Evaluating the tools used for inequality research with grant recipient Liyang Sun

We caught up with Stone Centre grant recipient Liyang Sun to hear more about her work evaluating one of the most widely used empirical tools in inequality research.

‍

Hi Liyang, thanks for joining us. Please tell us a bit about your academic background.

‍

I am a Lecturer in the Department of Economics at University College London. I completed my Ph.D in Economics and Statistics from MIT in 2021 and did my post-doc at the University of California, Berkeley in 2022.  I completed my undergraduate studies in Economics and Mathematics at Wellesley College in 2014.

‍

What is your research about?

‍

I am an econometrician with research interests in causal inference. I develop new methods for causal inference in a more realistic setting of treatment effect heterogeneity. Recently I received a UKRI Economic and Social Research Council grant to develop tools that allow researchers and policymakers to both estimate heterogeneous causal effects and optimally act upon them.

‍

I also contribute research on weak identification/model misspecification. For example, in empirical research on inequality, to study the impact of education on earnings, economists know that even after controlling for many observed covariates, education decisions may also be influenced by unobserved factors that affect earnings. To isolate education’s causal effect, economists use "instrumental variables", which are external factors that influence education level but not earnings directly. Weak identification arises in this example when these instruments are weak (barely shifting education levels). Model misspecification arises in this example if the impact of education is heterogenous across individuals. The Stone Centre internal grant I recently received supports this line of research.

‍

I have worked on improving the empirical implementation of synthetic control methods, particularly in high-dimensional and high-frequency settings that fall outside the classical framework.

‍

Why did you decide to study this topic?

‍

Instrumental variable (IV) analysis is a cornerstone of empirical research on inequality. Prominent examples include using geographic proximity to colleges as an instrument for educational attainment to estimate returns to schooling, and employing trade exposure constructed from initial industry composition as an instrument to assess regional wage and employment inequality. These designs are powerful because they exploit quasi-experimental variation to address endogeneity, but they can also fail when instruments violate exclusion restrictions, treatment effects are heterogeneous, or instruments are weak. In these cases, estimates may be biased or misleading. I decided to study this topic because it is both theoretically important and highly consequential in practice: improving the reliability of IV analysis can strengthen the empirical evidence used to understand inequality and inform policy.

‍

How does your research contribute to the Stone Centre’s mission of developing our knowledge of inequality?

‍

The Stone Centre aims to advance rigorous research on the causes and consequences of inequality. My project contributes directly to this mission by evaluating one of the most widely used empirical tools in inequality research: instrumental variable analysis. By systematically assessing how violations of key IV assumptions affect estimates in canonical studies of education, trade, and other drivers of inequality, the project will improve methodological transparency and help researchers better distinguish credible causal findings from results driven by misspecification.

‍

How will the Stone Centre grant help you?

‍

The Stone Centre grant will support collaboration with a PhD research assistant to conduct a structured survey of inequality studies using IV methods. Building on this survey, we will replicate benchmark studies and develop calibrated Monte Carlo simulations to evaluate the performance of IV estimators and inference methods. The Stone Centre grant creates a unique opportunity for me to focus on methodological research while grounding it in the realities of empirical work and the practical challenges faced by applied researchers.

‍

What will you produce as part of your research?

‍

As part of the survey, we will generate open-source simulation code and teaching materials that can be used by researchers and students to better understand IV methods. I am also preparing a practical guide for applied researchers on IV analysis under heterogenous treatment effects based on the survey. The collaboration with PhD students will contribute to joint research on new IV methods.  

‍

Useful links:

Liyang Sun's Google Scholar page

‍