Oriana Bandiera is the Sir Anthony Atkinson Professor of Economics at the London School of Economics. Much of her research has been on organizations and labor markets in both developing and developed countries, as well as on their intersection with gender. She is currently a co-editor at Econometrica, and directs the Gender, Growth and Labour Markets in Low-Income Countries (G²LM|LIC) programme at IZA.
1. You gave a 2018 RES Public Lecture on What Economists Really do, as part of an effort to help draw a distinction between popular perceptions of what economics is about and the breadth of issues that economists actually work on. What did you think economics was about when entering the field, and what got you into development?
To be honest, I had no idea what economics was until I stumbled into it almost by chance. I went to a high school with a strong focus on maths and my favourite subject was philosophy but everybody expected me to study law and join my dad’s firm. However, I had no inclination for law and a very strong desire to see the world so I had to find something so prestigious that it would justify deviating from the beaten path. Fortunately, I was admitted into the business administration course at Bocconi, in Milan. Unfortunately, my talent for business studies was worse than for law. As I was getting ready to go back home, a friend enrolled in tiny degree programme called “Political Economy” convinced me to attend one lecture. I switched degrees that evening and never looked back.
Coming from a very poor region of a fairly wealthy country, I cared about development even before I knew that it was a field of study. Then Nick Stern gave a guest lecture in my final year. He was coming from Palanpur, the Indian village he and Chris Bliss had been studying for decades. I remember two things distinctively. The first was that my grandma’s village in Sicily was much closer to Palanpur than to Milan. The second was that I could potentially have a paid job to study things that I actually cared about. I could not —and still cannot- believe my luck!
2. A trade-off researchers often face when working on evaluating the impact of some policy is how much to evaluate it with the best possible implementor and very close supervision to test some theory or perhaps an efficacy trial, versus working with an organization such as a government and having them implement a program under business as normal to see what the likely effectiveness might be if implemented as a policy. How have you thought about this in your own work, especially given your research on organizations?
This is a very important question. The answer depends on what it is that we are trying to establish. If we want a proof of concept -e.g. does vocational training improve job market prospects? - or evidence on a specific theory -e.g. are the poor stuck in a trap?— then it is essential to implement as well as possible otherwise we will never be able to tell null effects from implementation failures.
If we want to know whether a specific intervention can reduce poverty on a large scale then we need to test its scalability. And it might as well be that A is more effective than B under perfect implementation but B is easier to scale-up.
3. A related question is what you do you see as the most important determinant of the success of the different programs you have evaluated with BRAC in Uganda. Your experiments have found relatively large and lasting impacts of vocational training and ultra-poor programs. What lessons do you draw from the way BRAC implements programs that you think may differ from other organizations or offer lessons for governments attempting to implement such policies?
I think that the main challenge we face as researchers and policy makers is that we aim to solve problems faced by others in contexts of which we have limited, if any, experience. The hardest thing is to keep an open mind and avoid interpreting what we see in the data through the lens of our own experience in different contexts. BRAC do two things very well. They try to get to know the people and the context as well as they possibly can and they use evidence to find out what does not work rather than prove that it works. Evaluation is built in as part of implementation rather than done ex-post. But BRAC are sufficiently large and successful to set aside funds for evaluation. Unfortunately it is not common practice among donors to fund evaluations as part of program implementation, which makes it difficult for other NGOs to follow the BRAC model. I think it would make a great difference if donors were to see evaluation as part of implementation rather than a separate activity. This leads to better research and enables organisations to learn as they implement the program rather and can still implement changes.
4. Your webpage lists over 30 PhD students you have advised, most of whom are development economists. Given the great cluster of development faculty at LSE (as well as at nearby UCL), do you team advise, or tend to have a primary advisor for each student? What is your approach to advising PhD students in development, and what is the typical advice you give for those trying to find a topic to work on?
It is definitely a group effort! I find supervising PhDs quite difficult because what makes a good paper is the question it asks, and we are trained to teach how to answer questions, not how to ask them. My advice is not to be “strategic”, that is not to give in to the temptation to ask questions because they are on hot topics, or because you have data that nobody else does or whatever reason other than you find the question really really interesting.
Writing a paper takes a long time and the job market paper will stay with you for years, so if you choose a question you are not interested in at the start, you’ll find it impossible to work until you find the answer
5. As part of the IZA G²LM|LIC program that you direct, I see that you are teaching an online course with Robin Burgess on Development Economics, for students who live in South Asia (with applications due November 10). Can you tell us a bit about the impetus and goals for this course. Do you view the key development economics questions as being different in South Asia than in say Sub-Saharan Africa or Latin America, or is it more a question of time zones and which examples to use, but the basic curriculum being similar to what you would teach if focused on another region or for your general course at LSE?
Actually after you asked me I realised that the title could be misinterpreted and changed it to Development Economics, specifying later that it is for students who live in South Asia. I started teaching online courses during the pandemic. G2LM|LIC had agreed to organise capacity building courses in Africa and South Asia but obviously we could not travel out of the UK so I put together a “research skills” course to make up for it. The course taught how to prepare a presentation or a funding application, how to work with governments, how to run field experiments and how to publish. Besides myself, I managed to convince three leaders in each of the areas - Adnan Khan, Munshi Sulaiman and Imran Rasul. Demand was higher than expected and we couldn’t run it to cover all the time zones, so we ran two: one for Africa, one for South Asia. We got positive feedback and several requests to run a course that covered topics in development so we did, and for the same reasons we split the two audiences. Topics are the same —there is nothing specific to South Asia. The course addresses the question of why fundamentally equal humans living in different countries enjoy very different standards of living. The aim is to make the very latest research in economics on these issues accessible to anyone who is sufficiently curious about it. The course will run for 10 weeks between January and mid March. Interested readers who live in South Asia can apply here until November 10. Join us!
6. You have written multiple papers looking at the impacts of different ways of selecting workers and of different payment schemes. One thing that always surprises me when seeing studies with factories in developing countries is how high labor turnover can be. Donovan et al. find that labor market churn is much higher in developing countries than developed countries, but does not seem to reflect faster reallocation into more productive or growing sectors. Based on your different experiments on recruitment and payment schemes, do you have a view on what sorts of reward, recruitment, or payment schemes large firms in developing countries should be trying to do more of?
I have been wondering about how people select and are selected into jobs for a very long time , I suspect this is because I narrowly escaped becoming a lawyer, which would have been much worse for me and especially for my unsuspecting clients. Economists focus a lot on incentives to improve the productivity of the same individual doing the same job. But there is only so much one person can improve —20% is a good heuristic— while differences in productivity across people are orders of magnitude larger. In fact, every HR person I have ever spoken to, from the Ministry of Health in Zambia to the CHRO of a large multinational, says that recruiting and retaining the right talent is the most important determinant of productivity. I think differences between richer and poorer countries are due to differences both on the workers and on the firms’ side. In poorer countries agriculture still employs a large share of the population, and when demand for labor in agriculture peaks, workers go back to the fields. There is not much firms can do about this other than matching the peak wage, but it might not be worth it. Indeed, firms in high income countries spend quite a lot to reduce turnover because workers are highly specialised and the onboarding process very costly; to the extent that specialisation is lower in poorer countries (which we know to be true), it might simply make sense to bear the consequences of turnover rather than the costs to reduce it.
Here are our previous Six Questions with interview series:
· Six questions with Chris Udry
· Six questions with Rohini Pande
· Six questions with Mark Rosenzweig
· Six questions with Martin Ravallion
· Six questions with Andrew Foster
Join the Conversation