مجلة اقتصاد المال و الأعمال
Volume 10, Numéro 2, Pages 67-83
2026-06-30
Cross-sectional Dependence In Panel Data Models: Assessing The Accuracy Of Analyzing Of Factors Influencing On Foreign Direct Investment
Authors : Chekka Abdallah . Medini Atmane .
Abstract
The study focuses on presenting a comparative analysis between traditional fixed panel data models and models that account for cross-sectional dependence among units, using foreign direct investment data from four North African countries during the period from 1990 to 2021 as a case study. Traditional panel models, such as fixed effects and random effects models, assume independence across cross-sectional units, an assumption often violated in macroeconomic data due to common shocks, spatial spillovers, or unobserved global factors. This violation leads to biased inferences and unrealistic estimates because of the significant interdependence among sectors. The Pesaran CD test was employed, which indicated that the residuals across cross-sectional units are independent, thus allowing the use of classical panel models without the need for adjustments related to cross-sectional correlation. Additionally, estimation results were compared, highlighting the crucial role of trade openness and governance quality in attracting foreign direct investment. Through the adopted analytical approaches, this study contributes to clarifying the empirical and methodological implications of cross-sectional dependence in panel data analysis and provides guidance for policymakers on the necessity of handling economic data with high-level assumptions of interconnection and entanglement among economic units.
Keywords
panel model ; cross section independence ; estimation
Les articles similaires
Mosbahi Radhwane
.
Hafid Ilias
.
pages 517-535.
. Mohamed . Retia
.
Khemissi Gaidi
.
pages 31-48.
Bouhanna Keltoume
.
Louaradi Elhadi
.
pages 186-204.
Belkhir Fatima-zohra
.
pages 204-216.
