Allocating Stimulus Checks in Times of Crisis

We study the problem of financial assistance (bailouts, stimulus payments, or subsidy allocations) in a network where individuals experience income shocks. These questions are pervasive both in policy domains and in the design of new Web-enabled forms of financial interaction. We build on the financial clearing framework of Eisenberg and Noe that allows the incorporation of a bailout policy that is based on discrete bailouts motivated by stimulus programs in both off-line and on-line settings. We show that optimally allocating such bailouts on a financial network in order to maximize a variety of social welfare objectives of this form is a computationally intractable problem. We develop approximation algorithms to optimize these objectives and establish guarantees for their approximation ratios. Then, we incorporate multiple fairness constraints in the optimization problems and study their boundedness. Finally, we apply our methodology to data, both in the context of a system of large financial institutions with real-world data, as well as in a realistic societal context with financial interactions between people and businesses for which we use semi-artificial data derived from mobility patterns. Our results suggest that the algorithms we develop and study have reasonable results in practice and outperform other network-based heuristics. We argue that the presented problem through the societal-level lens could assist policymakers in making informed decisions on issuing subsidies.