budget for education and health programs, but people don't demand the right services for lack of knowledge about their potential effects.  How to do better marketing is thus a key challenge for the aid business - an area in which we have made only modest investments.  Relatively small investments in transparency measures can also help stimulate demand as well.  Local people are often told that there are not enough resources in the budget to do X, Y, or Z, but once local people are able to see what the money is actually being spent on, they are able to bring to bear pressures to get what they need.

Kenny closes the book with suggestions for how to catalyze the flow of ideas and hasten the uptake of innovations in health, education, and other areas.  This is the weakest part of the book, but at least he is offering a framework for further analysis and experimentation.  The bad news is, as Kenny says, we don't know how to make people financially richer. The good news is that we do know how to make them better off, and that fact should motivate us all.

Bad News for Foxes?

My friend April Harding sent me this talk at the Long Now Foundation by Philip Tetlock, who has spent many years studying how well experts perform at predicting future events.  Tetlock's research shows that foxes are better than hedgehogs at predicting the future.* And interestingly, the less confident people are about their predictions, the more likely their predictions are to be accurate!  (Foxes also tend to be less confident than hedgehogs.)

This result makes many of us (myself included) happy, because we consider ourselves foxes.  The only problem, according to Tetlock, is that even foxes barely edge out simple mathematical algorithms** when it comes to accuracy.

----------------
* Isaiah Berlin famously defined hedgehogs as people who have one big idea through which they interpret everything they see.  Ideologues of all stripes are examples of hedgehogs.  Foxes, by contrast, bring an eclectic set of perspectives to bear on their interpretation of events and phenomena.

**For example, things will continue unchanged.  Or, things will continue to change at the same rate as in the recent past.