Algorithm is Experiment: Machine Learning, Market Design, and Policy Eligibility Rules

Algorithms produce a growing portion of decisions and recommendations both in policy and business. Such algorithmic decisions are natural experiments (conditionally quasi-randomly assigned instruments) since the algorithms make decisions based only on  observable input variables. The authors use this observation to develop a treatment-effect estimator for a class of stochastic and deterministic decision-making algorithms. Their estimator is shown to be consistent and asymptotically normal for well-defined causal effects. A key special case of their estimator is a multidimensional regression discontinuity design. They apply our estimator to evaluate the effect of the Coronavirus Aid, Relief, and Economic Security (CARES) Act, where hundreds of billions of dollars worth of relief funding is allocated to hospitals via an algorithmic rule. Their estimates suggest that the relief funding has little effect on COVID-19-related hospital activity levels. Naive OLS and IV estimates exhibit substantial selection bias.