Application of Panel Data Analysis and Modelling to Economic Data: A Case of Determinants of Economic Growth for SADC and Zimbabwe

Abstract

In order to accurately forecast economic growth, it is important that growth determinants are identified. However Africa and Southern African Development Community (SADC) region in particular have not identified any determinants of economic growth that are peculiar to the SADC region. The aim of this research is to establish models that links economic develop ment as measured by GDP to the determinants of economic growth for the Zimbabwean and SADC economies.In this study determinants of economic growth are gathered and evaluated for sixteen SADC countries for twenty two years (2000 to 2021), that dictates use the panel data analysis, whereas panel data may have group effects, time effects or both. Data is taken from various sources but mainly the World Bank website for different SADC countries con tributing in the world economy. In this article, the comparison of ordinary least squares (OLS) model, fixed effects model (FEM), Machine learning (ML) and Random effects model (REM) for SADC nations panel data were carried out. F-test was used as a specification test to make a selection between OLS model and fixed effects model, The Breusch-Pagan test was used to choose between OLS and REM while the Hausman test was used as a specification test for FEM and REM. A fixed effects model with an adjusted R 2 value of 98% which is very plausible was realised to be the best model to handle the SADC community economic data.Imports, exports, external debt, international reserves, unemployment and labour force had positive impacts on the SADC community’s economic growth. Foreign direct investment negatively influenced economic growth. Inflation, exchange rate and interest rate had no association with economic development for the SADC community. As for country effects, it was established that South Africa had a positive impact on gross domestic product (GDP), whereas all other SADC nations country effects negatively affected economic growth with the exception of Comoros and Seychelles, whose effects had no significant effects on economic growth. For Zimbabwe Deep learning modelling and the convectional model with log transformations were the best models and had almost the same predictive powers, Exports,Foreign direct investment and Labour force positively influenced economic growth.Inflation, external debt, interest rate and exchange rate had negative impacts on GDP. International reserves,imports and unemployment rate had no association with economic growth. Forecasts were done for Zimbabwe’s GDP and it was realised that the GDP will increase for years 2022 to 2025. Based on these findings, the study recommends that policymakers in the SADC region prioritize areas such as imports, ex ports, external debt, international reserves, employment levels, and the labor force to stimulate sustainable economic growth. Furthermore, it is crucial to address challenges such as inconsis tent power supply and integrate trade regulations to foster economic development in the reg