Moving Past the Digital Smokescreen: How to Harness Big Data to Solve Social Welfare Problems

Big data and computational methods have generated great excitement and enthusiasm among scholars across the social sciences. In the form of government records and digital traces collected by companies, big data are abundant, constantly updated, and inexpensive to collect, giving researchers an unprecedented window into everyday behavior. For sociologists interested in solving social problems, big data likewise seem like an attractive source of evidence for developing policies and interventions. In this paper, we draw on lessons from our research on online fundraising and social welfare provision, which uses both big data analyses as well as conventional methods, to consider how big data can be used to solve unsolved social problems. We introduce the smokescreen analogy to characterize how big data can obscure rather than clarify social problems. Digital traces often serve as a veneer over longstanding issues, leading researchers into a "description trap" where they generate detailed accounts at the expense of identifying causes and solutions. However, behind every smokescreen is a smoke machine. We argue that analyzing the platforms generating digital traces can reveal their causal role in shaping social problems. Big data's relationship to problem-solving is ultimately ambivalent: it can both enable new solutions and distract from enduring causes.