new_data
Description
This study draws on a balanced panel dataset covering 14 ECOWAS countries from 2010 to 2023 in order to examine how political and institutional alignment in the rice sector relates to rice productivity over time. The database was constructed by combining institutional, agricultural, macroeconomic, trade, and climate variables from several complementary sources. Its main purpose is to capture not only changes in rice output performance, but also the broader policy and governance conditions within which rice production evolves across West Africa. The central explanatory variable is the Political and Institutional Alignment Index, designed to measure the degree of coherence between national rice sector governance and regional agricultural commitments. This index is built from four core dimensions: strategic alignment, regulatory alignment, multi-stakeholder coordination, and monitoring and evaluation. Each dimension was coded annually using an ordinal scale ranging from 0 to 2, where 0 reflects absence or very weak institutionalization, 1 reflects partial or emerging institutionalization, and 2 reflects full institutionalization. The coding process relied on information drawn from CAADP Biennial Review reports, ECOWAP-related monitoring frameworks, and National Rice Development Strategies. To improve data quality and consistency, the coding was reviewed and refined through a structured validation process involving national focal points familiar with the rice sector and the implementation status of national strategies. After validation, the four dimensions were aggregated into a synthetic indicator using Principal Component Analysis. The dependent variable is rice yield, used here as a proxy for sectoral productivity. In addition to the institutional index, the dataset includes a broad set of control variables intended to capture the main factors that can influence rice productivity. These include seed use, cultivated rice area, relative price incentives, GDP-related indicators, electricity access, household consumption per capita, and agricultural credit. Climate-related variables, especially temperature anomalies, were also incorporated to account for environmental shocks affecting agricultural performance. In some model specifications, trade variables such as rice import volumes or import values were added to reflect the role of external market exposure and competitive pressure from imported rice. Taken together, these data provide a multidimensional representation of the rice sector in ECOWAS. They make it possible to assess whether countries that are more institutionally aligned, more coordinated, and better equipped in terms of policy follow-up tend to achieve stronger productivity outcomes over time, while controlling for broader structural and climatic constraints.
Files
Steps to reproduce
Download the replication dataset and associated code files from Mendeley Data (DOI: 10.17632/ny43f8k5kk.1). Open the final country-year dataset covering 14 ECOWAS countries for 2010–2023 (196 observations). Verify that the variables include rice yield, the four institutional dimensions used to construct the PIA, government effectiveness, political stability, seed use, domestic-to-imported rice price ratio, cultivated area, household consumption per capita, electricity access, and temperature anomaly. Reconstruct the Political and Institutional Alignment (PIA) index using principal component analysis of the four institutional dimensions: strategic alignment, regulatory alignment, multi-stakeholder coordination, and monitoring and evaluation. The first principal component is retained as the PIA score. Generate the interaction terms PIA × government effectiveness and PIA × political stability. Reproduce descriptive statistics, correlation matrices, KMO statistics, PCA eigenvalues and loadings, and the panel diagnostic tests reported in the manuscript. Estimate the baseline two-way fixed-effects model with country and year fixed effects. Compute Driscoll–Kraay standard errors to account for heteroskedasticity, serial correlation, and cross-sectional dependence. Reproduce the robustness specifications using one-period lagged regressors, the dynamic specification with lagged rice yield, the standardized PIA, winsorization at the 1st and 99th percentiles, and country-clustered standard errors. Compare the generated coefficients, standard errors, significance levels, observation counts, and within R² values with Tables 1–8 and the Appendix. Minor numerical differences may arise from software versions or rounding.
Institutions
- Africa Rice CenterAbidjan Autonomous District, Cocody
- Nazi Boni UniversityHauts-Bassins, Bobo-Dioulasso