Analysing the relationship between digitalisation and poverty dimension in Zimbabwe

Abstract

This research delves into the dynamic interplay between digitalisation and various dimensions of poverty within the unique context of Zimbabwe’s ten provinces, using secondary data obtained from Global Data Labs and World Bank for the period 2006 to 2020. The endogenous variable representing poverty was the International Wealth Index (IWI) and the core exogenous variable of concern was digitalisation (DIG) which composed of access to internet, mobile cell phone penetration and access to computer, indexed using principal component analysis. Other control variables included education (EDU), health (HEA), household size (HHSE), house quality (HSEQ), electricity (ELE) and urbanisation (URB). A correlation analysis to determine relationship, stylised fact analysis, Dimistrescu-Hurlin Causality to determine the causal effect, System General Methods of Moments (SGMM) to analyse impact, were harnessed to address the objectives of the study. The results of the study showed that indeed there is a strong positive relationship between digitalisation and poverty, and a significant bi-causal relationship at (1%) significance level. Using the Haussmann test to gauge OLS, FEM, REM, and Arellano and Bond’s Difference GMM estimator through Windmeijer conditions, an Arellano-Bover/Blundel-Bond SGMM estimation model was adopted which addressed the nature and character of data (T<30, N=150) scenario, and resolved the indogeneity and instrumentation as well as over identification issues arising from the data. The SGMM findings showed that indeed digitalisation is a significant determinant of poverty, alongside electricity, household size, and urbanisation at (1%) significance level. The Two Step SGMM provided an impact coefficient of digitalisation on IWI of 0.30, meaning that digitalisation indeed reduces poverty through a (30%) or a 0.30 marginal effect increase to IWI. Pre-estimation tests carried included variance inflation factor for multi-collinearity, heteroscedasticity and stationary tests so as to validate the data. The pre-estimation tests resulted in the data being converted into logarithms, as well as prune education and asset variables as they suffered from serious multi-collinearity. The post estimation tests carried out included Sargan test for indogeneity, Wald test for panel significance, and Arellano Bond tests for serial autocorrelation which all produced plausible results. The study henceforth recommends policy makers to embrace digitalisation as an effective tool to reduce poverty, and efforts such as increase in investment in digital infrastructure, Analysing the relationship between digitalisation and poverty dimension in Zimbabwe appropriate taxation policies and price controls towards digital enablers, creation of a conducive environment for the digital sector to grow, amongst many initiatives. The findings are to provide valuable insights for policymakers, researchers, and practitioners navigating the intersection of digitalisation and poverty in the pursuit of inclusive and sustainable development and fight against poverty in Zimbabwe, from a provincial level, in the face of the unfolding fourth industrial revolution.