MIT Master’s Program

by Shotaro Beppu

First and foremost, I express my deepest gratitude to Ito Foundation U.S.A and Friends of UTokyo, Inc. Without their support, I would not have been able to experience this surreal opportunity.

As of writing at the end of June 2023, I am conducting my summer research to fulfil my requirements for MIT’s Master of Applied Science in Data, Economics, and Design of Policy (DEDP). For background, this master’s program is one semester and a summer program. I arrived in Boston in January 2023 and will leave for my next endeavour at the end of August 2023. Why so short? Because the first semester’s worth of coursework is done via an online platform which I did piece by piece from 2020. The courses are available to anyone, and only those who completed the online segment can apply to the residential program. The uniqueness of this program to allow doing coursework while we were doing something full-time made it possible for me to meet people from diverse backgrounds with the common aspiration to use data and policies to understand better and help solve the pressing issues in the world. Surrounded by inspirational classmates, I thoroughly enjoyed the coursework and interactions with professors whose papers inspired me to take the path of development economics.

< Cohort >

My cohort of 21 was nothing but diverse. My classmates were working in the public sector, international organisations, NGOs, development aids, research teams in platform and consulting companies, financial sectors, academic institutions etc. We also came from 15 countries, though, since people worked in so many other countries due to the international element of development policy, I somehow found people either visited or lived in every country imaginable. We had very different work experiences. From a freshly graduated undergrad to those who had already worked a couple of years in the field, who already had a PhD in social sciences, who were higher-ranked public servants, it seemed onset that we were very different people. Yet, what I found was a welcoming environment to understand each other’s interests that was always related to the desire to change the world through evidence and policymaking. We took most of the classes together and studied together in the same room for hours. We enjoyed our weekly brunches with dishes from one of our countries, where a couple of us presented what we were doing. We did many social/physical activities on and around campus together. Contrary to what I thought when I arrived, I strongly believe that the biggest benefit I got from this program is the relationships formed and the enjoyment I got from interacting with them. Development economics encompasses not just researchers but practitioners in many sectors. I am extremely fortunate to experience the potential and the vastness of this field in the early stages of my life.

< Coursework >

One of the things I initially looked forward to the most was the classes. As an engineering major focusing on data science with a strong interest in social sciences, I was in an ideal place where economics and data science are conducted with one of the highest standards. And I was never disappointed. The classes I took are below.

  • IAP (an MIT thing in January where you are free of full coursework to explore)
    • 6.S091 Causality and Machine Learning (half course)
    • 6.S191 Introduction to Deep Learning (half course)
  • Spring
    • 14.582 International Economics
    • 14.760 Firms, Markets, Trade, and Growth
    • 14.126 Game Theory
    • 14.320 Econometrics
    • 14.382 Econometrics
    • 14.123/124 Microeconomic Theory III/IV
    • 14.399 Master’s program seminar

Development economics and policy evaluation deal with the question of what-ifs. For example, what would happen to student performances if students received extra classes? A simple comparison between students who did the extra classes and students who didn’t is not enough to see the impact. Why? Probably because those who took the extra classes are more motivated compared to students who didn’t and would have a higher performance anyway. This question about causality is crucial to answering how certain policies or events led to changes. 6.S091, 14.320, 14.382 taught me the tools to help identify this causality. This is most certainly important in our world as we have more data for bringing “evidence-based policymaking.” I also took 6.S191 for the increasing usage of deep learning in our society in terms of data analysis and changes brought forth to the future of policies.

Then what kind of policy is useful? For this, we need to understand how the individual as well as the society works. For this, 14.123/124 and 14.126 are helpful in modelling people’s interactions. As I am interested in trade and urbanisation (essentially, how can trade and cities be helpful in the development of societies?), 14.76 and 14.582 gave me a deeper insight into these areas as well as the tools to understand more. A topic I took a particular interest in is how people gather in cities and how this affects economic development. From the birth of civilisation, we gathered as a group to create innovations. Quantifying to what degree a group of people formed in cities multiples each person’s ability and what sort of policies can best harness this power is arguably crucial in this era as cities such as Boston and Tokyo continue to be the hub of innovation while cities such as those in Japan’s rural areas are diminishing. As climate change changes the geography of this world, where cities locate and are connected, as well as how to make cities more sustainable, are of first-order importance. I will continue working on this topic after September at UChicago Urban Labs.

Lastly, I participated in several weekly seminars in the economics department, including 14.399, which was exclusively for our cohort. This helped me know more about what researchers in economics are trying to answer. One of the unexpected benefits was listening to questions from world-renowned professors, such as a couple of Nobel prize winners. Sometimes they asked questions I also had, which made me a little happy, and sometimes they gave crucial comments that could change the paper drastically. However, what made a lasting impression was how they truly wanted to know the background and the setting of this study. This impression inspires me to know better what I am doing research about and why I am using the tools I am using to investigate.

< Research >

My main research is a continuation of my graduation thesis for UTokyo but with more emphasis on economics. The gist of my undergrad thesis is using graph neural networks to digitise historical shipping documents and investigate the impact of steamships on trade and development in the 19th century. Currently, I am using historical census data in the US to see whether the changes induced by the transition from sailing ships to steamships led to different city growths and, thus, a long-term change in the economy of the US. This research fully uses what I learnt during the spring semester, and I believe it gives important insights into not only how technology can change the location of production but how the location of cities can affect long-term growth. I plan to produce my first draft by the beginning of August.

In addition to my own research, I am fortunate to participate in a couple of other research projects. Two of them revolve around using the world’s grid-level data (about 10 x 10 km). One is using agriculture productivity and actual production data to see possible misallocations in agriculture production. The question my professor and I are asking is, are places producing the crop that is most efficient for the world? This combines insights into international trade and environmental studies. The findings will further understand how trade can help in creating more sustainable production at a global level. Another is using population and income data to put a number on how much we see a productivity increase if we bunch people together (i.e., cities). To answer this, we are implementing a novel identification method in econometrics. As this is in the initial stage of research now, it is also interesting in terms of how one of the very top researchers whose papers inspired me to do research in this field in the first place formulates questions and how to tackle those.

< Concluding Remarks >

Staying somewhere I knew no one for the first time and studying in quite a rigorous environment was not easy for me. My days during the semester were more often than not filled with struggling over the same things over and over again and doing problem sets. Too many times, I procrastinated, and my research stagnated. Too often, I felt dumb and asked stupid questions to people I never ever wanted to. However, unbelievably, I never felt more conviction in what I was doing. I learned the most, cooked the most, exercised spontaneously the most, went out the most, and enjoyed the most in my life. How could this be? It was my cohorts who studied with me and told me they were doing exercises. It was my PhD roommate finding time to cook whilst doing experiments. It was the professors and the TAs who were clear in their teachings and their thoughtful answers to my questions. It was the seminars and the normal conversations with people which gave me inspiring ideas. I am extremely fortunate to be in this environment. As I have two more months to continue my struggle here, I will try to make the best out of this place. Once again, I am thankful to Ito Foundation U.S.A and Friends of UTokyo, Inc. for this experience.

Informal graduation ceremony