Augustin Bergeron is a postdoctoral fellow at the King Center and starting in Fall 2022, he will begin teaching as an Assistant Professor in the University of Southern California's Department of Economics.
Bergeron's research focuses on topics spanning development economics, public economics, and political economy, where his primary research agenda explores the determinants of state capacity and tax capacity. He received a PhD in Political Economy and Government at Harvard in 2021.
What drives your interest in state capacity and tax capacity?
I've always had interdisciplinary interests. As an undergraduate student in France, I chose a multidisciplinary path, the Classes Préparatoires B/L, which allowed me to very thoroughly study mathematics, economics, sociology, history, philosophy, and literature for two years. My favorite subjects were mathematics, history, and philosophy. Funnily enough, at that time, the prospect of becoming a researcher in any of these disciplines terrified me (I wrongly assumed–based on my limited experience–that research was a very lonely enterprise), so after those two years of Classes Préparatoires B/L I chose to major in economics mostly because it opened many non-academic doors.
As an economics student at the Ecole Normale Supérieure in Paris, I became increasingly interested in development economics. I was especially interested in understanding why some countries are poor and others are rich. My favorite books at the time were The Wealth and Poverty of Nations by David Landes and Why Nations Fail by Daron Acemoglu and James Robinson. Thomas Piketty, who later became my master’s thesis advisor at Paris School of Economics, encouraged me to focus on the role of tax institutions in explaining development and income concentration differences across countries. He made me realize that research can be collaborative and tackle important and policy-relevant questions. I was hooked and decided to apply to PhD programs in the US.
As a PhD student at Harvard, I met other doctoral students—Pablo Balan, Gabriel Tourek, and Jonathan Weigel—who shared my interests in understanding the determinants of states' fiscal capacity and became my coauthors. I was also lucky to have a primary advisor, Nathan Nunn, who had worked in the Democratic Republic of the Congo (DRC). He pushed my coauthors and me to go to the DRC and inquire about possible partnerships with the government to evaluate some of its tax policies using randomized controlled trials (RCTs). I wasn't sure his plan would lead anywhere, but he was right: that's what my coauthors and I have been doing over the past five years!
Your work has been mostly focused on the Democratic Republic of the Congo. You received the National Tax Association’s 2021 Outstanding Dissertation Prize for your research using large-scale randomized policy experiments in the DRC to study what policy levers can increase tax revenue. Can you discuss your research findings and the impacts the work has had in the DRC and elsewhere?
My primary research agenda explores the determinants of state capacity and tax capacity in particular. Most of my work has focused on improving the policy tools that governments of low-income countries can use to raise tax revenue. This is of primary importance given that low-income countries face severe revenue constraints which are associated with low-quality public services and infrastructures. A surprisingly broad range of policy tools can be used to improve tax collection: changing tax rates, investing in enforcement capacity, centralizing or decentralizing tax collection, changing the assignment of tax collectors, etc.
To study how these policy levers affect tax collection, my coauthors and I evaluate RCTs conducted by the Provincial Government of Kasai-Central in the DRC. In such RCTs, a set of individuals (control group) is randomly assigned to the status quo tax system, while another set of individuals is assigned to a different tax policy. Assignment to the control vs. treatment group is random, so we can be confident that any differences in tax revenues between the two groups are due only to the tax policy we are evaluating.
To give you a concrete example, in 2018, we partnered with the Provincial Government to evaluate the effect of empowering local elites (city chiefs) to collect taxes. Households in control group neighborhoods were assigned to tax collection by state tax collectors, and households in treatment group neighborhoods were assigned to tax collection by city chiefs. We find that the city chief collection raised tax revenue by 44 percent. We use an additional RCT to document mechanisms and show that chief collectors achieved higher compliance by using local information to more efficiently target households with high payment propensities rather than by being more effective at persuading households to pay conditional on having visited them.
In other projects, we use similar RCTs to study the revenue-maximizing tax rate and how it is affected by investments in enforcement capacity or how the assignment of tax collectors to teams and postings affects tax revenue. Importantly, many of these projects have led to policy changes. The Provincial Government of Kasai-Central has been eager to learn from these policy evaluations and to reform its tax system accordingly. Governments in Africa and the Middle East facing similar difficulties in collecting taxes have started reaching out to us, potentially leading to exciting future research projects and collaborations on taxation.
Were you expecting to see those results from your RCT on tax collection by local elites? Or were you surprised by your findings?
We expected chief collectors to outperform central collectors (as described in our pre-analysis plan). However, we were unsure which mechanisms would lead chiefs to perform better than central collectors. Chiefs might conduct more tax visits (e.g., because they live in the neighborhood), they could be better at targeting households with a higher ability to pay (i.e., because of private information they have about each household's financial situation), or they might be better at convincing households to pay (e.g., because they are more trusted).
