We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.SI

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Social and Information Networks

Title: Allocating Stimulus Checks in Times of Crisis

Abstract: We study the problem of allocating bailouts (stimulus, subsidy allocations) to people participating in a financial network subject to income shocks. 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 the types of stimulus checks people receive around the world as part of COVID-19 economical relief plans. 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 rations. Then, we incorporate multiple fairness constraints in the optimization problems and establish relative bounds on the solutions with versus without these constraints. Finally, we apply our methodology to a variety of 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.
Comments: Colorblind-friendly figures; Fix typos; Restructured Introduction; Reorganized approximation algorithm proofs; Add acknowledgements. To be presented at EAAMO '21 (non-archival track)
Subjects: Social and Information Networks (cs.SI); Computer Science and Game Theory (cs.GT)
Cite as: arXiv:2106.07560 [cs.SI]
  (or arXiv:2106.07560v3 [cs.SI] for this version)

Submission history

From: Marios Papachristou [view email]
[v1] Mon, 14 Jun 2021 16:17:50 GMT (4629kb,D)
[v2] Tue, 31 Aug 2021 15:03:31 GMT (4944kb,D)
[v3] Mon, 20 Sep 2021 19:54:09 GMT (9151kb,D)

Link back to: arXiv, form interface, contact.