Photo by The Research Desk/The Gazelle

AMA: Peter van der Windt

This week, The Gazelle’s Research Desk held a live Ask Me Anything session with assistant professor of Political Science Peter van der Windt. For an ...

Nov 21, 2015

Photo by The Research Desk/The Gazelle
This week, The Gazelle’s Research Desk held a live Ask Me Anything session with assistant professor of Political Science Peter van der Windt. For an hour on Nov. 18, Professor van der Windt answered questions submitted by our readers and editors. In the session, he discussed his work in the Democratic Republic of Congo as well as his thoughts on the study of development interventions.

What is so specifically insightful about Africa in the larger scheme of development studies?

Africa is the continent that most people think of when they think about development. Of all continents, Africa is probably the most plagued by war, famine and bad governance, among other things. Thus it makes for a good place to study these things. However, most poor people do live outside of Africa. Think of India and China.

How did you end up focusing your research on the DRC?

I know I should write a very inspirational story here, but I don't have one. I was always interested in Africa — I set up a student association related to development issues at my alma mater, and I studied for a year at the university of Pretoria in South Africa. But why Congo? My professor at [Columbia] graduate school asked me, "Peter, I have a project in Congo. Do you want to be my field man there?" And I said yes. I quickly learned French and I was off to the Congo. The first trip was only one month, but I immediately fell in love with the country and its people.

What has working with the World Bank been like?

Very frustrating, very often. I would need several hours to talk about this completely. It is a really large organization with a lot of bureaucracy, red tape and stupid rules. Things move slow, many people are involved in decision making. On the other hand, it also allows you to do research that otherwise would not have been possible: The budgets are large, many of the people that work for the World Bank are very smart and very dedicated. So it's a bit of a mixed bag.
I have tens and tens of stories about working with NGOs and the World Bank that would make for what I think would be a very fun book. And I say this after working with these organizations for only a few years.

If you write a paper about the benefits of an aid program and William Easterly responds proving you wrong, would you be thrilled by having been blessed with an Easterly refutal or devastated by his usual academic ruthlessness?

I would be very happy. It means the paper was good enough for Easterly to read and respond to. More seriously, this is exactly what we want in academia. One person writes a paper and really thinks they did the best they could. Then somebody that is smarter, has better data or a new point of view looks at the same question and comes to a different conclusion and then interacts with the original article. Then more people join the discussion. And that is how we learn. That's what we do — or at least what we should do — in academia.

Cash transfers: conditional or unconditional? Why?

Up to a few years ago, people would have said, "Conditional, of course. Otherwise people would squander the money. Men would buy alcohol, it's better to give cash only under the condition that the households’ children go to school, etc." Over the last few years, however, more and more research has come out arguing in favor of unconditional cash transfers. The idea is often very simple: People themselves know best what to use the money for.
Ultimately, a major problem with conditional cash transfers is that you have to be really sure that the conditions you're putting in place are good. Even pressuring kids to go to school could have negative consequences: They won't be able to work in the fields, which means less income for the family, etc.

Could you tell us about an exemplary international development intervention program? What made it so effective?

It's actually quite a difficult question and the reason why is that it is difficult to measure if a development intervention worked or not. It has only been 10 years since we started using randomized control trials, which allow us to make causal statements. What we have seen over the last years is that the more rigorous the evaluation techniques, like using impact evaluations or RCTs, the fewer programs that seem to live up to our expectations.

What are some of your hobbies?

I played poker with friends until late last night. (That's why I was yawning in class today.) I like hanging out with friends. I also like traveling. I'm in a plane a lot, probably a bit too much. But I really have this feeling that there is still so much out there in the world that just has to be seen and explored.

Could you tell us more about the student group you founded in Tilburg University related to development studies?