To test for mechanisms, we collected a wide range of survey data and designed an additional treatment arm in which state agents collected taxes after a half-day consultation with the local chief. The results from this treatment arm suggest that chiefs collect more taxes by leveraging their private information to target households with a higher ability to pay. This mechanism illustrates the value of the information possessed by local elites in tax collection and the potential returns to the state's attempts to codify and transmit local information to its tax collectors.
Can you elaborate on the process of working with public officials in the DRC?
Starting a research project in the DRC is a slow process. It typically involves several months of intense discussions with authorities and public officials about the tax policy issues they face and possible solutions. My coauthors and I typically enter these discussions as active listeners: our goal is to support the government in evaluating the policies they want to implement. Logistically speaking, there is always a detailed pilot and evaluation phase, followed by the implementation of the RCT at scale by the government. At the end of a multi-year RCT, we submit a lengthy report on our findings to the Provincial Government and hold discussions with the Governor and Provincial Ministers about the implications of the results for their tax policy.
You wrote a paper in 2020 examining the effect of colonial Christian missions on social ties. How did you become interested in the origins of social ties? How do social ties affect development?
I became increasingly interested in development economics when I was a master's student at the Paris School of Economics. I was especially fascinated by how institutions and culture affect economic development. At that time, Nathan Nunn and Leonard Wantchekon had just published a paper showing that the negative impact of Africa's slave trade on development was partly mediated by lower levels of interpersonal trust. Alberto Alesina and Paola Giuliano were also doing fascinating work showing the effect of family ties on economic behaviors, labor market regulation, and political participation.
When I started going to the DRC as a PhD student, I wondered how to contribute to this research agenda. At that time, I was reading many books on the DRC. One of them was Congo: The Epic History of a People, written by David Van Reybrouck, a Belgian cultural historian. In one of the book's chapters, he argues that missionaries in the DRC actively sought to weaken family and ethnic ties to create a cross-cutting Christian identity and convert populations. I started thinking about whether individuals' identity (familial, ethnic, national, religious) can co-exist or crowd out one another and the effects of identity on economic behaviors.
My advisor, Nathan Nunn, encouraged me to study these questions by collecting new data in the DRC. I started collecting archival records on colonial Christian missions in the DRC. I then combined these data with surveys and lab-in-the-field experiments that I conducted with a sample of 1,000 Kananga inhabitants in 2019. By leveraging the distance of the respondents' village of origin to the nearest colonial mission station, I was able to show that missions are associated with a positive bias towards Christianity and Christians and weaker family and ethnic bias. These results suggest that these identities are malleable and can crowd out one another.
Maybe more surprisingly, the paper also shows that exposure to missions affects the composition of individuals' social networks, resulting in a higher proportion of co-religionists and a lower proportion of family members or coethnics. Finally, I use a lab-in-the-field job referral experiment to show that the identity change induced by exposure to missions also affects economic decisions, such as which member of their social network individuals recommend for a job opening. Exposure to Christian missions is associated with more recommendations of co-religionists and fewer recommendations of family members or coethnics.
What are some challenges you have encountered in your research, particularly when conducting surveys in low-income countries? What have been the highlights?
There are multiple challenges in monitoring and evaluating randomized controlled trials in the DRC. One of the main challenges is to find the respondents we already surveyed at baseline for follow-up surveys or activities. Finding these respondents is difficult as they might not own a phone or it might be out of battery since very few households have access to electricity. Moreover, the city of Kananga is hilly, hot, and the roads are bad, complicating every trip to find a respondent. Thankfully, I have been lucky to work with a fantastic and growing team of enumerators. They are local Congolese who work with us to design and implement surveys. We have developed a great relationship over the past five years, and my research in Congo would not be possible without their ideas, perspective, and hard work.
You have been at the King Center as a postdoctoral scholar for a few months now. What has your experience been like at Stanford and the King Center?
I am very grateful for my time at the King Center. Pascaline Dupas and Kate Casey have been amazing mentors. They gave me fantastic advice on my research when I presented it at seminars, during one-on-one meetings, and after reading my paper drafts. I also enjoyed being part of the development group. I always look forward to the development tea and seminar or fun events at Pascaline’s house. I am also very thankful that so many of the faculty in that group – Alessandra Voena, Melanie Morten, Arun Chandrasekhar, and Marcel Fafchamps in particular – took the time to meet with me one-on-one, engaged with my work, and gave me great feedback. I’ll be sad to leave, but I am also very excited to move to LA and start as an Assistant Professor in the Department of Economics at the University of Southern California!