iNshiva/food-price-inflation-eswatini

Econometric analysis of food price inflation in Eswatini (1994–2019). OLS regression, ADF stationarity tests, Durbin-Watson diagnostics, and 3-scenario CPI forecast. Built with Python and statsmodels. # Food Price Inflation in Eswatini (1994–2019) ### Econometric Analysis | OLS Regression · Time Series · Scenario Forecasting --- ## Overview This project investigates the key drivers of food price inflation in Eswatini (formerly Swaziland) over a 26-year period from 1994 to 2019. Using IMF World Economic Outlook data alongside local agricultural data, the analysis moves through a full econometric pipeline — from data cleaning and exploratory analysis through to regression modelling, diagnostics, and scenario-based forecasting. The central research question: **what economic and agricultural factors most strongly predict food price inflation in a small, landlocked, import-dependent economy?** --- ## Key Findings - **GDP is the strongest predictor of CPI** (coef = -2.30, p = 0.003) — as the economy grows, inflation tends to ease, consistent with supply-side expansion outpacing demand - **Brent crude oil is a significant secondary driver** (coef = +0.058, p = 0.042) — Eswatini's import dependency means global oil shocks pass through directly to food prices - **The model explains 47.5% of CPI variation** (Adj. R² = 0.33), with all OLS assumptions satisfied - **Imports, maize production, and rainfall** show expected directional effects but are not statistically significant at the 5% level — likely due to the small sample size (n=24) - **GDP non-stationarity** (persistent even after second-differencing) suggests Eswatini's nominal GDP may be heavily influenced by rand/USD exchange rate fluctuations rather than real economic activity — a structural limitation noted as a key caveat --- ## Project Structure ``` food-price-inflation-eswatini/ │ ├── food_inflation_analysis.py # Main analysis script (all phases) │ ├── data/ │ ├── raw/ │ │ └── Data 1.xlsx # Maize production & rainfall (1993–2023) │ └── processed/ │ ├── master_food_inflation.csv # Merged master dataset with lag variables │ └── cpi_forecast_scenarios.csv # Sc …