A data-driven approach to improving social impact
Whether seeking to feed the hungry, improve access to health care, or increase literacy rates, social sector agencies are driven to launch international development projects because of an earnest desire to help those in need. But what information do decision-makers use to craft these critical interventions? How do they determine if their good intentions are actually yielding effective results?
Evaluating a program’s impact is a multifaceted and challenging problem, and one Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) alumnus Deng-Tung Wang, A.B. ’17, considers daily as an associate at IDinsight. The non-governmental organization partners with client organizations around the globe to study challenges faced in developing nations and then design and test potential solutions.
“Our line of work really deals with using data-driven evidence to influence decision-making for a range of actors, including governments in developing countries, foundations, and bilateral agencies,” said Wang, A.B. ’17, a joint concentrator in environmental engineering and environmental science and public policy. “We work with organizations like the Gates Foundation or UNICEF that have the agency and willingness to adopt our recommendations. With projects aimed at improving lives, these entities use our evidence to learn, grow, and increase impact.”
Operating out of IDinsight’s office in Lusaka, Zambia, Wang is currently evaluating a project aimed at increasing vaccination rates in northern Nigeria. He began by conducting a baseline study to better understand the landscape of immunizations in this remote region; Wang found that only one in 10 children in northern Nigeria have received the full package of immunizations. His study also showed links between vaccination rates and socioeconomic status, as well as the education level of the caregiver.
A member of the nonprofit's program staff confirms that a mother has received her incentive after vaccinating her baby. (Photo provided by Deng-Tung Wang.)
Building off that data, the nonprofit client is planning to implement a cash transfer program, incentivizing caretakers with money as an enticement to ensure their one- and two-year-old children receive vaccinations. The team will oversee a two-year long randomized control trial of families in the program, and then analyze that data to see if the incentives provide a measureable improvement.
“With immunizations, there are tens of thousands of lives in this part of Nigeria that could be at stake,” he said. “If we do find that this program works, there is potential for it to scale up across a larger region, even across the entire country. The number of lives that could be impacted by a working program is huge.”
But the challenges of Wang’s work loom large, as well. It can be very stressful to work on such far-reaching projects in a deadline-driven environment, especially in unfamiliar territory. For instance, Wang traveled throughout rural northern Nigeria for weeks to collect data related to immunizations—an assignment well outside his comfort zone. When there are lives at stake, analyzing a brand new problem adds an extra element of tension, he said.
Wang and two of his coworkers eat a meal in the village head’s home. Households in this village eat from the same pot. (Photo provided by Deng-Tung Wang.)
And sometimes, after weeks or months of data gathering, the evaluations conducted by IDinsight show that a program is not working as intended. Sharing that information with a client leads to difficult conversations.
“Sometimes, clients have devoted 20 or 30 years of their lives to working on this intervention,” he said. “Articulating a program’s failures in a positive light is very difficult. When we find that a program doesn’t work, we emphasize ways that it could be improved, and underscore that their data is still important and could inform future projects.”
For Wang, the problem-solving skills and analytical abilities he developed at SEAS help him navigate the day-to-day challenges of his work. After initially planning to pursue a corporate consulting career, he changed tracks because he felt drawn to work in an arena where profits weren’t the driving goal. When asked during an interview why he wanted to work in corporate finance, Wang realized he didn’t have an answer. He got up, walked out of the interview, and never looked back.
Wang in the field with his team of enumerators after just arriving in a village to begin data collection. (Photo provided by Deng-Tung Wang.)
He’s found the perfect outlet for his skills at IDinsight. The work is similar to financial consulting, but instead of measuring profits, he’s evaluating success with different metrics. He also relies on his engineering background often, whether conducting a detailed mathematical analysis or thinking through a complex problem by breaking it down and objectively critiquing each element.
Fresh out of Harvard, Wang’s first few months on the job have taught him a great deal about how to conduct analytics in an applied setting, how to collect good quality data, and the many cultural and infrastructure challenges of running programs in under-resourced areas.
“The biggest reward of this work is knowing that a report I sent, or the data I’m gathering, will influence some kind of decision by an organization that could improve many people’s lives,” he said. “That gets me out of bed everyday.”