Optimal resource allocation: Convex quantile regression approach

TL;DR

This paper introduces a convex quantile regression approach for optimal resource allocation, enabling efficient decision-making in multi-unit systems by accounting for noise and heteroscedasticity, and demonstrating significant productivity gains.

Contribution

It develops a novel quantile allocation model using convex quantile regression to improve robustness and flexibility in resource allocation analysis.

Findings

Significant potential for productivity improvements in Finland's business sector.

The model effectively handles heteroscedasticity and noise in data.

Resource reallocation can lead to substantial efficiency gains.

Abstract

Optimal allocation of resources across sub-units in the context of centralized decision-making systems such as bank branches or supermarket chains is a classical application of operations research and management science. In this paper, we develop quantile allocation models to examine how much the output and productivity could potentially increase if the resources were efficiently allocated between units. We increase robustness to random noise and heteroscedasticity by utilizing the local estimation of multiple production functions using convex quantile regression. The quantile allocation models then rely on the estimated shadow prices instead of detailed data of units and allow the entry and exit of units. Our empirical results on Finland's business sector reveal a large potential for productivity gains through better allocation, keeping the current technology and resources fixed.