Skip to content
  1. Latest From the Blog

Q&A with CHIBE Post-Doctoral Researcher: Ellen Moscoe, ScD, MA

Ellen Moscoe headshot

Ellen Moscoe, PhD, MA, is a post-doctoral researcher whose research lies at the intersection of global health and behavioral economics. She focuses on poor populations in sub-Saharan Africa as well as marginalized groups in the United States.

Q: What kind of quasi-experimental methods do you use to measure the causal effects of policies on health behaviors and mental health?

When we are interested in measuring real-world impact, we are usually working with policies that have been implemented and data that is already being collected, either administrative data or data from national surveys, for example. So the main thing your methods are trying to do is find a good comparison group: which group of people provides a good comparison that will show you what would have happened if those who got some intervention or were affected by a policy had not been? Difference-in-differences studies do this by comparing the evolution in an outcome in two (or more) groups over time, and comparing the changes before and after the policy. It’s a great class of methods to use in the U.S. where states often adopt policies at different times. Another method that has gotten quite popular lately is called regression discontinuity design. These studies find a comparison group by looking at policies or treatments that are given based on some kind of threshold rule, and then comparing groups close to the threshold to see how outcomes differ for those who got the treatment compared to those who just barely missed out. Regression discontinuity is powerful because it can be applied to a wide variety of settings, from medical treatments given when a test result is above/below some cutoff value, to access to social services where eligibility is based on one’s birth date or income level.

Q: What global health issue would you like more attention/research on?

This is a hard question because I have plenty of answers. I attended a health economics conference in July and was surprised and excited to see many sessions focused on mental health. There is increasing attention to this as an important health issue but I think it’s still not given the weight it deserves in low-income country contexts. The evidence linking poverty to stress has really big implications if we believe that constant stress, especially in early life, can both impact decision-making and put people at risk for chronic illnesses. I’d love to see, and be involved in, more work in this area! If I get to pick a second thing, I would say we need a better way of tracking the results of randomized trials in the social sciences so that we know what did not work. Asking journals to publish null results might be possible, but it could be more efficient to require publication of a short summary of findings linked to the trial registry and make those searchable, so that it would be easy to, for example, find all the trials of financial incentives for exercise and quickly assess what we know from them. If I were a billionaire looking to invest in a vanity philanthropy project, that’s probably what I would spend my money on.

Q: What do you find most rewarding about your work?

On a day-to-day basis, I find it really rewarding to work with very talented and smart colleagues. My favorite collaborations have been projects that I’ve worked on with colleagues who are first and foremost friends, because it makes is really enjoyable to learn from each other. Taking a step back, I think my main motivation in this job comes from the potential to make a measurable impact with my work. Of course most studies will make modest contributions, but the kind of work that CHIBE does is so closely tied to policy that once in a while something will really stand out and have a big effect on well-being. I hope that I have one of those over the course of my career. And work that doesn’t translate directly into policy change can contribute to an evidence base that will chip away at a complex problem over time. And even studies that completely fail will still have taught us something about why they fell apart, so we can do better the next time.

Research Areas