AI, Fiscal Space, and Social Protection in Four Latin American Economies: ;A Scenario-based Threshold Framework
By Vicenzo Scavino
Can fiscal space from artificial intelligence finance recurrent social protection in Latin America? This paper develops a prespecified, scenario-based feasibility test that translates an AI dividend into income and fiscal capture, compares it with policy costs, and applies debt-consistency and persistence screens. For Peru, Chile, Colombia, and Mexico, no deterministic crossing occurs under non-stress scenarios because translation and capture jointly fall short. Crossings arise only under diagnostic disruption: Chilean guaranteed minimum income is conditionally feasible and debt-consistent; Peru’s social pension crosses accounting but fails the debt guardrail; Colombia and Mexico do not cross. Crossing frequencies are model-implied, not predictive probabilities.
Source SSRN
