This work presents a multi-objective optimization strategy for fair vaccine allocation through different fairness schemes. The proposed approach considers a diverse series of parameters related to different public health data and social behaviors that influence the correct distribution of vaccines, such as corruption and crime. Simultaneously, the formulation includes prioritizing those groups with the highest risk based on the epidemiological traffic light. Furthermore, the presented strategy involves different budget constraints that allow identifying trade-off solutions through Pareto fronts. Therefore, vaccine allocations are obtained by combining fairness concepts with multi-objective optimization. The applicability of the model is illustrated using the case study of Mexico. The solution to the proposed scenarios was carried out using different justice schemes and an economic objective function. The results show the compromises between a satisfaction index and costs, which are shown through Pareto optimal solutions that allow selecting the solutions that balance the objectives. The solutions provided by the social welfare scheme suggest a greater allocation of vaccines to those states with higher epidemiological risk, which may be helpful in the first stage of vaccination. On the other hand, the Rawlsian scheme provides more balanced solutions that can be useful in situations with lower rates of infection. Finally, the Nash scheme is the one that provides the most balanced solutions, favoring to a lesser extent the areas with the highest epidemiological risk, which may be useful in the later stages of vaccination.
Keywords: COVID-19; Fair allocation; Optimal distribution; Vaccines.
© 2023 Elsevier Ltd. All rights reserved.