Research statement

I am a Macro Development economist interested in what drives growth processes in developing countries. Methodologically, I combine rich micro data to uncover fundamental mechanisms of economy-wide changes with structural models that allow to think about and quantify aggregate effects. Throughout, my research focus is on longer-run changes (processes that play out over decades).

Current work

My current work can be structured into two primary areas of interest: first, better understanding the drivers of growth and structural change. And second, the political economy of development. I discuss each in turn and then end with my research agenda for the next years.

The drivers of long-run growth and structural change: Economic development is characterized by dramatic structural changes in the economy. For example, the majority of people in the poorest countries in the world work in agriculture, household subsistence and self-employment. However, as countries grow richer workers move (1) into manufacturing and services (Herrendorf et al 2014), (2) into market production (Goldin 1995) and (3) into salaried wage work (Gollin 2008, Jensen 2022). At the same time, we know that many of these processes take time. For example, firms take time to hire or lay-off workers, and workers only slowly reallocate jobs across occupations, sectors and firms. A natural question is thus how much of these processes are driven by current changes in the economic environment such as new government policies and how much is simply driven by economies still adjusting to previous changes in the economic environment.

In my Job Market Paper, which is joint work with Oscar Fentanes, a fellow PhD student at TSE, we answer this question quantitatively by looking at the drivers of economic growth in Indonesia. Building on a structural model and 40 years of establishment-level data, we show that Indonesian establishments at any point in time still heavily adjust to past changes in the economy. Importantly, adjustment processes do not become less important over time, because establishments are not catching up fast enough to keep up with new changes in the economic environment. As we show, failing to take this into account, leads to misattributing economic growth to current policy changes. Policies also differ in whether they create new potential for growth or lead to faster catch-up growth.

While my JMP focuses on the supply side of the dynamics of development, in my ongoing work on Income Inequality and the Dynamics of Structural Change, I look more closely at the demand side. I start by establishing two empirical facts. First, by drawing on extensive Indian household-level consumption data, I show that changes in India’s income distribution drive changes in sectoral and product-level demand that can in turn explain economy-wide structural change out of agriculture and into manufacturing and services. The reason is that as households become richer, their demand changes due to non-homothetic preferences. Second, I link these demand changes to productivity growth. The idea is that if household income increases, demand increases differentially for products that are produced with similar technologies but face different income elasticities of demand. Preliminary results surprisingly show that increases in demand do not systematically drive productivity growth, in contrast to standard growth theories. I am currently exploring the second result at different levels of aggregation and then plan to incorporate both facts into a quantitative model of demand-induced structural change and growth. I have already applied to a STEG CEPR grant that would finance the continuation of this project.

The Political Economy of Development: A major determinant of development is the interaction between political and economic institutions. I study this question from two perspectives.

On the one hand, I study how economic and political elites may collude for their own benefit, potentially undermining the workings of the economy. In my paper on the Aggregate Costs of Political Connections, which is currently under submission, I draw on Indonesian micro-data, a natural experiment that identifies connected firms, and a structural model of endogenous rent-seeking to quantify the welfare costs associated with giving favors to a few politically connected firms. I find that the few connected firms — only 1% of firms in the Indonesian data — cost the economy up to 5% of aggregate output and that the majority of the costs come from the costs of subsidizing connected firms rather than the costs of misallocating resources, including resources spent on wasteful rent-seeking.

On the other hand, I study how the government interacts with the private sector over the course of development: for example, via the provision of public goods (work on Political Connections and on Decentralization), or by affecting who works for the private and public sector over time (work on Selection of Government Workers). For example, in my work on the selection of government workers, I show how to measure and compare government worker skills if government wages are uninformative about skills, but researchers have access to skill-related observables. I then use this method in an application to detailed Indonesian worker-level data to evaluate whether public sector hiring reforms led to the better selection of government workers over time.

Research Agenda

In my future work, I aim to build on my current research and develop it further in three main areas: The study of (1) the dynamic evolution of markets, (2) the drivers of growth, and (3) Macro Political Economy. I discuss each in turn. I expect to apply for (additional) research grants that would cover the extensive data efforts involved in some of the research plans. I also expect future PhD students that I advise to actively participate in this research agenda.

Dynamic market evolution: A common practice at the current research frontier across fields such as Macroeconomics, Development and Industrial Organization is to abstract from important changes in markets over time. Technically, this shows up as steady state or balanced growth path assumptions that are made for tractability reasons. Substantively, this implies that many important economy-wide adjustment processes remain understudied. For example, what are the welfare costs and optimal policy responses to slow gentrification processes that drive poorer households out of expensive neighborhoods? How do big firms and national leaders emerge within industries over the course of development and how can policy influence this process? And how do economies diversify into increasingly more products over the course of development?

Besides missing important processes, there is also increasing evidence — including my own Job Market Paper — that economies are poorly described by quickly transitioning between different steady states as policy and technologies change. This is because aggregate adjustment processes can take long since workers reallocate slowly across locations, jobs and occupations, and firms enter, exit, and grow slowly. If changes in policy and technologies are frequent and sizeable, then economies never stop adjusting and are always far away from their steady state. In my Job Market Paper I provide tools — particularly estimation tools — to tractably study such settings. Applying and extending these tools to different markets and questions is an exciting area for future research that I aim to push forward. For example, I am currently actively investigating research projects on:

gentrification dynamics and rent dispersion (joint with Tuuli Vanhapelto at Glasgow & Elena Lutz at ETH Zürich) building on Swiss apartment-household matched administrative data and the universe of rents
the evolution of market concentration over the course of development (joint with Oscar Fentanes), drawing on 20 years of Tortilla shops in Mexico, a large industry with many local markets producing a homogenous good
the evolution of the product space over the course of development building on comparable manufacturing censuses across countries and within-firm expansions in their product scope

All three projects are currently still preliminary, but in all cases, I have already secured the necessary data access and I am the co-PI on a separate CEPR STEG grant application that would cover work on the evolution of market concentration project.

The drivers of growth: We have a large variety of theories on what drives economic growth in developing countries including among others: technology diffusion, learning- by-doing, R&D, human capital externalities and knowledge spillovers. Being able to better distinguish between these drivers and quantifying their relative importance is key to designing effective policies. I see two main ways in which my research can contribute here. First, as I show in my JMP, there are ways to non-parametrically identify sources of economic growth in spirit of “sufficient statistics” approaches that are in line with larger classes of growth models and which allow to quantify different drivers of economic growth without relying on any one specific model. Second, new sources of empirical variation can be better exploited to pin down drivers of economic growth. I am currently actively pursuing this in my work on Inequality and Structural Change where I show that household heterogeneity coupled with empirically observed non-homothetic product demand can be used to construct novel demand shifters that pin down demand drivers of productivity growth.

Macro Political Economy: Lastly, I aim to build on my previous work on the political economy of development. In my view, this literature leaves a number of important big picture questions unasked and unanswered. For example, how should the government select workers over the course of development? Is it always best to attract the most talented workers for government jobs or does this crowd out talent for the private sector? And how does this answer change quantitatively if government jobs are partly used for rent-seeking? These are questions that would benefit from more systematic stylized facts, formal models, and new data. The current availability of matched employer-employee data that often covers the universe of public sector workers could for example be exploited to document how relative skills of government workers systematically vary over the course of development, across and within countries.

Together, I believe these research projects can further our understanding of the development process, inform important policy decisions and make important contributions to the field of Macro Development.