Educational Objectives

How should managers measure, and increase, the social impact of their projects and businesses? How should donors assess the impact of potential beneficiary organizations? Unlike profits, metrics of societal impact cannot be inferred from accounting statements. Yet measurement is critical not only for demonstrating impact, but also for massively amplifying it. Fortunately, tools from data science and econometrics have been developed to navigate the nuances of assessing impact.

Data for Impact (DFI) is intended to train students to become informed and discriminating consumers of evidence so as to enable the more effective management of impact. The course aims to develop data literacy even amongst managers who never plan to implement statistical analyses themselves.

DFI will be of core interest to students with aspirations in social entrepreneurship, socially responsible business, nonprofits, government, impact investing, and effective philanthropy. Indeed, DFI presents students with numerous examples of companies of all types that use data to achieve outsized impact. But, while class discussions will center on social impact, the methods utilized in class extend to any question of causal inference, including questions about whether a particular endeavor increases a firm’s profits, raises customer engagement, etc. It will therefore also be of interest to any manager that aspires to commission and evaluate data analysis as a part of their workflow.

This course requires an openness to -- but no prior background in -- statistical analysis and quantitative thinking.

Course Content and Overview

DFI rests on two foundations: rigorous frameworks for assessing social impact, and an introduction to the tools required to measure it. The course will provide a broad exposure to a variety of analytical methods, presented so as to be accessible to students with no prior background in statistics. Students will learn the conceptual underpinnings and limitations of methods such as randomized controlled trials, difference in differences analyses, regression discontinuities, and statistical prediction among others. Students will also benefit from abundant practice interpreting existing evidence and applying it to pressing managerial challenges.

Grades will be based on in-class participation, several at-home statistical exercises, and a final paper (completed in groups of up to 3), in which students will assess the impact of an existing or aspirational enterprise.

Copyright © 2023 President & Fellows of Harvard College. All Rights Reserved.