Natasha Mathur, MSCAPP’19, on the Connection Between the Real World and Data

Natasha Mathur, MSCAPP’19, a New York native and graduate of New York University’s Stern School of Business, was drawn to the Harris School of Public Policy by its rigorous MSCAPP program, a Masters in Computational Analysis & Public Policy offered jointly with the Department of Computer Science.

“It was the only program like it at the time,” says Mathur, 30. After doing data analysis on Elizabeth Warren’s and Joe Biden’s presidential campaigns, then a stint of political consulting, Mathur became the director of data at Maisha Meds, an app that helps more than 3,200 businesses deliver affordable health care in sub-Saharan Africa.

Have you always been obsessed with data?

Growing up, I was good at math, but also a big reader. Sometimes people would say, “Either you’re a STEM person or you’re a literature person,” and I never thought that made sense. It’s the data that feeds liking puzzles, liking problems, and figuring things out.

What was your experience like, getting a crossover degree in the MSCAPP program?

I loved it. We had so much work going on, but we were always hanging out together. There were amazing opportunities, like when the Japanese students organized a class trip to Japan during spring break of our second year. We went to different government departments in Tokyo, met officials, and they talked to us about different issues that they’re facing. We also met Harris alums from many years ago who were working for government there.

What Harris professors or classes particularly made an impression on you?

I loved Science of Campaigns and Elections, taught by Alexander Fouirnaies. Those classes made a big impact.  Also, Dan Black’s advanced statistics and program evaluation classes. I learned a lot!

How has the Harris curriculum and culture informed your career?

It was the idea of approaching everything in a mathematical way. Every problem can be modeled somehow. I found that idea especially helpful with political campaigns. Sometimes we were faced with situations that there was no clear framework for, and you need a way to break that down. How do you model an election in the middle of a pandemic? On Senator Warren’s presidential campaign, we were focused on grassroots fundraising, so we had to figure out how to answer those questions with a strong basis in theory or on actual calculation–not a hunch. It was funny because when you walked around the Warren campaign office, you would see UChicago pencil cases everywhere. (I have one, too.)

How does the emphasis on data affect presidential elections?

Whenever I try to explain this to people from other countries, they’re like, What do you mean the data comes in a different format from every state? But it absolutely does. When you vote by mail, the state makes a record of it, and if you request multiple ballots or records of that, or if it gets rejected, that data is stored in the most confusing way possible–and it’s different in every state. So, we did a lot of work to collect and clean that data, so we actually knew what was going on in each state. In an ideal world, this should be obvious. But it isn’t. 

You’ve done a lot of work in AI. What are the biggest misconceptions about it?

At the end of the day, AI is really fancy statistics. There are levels on that, but any issue that you might have with a simple model you make can also be an issue with AI. For example, Google used to have an AI feature that was giving ridiculous results like, “You should put glue in your spaghetti,” because it was based on all the information on the internet, and the information it happened to pull was a strange joke made on Reddit, which AI didn’t have context for. AI should be thought of as a very useful tool, not a prelude to I, Robot.

What led to your pivot from political campaigns to the policy world?

I got to do my share of campaign work. People will ask, “Oh, did you watch the debate?” And I say, I never want to watch a debate again and wanted to do something where I could have a continuous and long-term impact. Maisha Meds is more conducive to the kind of impact I was hoping to make on a regular basis.

What can you tell us about Maisha Meds?

We build mobile software for private pharmacies and clinics in sub-Saharan Africa. It's built around an Android app that helps them run their business, and we also use  it to reimburse pharmacies for providing high-quality care at a low cost to people who visit them. 

Global aid often builds programs around public hospitals and research centers, but for a lot of people, their main point of care is a private drug shop or pharmacy close to where they live. It’s important to have a path to help people through either of those ways. 

My team and most of the people I work with are in Kenya, and we also have teams in Uganda, Tanzania, and Nigeria. You want to see how people actually use the app in real life, and what problems they come up with. Data always needs context. We’re helping with malaria tests. We’re testing out new ways of delivering HIV PrEP through private facilities. And we’ve done a lot of work making long-acting contraceptives more accessible in private pharmacies where short-acting options are more common.

What drives you to do the work you do?

I like when the data is close to the ground. The link between the campaign work and Maisha Meds is that there’s a field component to every data decision that gets made. I like both knowing where the data comes from and what the impact will be. I know that literally, even though the data feels removed, it’s going to lead to someone getting more affordable contraception. You can’t beat that. It’s that connection between data and anything in the real world I find really interesting.

What do you do in your spare time?

I like to go swimming. I used to swim at Ratner. I like doing jigsaw puzzles, a pandemic hobby that I’ve kept. And I love to travel in general, so this job has been great. I do a lot of it individually too. I’m in New York, so just a lot of trying out new restaurants and things like that.

What words of wisdom would you give to Harris students?

There is a lot going on across campus in terms of speakers and events. Take full advantage of that, because you’re not going to have easy access to that later on. And in terms of careers, a lot of people have told me my career is not linear or this and that. So I’d say: Get information from as many places as you can, then think very critically about how you want to make an impact and use your skills. Obviously, you sometimes have to make compromises in your career, but you can find a job that balances what you’re interested in, what skills you want to use, and where you want to live.