People

How COVID Changed the Research Landscape: An Interview with Ayesha Mahmud

By Amanda Glazer

February 1, 2022

This article is part of STEMinism in the Spotlight, a monthly interview series.

Ayesha Mahmud is an assistant professor in the demography department at UC Berkeley. She earned her Bachelor’s degree in Economics and Physics from Carleton College, her PhD in Demography and Social Policy from Princeton University and was a Rockefeller Foundation Planetary Health Fellow at Harvard University. She has conducted research in a number of different fields including astronomy, health care, and demography (the study of populations). Much of her current research is COVID-related. In general, she seeks to answer questions such as “why do you see outbreaks at certain times of the year?” In this interview, we discuss her career path, demography research, how COVID has changed the research landscape, and her advice for students. She tells me that she has “meander[ed] a bit but that’s okay. I always tell my students that it’s okay to try out things. You’ll eventually find what you like.”

Amanda Glazer (AG): How did you first get interested in STEM and how did you end up in Demography?

Ayesha Mahmud (AM): I was always interested in science in school. The science departments at Carleton, where I did my undergrad, are really strong and so those classes are really fantastic. I took some social science, math, computer science and physics classes, and I liked science the best. That’s sort of the main reason. I had some great professors. I still keep in touch with them after so many years. I also went to a small liberal arts school which is nice because there’s no grad students, so we were able to get involved in research with the faculty. That part was really a major motivator for a lot of people who ended up doing science at Carleton. And then I did a couple of internships during college which got me interested in science. I was really interested in astronomy and planetary sciences. The internships are also how I decided that I didn’t want to do science research. They were both research focused and were designed to give undergrads a taste of what grad school would be like.

AG: Where were the internships?

AM: I did one at the Space Telescope Science Institute, which at the time was doing all of the Hubble work. Essentially I worked with some folks analyzing images from Hubble and studying star formations. Things that I could talk about for a long time. It was really fun. Then the next summer I worked at the Lunar and Planetary Institute, which is based right outside the Johnson Space Center. It essentially was set up when the first moon missions happened and there were samples brought back and geologists and planetary scientists were analyzing all of those samples. They now do a whole variety of different sorts of planetary science research. I was studying the atmosphere on Neptune and the chemical composition on Neptune’s atmosphere using Spitzer telescope data. So that’s very far from what I do now, but it’s interesting. I loved it, and then I didn’t want to do it anymore.

AG: What drew you to it initially? Through classes?

AM: Yeah, I had a really great astronomy professor at Carleton. I took a bunch of astronomy classes and then when I got those internships, each step built on itself. I also was doing Econ as a major because I was kind of interested in social science, and I also had great professors and a great mentor. I enjoyed the classes a lot. I sort of made a last minute decision to major in Econ, because it was a little bit difficult to double major at Carleton because you had to write two senior theses. In terms of timing it’s just hard to do. But I decided that I liked research enough to spend my time doing that.

Then after I finished at Carleton, I first worked for NORC, which is a large survey research organization based out of the University of Chicago but sort of independent and has offices in many places in the US. I did a lot of health care research. A lot of things I had studied a little bit in college in health economics and other classes like that. I looked at things like access to care in rural areas, Medicare / Medicaid beneficiaries and their experiences and other things like that. Then after that I worked for a health economist for a year. I was interested in health.

When I applied to grad school, I applied to all sorts of programs: a couple of econ programs, one demography and one health policy program. When I visited Princeton Demography that seemed like the best fit. Demography is nice in that the theory of demography draws from many other fields, so you don’t have to be as tied to the pedagogy of one particular field. Whereas if I’d done Econ, it’s a very specific type of training and very specific path. It’s a great path but it just wasn’t for me. With demography you can kind of choose your path because it’s more interdisciplinary by nature. My PhD advisors were both ecologists which is how I got into disease ecology. I spent my time half in the ecology department and half in the demography department, which is how I got into the type of work I do now. So, that’s the whole long, winded path to go from Physics and Econ to demography.

AG: That’s awesome, and it sounds like you have a pretty broad range of interests.

AM: Yeah which has made me meander a bit but that’s okay. I always tell my students that it’s okay to try out things. You’ll eventually find what you like.

AG: Or a few things you like! How would you describe the field of demography, in general?

