Distributional and Tail‐Dependent Perspectives in Economic Relationships: A Review of Quantile Regression Application
Shajara Ul‐Durar, Konstantinos A. Dimitriadis, Noman Arshed, Marco De Sisto, Hamidreza Harati
Abstract
ABSTRACT There is an increased proportion of studies using quantile‐based regression methodology (QR) in economics. They offer a robust alternative to classical mean regressions, which can estimate non‐normal variables with distributional heterogeneity in the dependent variable. This study synthesizes the theoretical foundations, methodological advancements, and empirical application of QR in economics that traces the evolution from the foundational work of Koenker and Basset. The targeted studies are from Scopus. Bibliometrix library in R is used for bibliometric analysis, and Structured Literature Revie (SLR) is conducted on selected studies. The Scopus query started with 250 studies and was finalized at 53 studies that focused on the motivation of using QR, comparison of ordinary least squares (OLS), and QR in economics. The systematic review has spanned the past decade. The consolidation of fragmented evidence showed that QR can advance econometric debates by providing superior data insights. The insights presented in this review are aimed at bridging the gap between econometric development and applied economic policy research. This paper contributes to a deeper understanding of distribution‐sensitive modeling strategies, offering valuable implications for economists in academia, government, and industry.