Writing sample · 27 September 2026
Food Price Monitoring in Nigeria
A Source-Audited Baseline of Spatial Dispersion and Persistence
Abstract
Administrative food-price series inform early warning, social-protection design, inflation surveillance, and the allocation of scarce public resources. Their value depends on disciplined interpretation. This paper establishes a source-audited economic baseline for the National Bureau of Statistics (NBS) Selected Food Price Watch. It documents reported food-price dispersion and conditional relative-price persistence across Nigeria’s states and the Federal Capital Territory (FCT). We reconstruct a fixed-vintage panel of 43 source-labelled food items, 37 jurisdictions, and January 2017–April 2019. The source contains 44,548 state/FCT–item–month records, of which 44,547 are finite and positive. Median item-level cross-jurisdiction log-price dispersion is non-monotonic, ranging from 0.134 in April 2017 to 0.186 in March 2018. In a state-item fixed-effect model of 42,955 consecutive observations, the conditional one-month persistence coefficient is 0.696 (two-way-clustered 95% confidence interval: 0.655–0.737). On a matched 26-outcome-month window, an equal-length split-panel point diagnostic is 0.858, showing that finite-time dynamic-panel treatment materially affects the exact magnitude. The paper establishes two empirical facts about this administrative series. Reported prices are geographically heterogeneous and short-run relative-price deviations are persistent. This reproducible baseline gives policy users and subsequent empirical work a verified account of the series before it is used to interpret deeper market processes.
What the results establish
Reported food prices vary geographically, and relative-price deviations persist in the short run. In the primary specification the conditional one-month persistence coefficient is 0.696, with a two-way-clustered 95% confidence interval of 0.655–0.737.
An equal-length split-panel point diagnostic on a matched window is 0.858. It has a different inferential role and no separately validated uncertainty procedure. It is not an interchangeable precision estimate.
This is a historical, pre-COVID monitoring baseline. It does not identify causes of price differences, establish market integration, or directly measure household food security.
Read the writing sample
26 pages · PDF. If your browser cannot display it here, use the direct PDF link above.