AM: I think the field is very broad in terms of the types of topics that we cover: anything having to do with population processes or things that affect humans are fair game for demographers to study. We are interested in three main things: mortality, fertility and migration. That’s how we define the broad umbrella of demography, but almost anything can fit within that. Any process that affects human health or has consequences for mortality, fertility or migration. There’s a core set of formal demographic methods that we use to understand population processes, which I think are unique to demography. It’s a way to understand how mortality and fertility rates affect population growth or change over time. We have some core methods that we use for studying populations and ways in which we think about exposure to risk and modeling risk. But the topics we study are very broad. Demographers study almost anything that social scientists or public health folks would.

AG: How did you find the Demography program at Princeton?

AM: That’s a really good question. I was looking at programs that had a health economics focus and the Princeton program had a population health track. They’ve now reorganized the program so it’s not so segmented, with the tracks, but they do have a focus on health. It’s part of the office of population research and also the Wilson school. Both of those groups have a very strong history in doing health research, so that’s why I was interested. I had encountered demographers in my work, and I knew a few people from my previous job, after college, who had gone to the Princeton program. So, I knew a little bit about the program through them as well.

AG: Did you always know you wanted to go to grad school or did you consider doing something else?

AM: I always wanted to go to grad school, because I liked doing research and being in school. I don’t think I was always sure I wanted to stay in academia. That’s a relatively new thing. Even while I was doing my PhD I was exploring other options. I did an internship during my PhD partly to try and see if there were other things I might be interested in doing. I didn’t go into my PhD saying I want to be a faculty, but I thought maybe I want to be a faculty. I was definitely more interested in non-academic jobs, but I still wanted to do the PhD because I thought that there was value in doing it even if I didn’t want a faculty job.

AG: And then how did you decide you wanted a faculty job?

AM: Well I think the fact that I got a faculty job kind of made that decision in some ways. I applied for the Berkeley faculty position when I had just finished my PhD. I had finished my PhD in the summer and I had started a post doc in the Fall when the posting happened. Demography is a tiny, tiny world. There’s not that many demography departments. A lot of demographers are based at various centers that are affiliated with other departments. So very few schools have a stand-alone, PhD granting department. Berkeley is one of them, and Princeton and Penn are the others. So, you only get a demography job opening every, I don’t know, 5 years or something. I had a good friend in the demography department at Berkeley who forwarded me the job posting and told me I should apply. I wasn’t sure because I had just started my two year postdoc. I definitely wasn’t on the job market or looking for a job at the time, but everyone encouraged me to apply. And I did, and I’m really glad that I did. I don’t think I ever made a conscious decision to stay in academia. I was really fortunate to get the job, and it worked out. So, in some ways I didn’t have to make that choice because it wasn’t like I was applying to a bunch of different jobs and choosing between options or anything like that. This was just my option.

AG: You came to Berkeley right before the pandemic, correct? What was that transition like?

AM: I think it helped that the demography department was really generous and allowed me to finish my post doc. So, I actually knew that I was going to start at Berkeley for a year and a half before I started. Postdoc life is really different from faculty life. I started talking to faculty members to get a sense of what it would be like, and I think the fact that I knew for so long that I would transition to the demography department here and I was allowed to finish my post doc was really helpful.

I started in Fall 2019, so that was right before the pandemic. The transition is really hard. As a postdoc, I felt I had a lot of time to really dig into every aspect of my research and I didn’t have any obligations. I didn’t have any students or classes I had to teach. It’s a lot of time that you can devote to just research. And there’s someone, your PI, who will guide you, so you have less chance of just wandering off and doing something that is less fruitful.

I think my first semester here was hard. I feel like I spent all of my time figuring out how to teach well. Everyone tells me it gets easier the more you teach, but it was my first time and I spent all of that semester just working on teaching material for my class. That was hard because I kind of missed doing research. It’s just a different pace. I was on leave for two semesters after that and then I was back again this last semester. Last semester, I was teaching one class but it was a graduate seminar which is a lot more discussion and talking to the students about various topics they are interested in. Last semester was a lot easier in some ways to find the balance between teaching and research but it’s an ongoing struggle for most faculty to figure out that balance.

AG: Berkeley demography is a smaller department. How do you feel like that impacts your job? Especially in terms of mentoring and teaching responsibilities?

