Ensuring multidimensional equality in public service

https://doi.org/10.1016/j.seps.2021.101176Get rights and content

Highlights

We consider optimization problems to allocate multiple benefits to multiple users.

We propose two multi-objective models to deal with equity and efficiency concerns.

We demonstrate the use of the approaches on a public education planning problem.

The results show the benefits of considering equality in public service planning.

Abstract

Service planning problems typically involve decisions that lead to the distribution of multiple benefits to multiple users, and hence include equality and efficiency concerns in a multidimensional way. We develop two mathematical modeling-based approaches that incorporate these concerns in such problems. The first formulation aggregates the multidimensional efficiency and equality (equitability) concerns in a biobjective model. The second formulation defines an objective function for each benefit, which maximizes the total social welfare obtained from that specific benefit distribution; this results in an n-objective model, where n is the number of benefits. We illustrate and compare these approaches on an example public service provision problem.

Keywords

Equality
Equitability
Fairness
Public service provision
Public education
Knapsack problem
Equity
Epsilon constraint algorithm
Bi-objective optimization

Özlem Karsu is an Assistant Professor of Industrial Engineering at Bilkent University. She received her Ph.D. degree in Operational Research from the London School of Economics. Her research interests include inequity-averse decisions, multi-criteria decision making approaches and various applications of multi-objective optimization.

Damla Akoluk is a Ph.D. candidate in Multi-Actor Systems Department at Delft University of Technology. She received her M.Sc degree from the Industrial Engineering Department of Bilkent University and B.S. degree from Sabancı University, Industrial Engineering Department. Her research interests center around multi-objective optimization and advance multi-objective decision support.

View full text