Forecasting household affordability with dynamic price monitoring in Lusaka, Zambia

Abstract

This research examines the efficacy of Dynamic Price Monitoring (DPM) as a policy tool for predicting and improving the affordability of essential commodities (ECs) for households in Lusaka, Zambia. The study fills a significant void in existing literature: although prior research has identified descriptive and causal relationships between price monitoring and welfare outcomes, the predictive capabilities of DPM have yet to be thoroughly examined. In Lusaka, where changes in the cost of essential commodities are linked to structural poverty and economic inequality, it is important for good governance to predict affordability shocks before they happen, such as welfare losses. This article formulates and evaluates econometric forecasting models that correlate the dynamic price index (DPI), price volatility, household income, and AEC. The study utilizes a quantitative, longitudinal framework, amalgamating cross-sectional household survey data from 384 participants across three economically distinct regions (Woodlands, Chalala, and Zingalume) with thirty quarters (Q3 2017–Q4 2024) of secondary time-series price data. The study used descriptive statistics, stationarity and cointegration tests, Vector Autoregression (VAR), and Vector Error-Correction Modeling (VECM) as its methodological framework. The results show that there are strong long-term correlations between the variables. For example, a 1-unit rise in DPI is linked to a 0.45-unit rise in AEC, and a 1-unit rise in volatility is linked to a 0.32-unit drop in AEC. The error-correction term of -0.38 (p<0.01) shows that the system tends to go back to the long-run equilibrium after the aftershocks. Prediction simulations show that models that include DPI lower the prediction error (RMSE) by 33.7% and give an average early warning of affordability stress 1.5 quarters sooner than models that don't include DPI. Impulse response functions show that positive shocks to DPI lead to long-term increases in AEC over 6 to 8 quarters. In contrast, volatility shocks cause rapid welfare losses that last for several periods. The findings validate that incorporating predictive econometric modeling converts DPM from a purely descriptive instrument into a proactive welfare governance tool. The study finds that DPM greatly improves affordability estimates, enabling proactive policy changes. It recommends that Zambia's national monitoring framework should include a DPI as a permanent part of it. It also recommends that the country invest in high-frequency data collection and targeted social protection programs to protect vulnerable households from affordability shocks and ensure that everyone has equal access to essential commodities.