AM: I guess it’s hard for me to compare because this is sort of the only department I know. I don’t necessarily feel that I have more responsibilities than if I were in a different department. We don’t have an undergraduate major so there’s a little less pressure than in some other departments which have lots of undergraduate majors, where you definitely have to teach a certain amount of undergraduate classes. We still teach undergraduate classes because we think that’s really important. We have some great undergrad classes. We have graduate students but our cohorts aren’t massive. They’re small relative to the size of cohorts in other departments. Our core faculty is small, but we have a lot of affiliated faculty who mentor students as well. So I don’t feel particularly overburdened being in a smaller department but this is the only department I’ve ever been in so it’s hard to know.

AG: What’s your favorite part of your job?

AM: That’s evolving. I really like research and I always have. I like doing research with other people. I like collaborating with students and I’ve been doing a lot more of that in the past few months which is great. I was really nervous about teaching, because I’d never really done it and I just wasn’t sure if I’d like it. But I actually really enjoyed the undergraduate class that I taught and the graduate class. The undergraduate class was the first one I taught, so I think it highlighted to me that I did like that aspect of my job. I really enjoyed it, and I had a great set of students.

AG: What was the class?

AM: It’s called Demography 110: Introduction to Population Analysis. It’s essentially giving people who haven’t had any exposure to demography a taste of the methods that we use. We talk about topics like how to calculate life expectancy and measure fertility, and what to do if it’s changing over time. I had a really great class. Some of the students did a few research projects with me which was really fun. The Berkeley undergrads I’ve encountered are great. I’m looking forward to doing that class again in the fall. I think it’ll be easier, because I’ve done it once. I think the hardest part of teaching is just figuring out the balance so it doesn't consume all of your time and energy.

AG: Tell me about your research.

AM: My current research is mostly COVID related. In the past a lot of my work has focused on studying childhood infectious diseases. I think a lot about seasonal patterns, so why you see outbreaks at certain times of the year. I did a little bit of work thinking about whether some of that is driven by demographic processes, so people gathering more at certain times of the year or migrating more, versus climate, so the effect of temperature, rainfall, humidity on all these things. I think about both direct and indirect effects. So, if it’s really raining and people are gathered more indoors that could lead to more transmission, or very dry conditions, for example, can influence the transmission more easily. Those have effects on seasonality as well. I’ve also been working a lot on using digital trace data and data from mobile phones to try to understand the movement of people and using that to track how outbreaks spread. A lot of that work is in Bangladesh. I’m trying to understand what the role of Dhaka city, which is a very large urban hub right in the center of the city with a lot of people moving in and out on a daily basis, is on how outbreaks are sparked from parts of the country. It’s hard to study that kind of population movement with census data, so we’ve been using other kinds of data such as that from mobile phones to try and understand those movement patterns. So, some work on that and then with COVID we’ve been doing a lot of work in Bangladesh. We’ve been using different types of data to try to understand the outbreak, because surveillance is so patchy with testing. Testing rates are very different across different parts of the country. Again thinking about population mobility, we’ve been partnering with people who have been looking at the genomic data and thinking about ways to pair both the genomic data and the mobility data, because it tells a parallel story of the disease spread. We think about how you control this disease spread if there’s so much movement in and out of these travel hubs. Broadly, I think a lot about population processes, population movement, fertility and things that affect diseases. I also think about environmental factors such as flooding, monsoon season and things that might affect the spread of the disease.

AG: How do you choose which diseases or geographic areas to look at?

AM: That’s a really good question. For my PhD I was really interested in childhood diseases and there’s a lot of great data from South American countries who are under the Pan American Health Organization (PAHO) System. So, that was where there was the data that I needed for my PhD. But I’ve increasingly become more interested in doing work in Bangladesh, which is where I grew up. I have a lot of great collaborators there. I care about things that happen within the public health domain in Bangladesh. Especially with COVID I’ve been doing a lot of work in Bangladesh.

Where my collaborators are has an impact on the work that I do. We just did some work in Chile, because one of my closest collaborators, who I’ve been working on various projects on Bangladesh with, happens to be from Chile. So, when COVID hit she was interested in studying what was happening in Chile.

I also think a lot of it is driven by where the need for this kind of work is. We did some work in parts of Africa, looking at the impact of the cyclone on cholera outbreaks. We’re doing some work now looking at flooding risk in India and South Asia more broadly due to monsoons and cyclones. Some of that is just that those are the parts of the world that are affected by these climate events. So that’s where this kind of research is most needed. So, it’s a little bit where the data are, what countries are in most need for this research, and where my collaborators are based.

AG: Do you have much interaction with policy and government afterwards? How do you feel your research has an impact?

