Quantitative Work with a Direct Social Impact: Harris Alumni Discuss Working for Mathematica Policy Research

Not long after getting his PhD at Harris in 1992, Matthew Stagner served as a federal project officer for the national sexual abstinence evaluation for the U.S. Department of Health and Human Services. When he told people what he was working on, after receiving a barrage of predictably emotional responses, he had a revelation: Most everyone he came into contact with was eager to accept the evaluation if it upheld their beliefs—and reject it if it didn’t. “It was like that old joke,” says Stagner: “They used evidence the way a drunk uses a light post: for support rather than illumination.” He began to wonder what the purpose of producing rigorous evidence was in a world where people were already dug in on their version of the truth.

The organization that conducted that sexual abstinence evaluation was Mathematica Policy Research. Twenty years later, Stagner was heading up Mathematica’s Chicago office, absorbed in data and social science to produce rigorous, evidence-based solutions for government agencies—and continuing to ponder the implications of his revelation.

A Princeton, New Jersey–based company that uses data, analytics, and technology to address pressing social challenges, Mathematica has long served as an ideal landing place for Harris graduates. The employee-owned company boasts a team of 1,600-plus data and social science experts in eight locations from Seattle to Cambridge, Massachusetts, each of them aiming, through quantitative work with a direct social impact, to add knowledge and understanding to what these government programs accomplish—and how they might accomplish it better. This makes Mathematica a natural extension for those who have absorbed Harris’ core mission and educational approach.

 “Harris is dedicated to the quantitative and the rigorous, but it's also built around the edges much more about leadership and use of evidence and ways to manage—as opposed to just test and evaluate,” Stagner said.  

April Yanyuan Wu, PhD’10, a senior researcher at Mathematica employee since 2014, says the company has long had a great reputation among Harris students. “It’s a place for big ideas, and Harris taught us the same thing Mathematica does: Always try to make a difference,” Wu says. “We are trying to use the data we have, the technology and the methodology we have, to try to understand how to implement government programs—and understand how this program touches people's lives.”

Mathematica’s vision and reliability has served it well in partnerships with federal government agencies, state and local government agencies, foundations, international aid organizations, and commercial health organizations. Whether addressing the effects of climate change, disability policies, or the efficacy of birth control education, the goal is to improve programs, refine strategies, and, above all, deepen understanding of the world around us. These values dovetail perfectly with the skills of dedicated researchers like Wu, who flourished under Harris’ emphasis on quantitative studies. “A lot of things I learned from Harris are still the core to the work that I do – like the statistics econometrics, programming valuation, and cost-benefit analysis,” says Wu. “It’s aligned with my vision.”

At Mathematica, where the workload is driven by the government’s agenda, Yu focuses on employment of older workers, individuals with disabilities, and other vulnerable groups, and covers a broad area of topics in public finance and labor economics. Some Mathematica employees remain on one topic for years; more light-footed workers straddle multiple areas simultaneously, a dance Stagner likens to a game of Twister. “You get put onto a red square or a green circle, and then you are given the opportunity to stretch over to the yellow triangle,” he says. “And you can be in two, or three, or even four different types of projects at once—but you also have to be careful not to stretch yourself so far and try to touch so many things that you collapse.”

Laura Meyer, MPP’15, is playing that game right now. After a varied educational career that included an Arabic Studies degree at Notre Dame, an internship in Delhi, volunteer stints in Geneva, Egypt, and South Sudan, and a year as a graduate research assistant at the UChicago Crime Lab, now housed at Harris, Meyer graduated when Mathematica was about to launch its international unit into a standalone division within the company. This sounded like a perfect fit.

Since landing at Mathematica’s international unit in Seattle in 2015, Meyer has juggled a full slate of research and evaluation projects that focus primarily on international development plans for governments and foundations, many in different stages and different parts of the world. “Two weeks from now, I may spend my morning on women’s livelihoods and business micrometric entrepreneurship in Jordan,” Meyer says. “Or maybe working on early childhood development in Tanzania, Kenya, and Mozambique. I have a domestic education study in a couple of school districts, including one in Chicago. Projects come up and that keeps it interesting.”

Jonah Deutsch, PhD’13, also went directly from Harris to Mathematica’s Chicago office, where he contributes to three projects for the U.S. Department of Labor. He and his team design, conduct, and evaluate behavioral interventions, one involving compliance assistance for employers with the Wage and Hour Division; another the Trade Adjustment Assistance program; a third includes three separate behavioral trials across different agencies. All of it has enormous implications on governmental policy. “What I do feels important and inspiring,” says Deutsch, a senior researcher. “And it really affects people’s lives who need good services.”

At the same time, Deutsch is also knee-deep in a massive nationwide study of apprenticeship programs as a new form of workforce development and training. “Apprenticeship programs are becoming increasingly popular, but there's very little hard evidence about whether they actually work,” he says. “So I’m very excited that I'm leading a study that will provide the first really rigorous evidence.”

Conducting research studies is only part of the equation. For many of Mathematica’s researchers, a central challenge of the job involves breaking down highly technical work and presenting it to an audience that’s not as well-versed, or interested, in the details—an audience with the power to implement or reject a program. “How do we convey that we thought about all of the minute details of this, so you can trust our answer, without talking about any of those 19 details?” asks Deutsch. “Sometimes they’re subtle points, and we have to figure out how to get them across when we only have five minutes.” Likewise, Wu says that learning how to write in plain language for policymakers and practitioners represents her steepest learning curve. “We want to show the bigger picture of what we do and make a difference,” says Wu. “But if you don't effectively get the word out, how do you make a difference?”

It’s safe to say that Deutsch, Meyer, Wu, and Stagner – who count among their Mathematica colleagues a number of other Harris graduates – have all at one point or another encountered a program that had not been rigorously evaluated before it began. “So many programs that get implemented don't have the same effect that people sitting in a room in D.C. think they're going to, because they don't really take human nature into account,” says Meyer. “There are so many counterintuitive findings because people don’t always behave the way you think they will.” Yet as Mathematica’s employees spend their days trying to build better mousetraps, one trait unites them all: a confidence in the power of government to effectively improve lives.

“Whether you’re doing qualitative work, data analytics, using survey data, using big data, or being what we now call learning and strategy consultants as opposed to evaluators, you need a belief in what these government programs are trying to do,” says Stagner. “I personally hope to help people improve the policies so they can impact the people that they’re trying to serve.”

As for the question Stagner posed—what is the purpose of producing rigorous evidence in a world where people are already dug in on their version of the truth?—the answer, he says, lies in who’s planning to use the evidence. “The larger political world and Twitterverse may be less inclined to care about evidence,” says Stagner. “But just like businesspeople use data and evidence to improve their bottom line, the operators of programs will strive to use data and evidence to achieve their key outcomes. They will always care deeply about evidence.”