The main reason why it was founded was a frustration from my side with my own study, economics, and others at the university. It was all focused on business, earning money, etc. Important international issues such as human rights, poverty and the environment were not discussed. The idea was to create a forum where these topics did appear. The association was very popular. I think within a year we had over 200 members and 40 active members. It still exists: AWAke International.

What have been the major ethical issues you have faced as a researcher in the field? How have you dealt with them?

You'll face them often when doing fieldwork. A drunken rebel might stop you on the road and ask for money. Do you give it? You leave a small village in a big 4x4 to the city and have a space free, someone in the village is sick and needs a ride. Do you take them with you? Probably yes. What if two people are sick and you only have one space available, but you know you have a satellite phone in your pocket and money in your bank account to have another car sent from the city to the village. Do you do this?
The most complicated episode I faced was during the Voix des Kivus program, in which we gave individuals in the DRC cell phones to report on events in their village. People often reported on violent events, rebel group presence, often even telling us the name of the rebels or government soldiers that were committing crimes in the villages. It could have happened that a rebel group learned about the program and killed the reporters. What happened [in our case] was less extreme.
Now you have your data; with whom do you share it? Everyone publicly, no? You're an academic, you would be maximizing the number of people able to respond and help these communities. Well, maybe not — then rebel groups would also know which villages are reporting. There were many more of these ethical issues that came up. We felt horrible. We were possibly putting people at risk. We had to censure people, etc. We did many things in response, some of which I wrote about here.

For Voix des Kivus, you randomized the villages from which to crowdseed, but not the reporters from whom you received the information. Why not randomize whenever one can?

Always have in mind, "Randomize wherever you can." Yet in Voix des Kivus the goal was to obtain high-quality data from events in a village. In Congolese villages, different people have knowledge about different things. The chief is the go-to person for conflicts over land or the arrival of new internally displaced persons. The head of the women's association usually knows about things like sexual violence against women, which the chief wouldn't usually hear about. We wanted to learn about all of these events within the village, so we randomly selected the village, but not the reporters.

What’s the difference between crowdseeding and crowdsourcing data? Do you see the potential for crowdseeding in relatively developed countries?

Crowdsourcing is when you get data from an anonymous public. Crowdseeding entails selecting certain people to supply you with information. Crowdsourcing is good because you gets lots of data; however, it also has a lot of drawbacks in terms of the quality of the data: It’s an anonymous public, so people can provide you with wrong information that they send you strategically. Crowdseeding overcomes this problem since you build up a relationship with the reporters. Moreover, with crowdseeding you can randomly select the reporters, which makes your data representative.
Would this work in the developed world? In a study I conducted in the DRC, I showed that crowdseeding can be used to obtain high quality information from very isolated places, places that otherwise would not have gotten their information out to the wider world. We don't have those problems in the developed world. However, I definitely see the value of crowdseeding in the developed world — the idea of having a large database of people that you can quickly ask questions to.

How much are research agendas in the field of development affected by grants and donors?

Quite a bit. For example, together with colleagues from Wageningen University, I recently got a 1 million USD grant to study Ebola resilience and preparedness in Sierra Leone. I have been working on Sierra Leone for a while now, but I know little about Ebola. Now, I am learning a lot about Ebola and working with medical doctors, something I would not have done had I not received the grant. However, grants are increasingly informed by what research finds and where the holes in research and knowledge are. So it’s not all donor-imposed research agendas, but also very much the other way around.

How has community involvement in development interventions changed over time? Is community participation always effective or desirable?

Community involvement has a long history in development. However, particularly over the last decade, it has become "sexy" among NGO projects. People want projects no longer to be implemented from the top-down, but bottom-up. The World Bank has been spending billions in the last years on participatory projects. There are, at least from the outset, good reasons to do so: When people participate in the process, they may feel more ownership of the project; the project is more likely to reflect what people actually need.
Are they always effective or desirable? Probably not. People don't always know what is best for them. Others also argue that participatory programs can be very expensive compared to more top-down projects.
gazelle logo