AM: That’s really hard, because often you do this research and then no one reads it ever. But with some of the work we’ve had an impact. We did a couple of projects in Mozambique after the 2 cyclones hit in 2019. We were working with a partner in Direct Relief, which is an organization that works on the ground to support the World Health Organization (WHO) and other groups and provide relief efforts to areas affected by various kinds of natural or humanitarian crises. So, we had a direct connection with folks on the ground. We shared some of our data directly with them before we published any of it. That’s an example of where we felt there was some use for the work. I think with COVID it’s been really different because there’s been a lot more collaboration between world health organizations and academics all over the world. The majority of work I’ve done for COVID in Bangladesh is not even at the stage where we are writing a publication. It’s all been to aid, for example, what are the things they are going to show on their dashboard or what are the sort of metrics they should be looking at and thinking through all those things. A lot of that is just for policy. It’s been a change of pace, because it all has to be very timely whereas with a research project you could work on it for years and have it be under review for months. It’s been kind of nice to see some of the work hopefully having an impact. I hope that more of that happens in the future.

AG: Do you think the way COVID has shifted the public’s relationship with research will have a longer-term impact?

AM: I hope so. Some of the conversations, at least in Bangladesh, they’re having are about strengthening surveillance in general. So, not just for COVID, but how we can strengthen surveillance and collect data in a way that we can then use it for understanding what’s happening. There’s more interest in collecting data better and using data to gain insights. We did one workshop in the past, but we’re talking about doing another this summer in Bangladesh with public health agencies, where we train them to use the data and help them think through what insights would be most helpful for them. It’s been an ongoing conversation we’ve had for a while now and I think with the pandemic folks are just more interested in making sure the data is being collected and used and helping public health agencies respond to disease outbreaks.

AG: Is it difficult to collect the mobile data? Are there privacy issues?

AM: I think things have changed with COVID. We were working with mobile phone companies, and we’ve also worked a little bit with Facebook’s Data for Good team. One of the main things we really stressed was that we don’t want and we don’t need to see any individual information. We definitely want to make sure that there are no concerns about privacy. So, all of the data we’ve worked on has been aggregated. We come up with these algorithms that mobile phone companies can run on their end to aggregate it so we don’t even see any of the individual level information. So, all we get is that 1000 people moved from this city to that city on a given day. It’s practically impossible to identify anyone. That’s partly because we were very concerned about privacy issues but also because countries have very strict rules and laws about who can look at and use these kinds of data. Everything we did was in collaboration with the ministry of health in Bangladesh for example. I think with COVID a lot more private companies, like Google, Apple, and Facebook, have started making some of these mobility measures publicly available. Again, they’re aggregated. But even beyond just identifying individuals there’s ethical concerns about individuals consenting to this data being used and that’s a really tricky issue. All these companies have just made this data available, so I think there needs to be more conversation about consent and what this data can be used for.

AG: I saw that you had some research on predicting adverse police events, how did that come about?

AM: I did an internship in grad school, again because I wasn’t sure what I wanted to do and I was trying to decide about academia versus something else. So, I did an internship at the University of Chicago called “Data Science for Social Good.” Essentially what they do is pair mostly PhD students, but also some post docs and faculty, with a nonprofit who has a very specific question that they need answered. The year I did it they expanded their reach to government or police organizations. It’s teams of 4 academics paired with an organization. In this case we were paired with the Charlotte-Mecklenburg police department who were really interested in seeing if all this data they were collecting on police behavior they could use somehow to try and understand who might be at risk of having some sort of adverse, avoidable interaction with the public. The idea was if there’s some early warning they could pull them out for more training or counseling and other things. So we did some predictive modelling to try and see if it works. Our internship was just for a summer but the Data Science For Social Good team has been working with them ever since then. This program has gone through several iterations, but I believe the police department is still using the system. Other teams have followed up and done work to understand the effectiveness of the system and all that.

AG: What advice would you give students?

AM: I think when you are a student it feels like every decision is monumental and you should make the exact right one. No, you should just try out things. It’s okay if you don’t like it, you have the rest of your life to figure it out. I took a very meandering path and it was fine. I understand that’s not for everyone and a lot of people don’t have the luxury to take a meandering path, but I think often when you are a student you are very worried because you think everything you do decides your whole life path. But I don’t think it has to. Especially now people switch paths and as long as you have the motivation to do whatever you want to do it’s okay to try different things.


Amanda Glazer is a graduate student in statistics

Photo courtesy of Ayesha Mahmud

Notice something wrong?

Please report it here.