What keeps me up writing code

I’m interested in understanding economic behaviour through rigorous measurement, applied casual analysis, and grounded testable theory that disciplines what we’re doing. My interest sits at the intersection of survey methods, and labor markets.

A central theme in applied micro over the last decade has been the development of effective identification strategies. However, one key linchpin in this development has been the assumption of “good” data, as we know that solid identification is essential—but it doesn’t rescue you from bad data. Dillon et al. (2021) make this point explicitly: progress depends on pairing strong designs with serious attention to how the data are generated. (ScienceDirect)

That’s basically how I think about measurement. The data-generating process, both in terms of modeling the estimand we are interested in and how the data is produced, matters. I am interested in understanding the implementation details and modeling their impacts on measurement. Take, for example, mode of a survey; we know from many years of survey methodology research that respondents react differently in person versus online, partly due to certain biases. This fact is only exacerbated by selection effects of who tends to answer each of these modes in comparison to the other. However through careful study and intentional design choices, such as thinking about benchmark variables or other ways of correcting the biases we can attempt to correct for these descripencies.

I am currently working on a project with which attempts to qauntify the measurement error and impact on treatment effects introduced by survey design choices such as how to order respondents in follow up questions. I am working with IPA, JPAL and the World Bank to develop a database of surveys to begin to conduct a meta analysis of these design choices and their impacts on treatment effects.

Labor markets and compensation

My interest in labor markets stems from my time in Nairobi, Kenya. Nairobi has an extremely interesting labor markets and firm behaviors. The majority of labor is informal and normally self-employment, furthermore these jobs tend to be extremely similar Labor markets are where workers, firms, and institutions bargain under frictions. I’m especially interested in compensation design—how pay structures shape worker behavior and who sorts into which jobs. Personnel economics is a useful backbone here because it treats pay schemes as an equilibrium object: incentives matter, but selection and sorting matter too. Lazear’s classic evidence from a shift to piece rates shows both channels in a way that’s empirically clean and conceptually tight. (American Economic Association)

Substantively, I’m interested in:

• how firms split compensation between cash and non-cash margins,
• how wage-setting interacts with frictions (search, bargaining, mobility),
• and how we measure job quality in a way that goes beyond “wage = welfare.”

Bringing it together

The way I see it: measurement is upstream of everything. Better measurement makes labor market inference sharper, and serious labor market work forces you to model firms rather than treating them as a black box. My goal is to do work that’s empirically credible because it’s grounded in how the data were produced, and theoretically grounded because the theory clarifies what should move, for whom, and why.