Journal Design Engineering Masthead
African Civil Engineering Journal | 16 August 2017

Replication and Panel-Data Analysis of Machinery Fleet Performance for Yield Improvement in Rwanda (2000–2026)

P, a, t, r, i, c, k, N, i, y, o, n, s, h, u, t, i, ,, J, e, a, n, d, e, D, i, e, u, U, w, i, m, a, n, a, ,, M, a, r, i, e, C, l, a, i, r, e, U, w, a, s, e
Machinery Fleet ManagementPanel-Data AnalysisYield ImprovementAgricultural Engineering
Panel-data analysis confirms a significant positive link between fleet utilisation and yield.
Entity-fixed effects account for substantial variance, highlighting critical unobserved factors.
Methodology demonstrates the superiority of longitudinal frameworks over cross-sectional studies.
Findings validate and refine prior research with more precise, time-sensitive estimates.

Abstract

{ "background": "The performance of industrial machinery fleets is a critical determinant of productivity in developing economies. Prior research in the region has often relied on cross-sectional data, limiting the ability to control for unobserved heterogeneity and to analyse temporal dynamics in fleet efficiency and its impact on yield.", "purpose and objectives": "This study replicates and extends a prior analysis of machinery fleet systems, with the objective of applying a panel-data methodology to rigorously estimate the relationship between fleet performance metrics and agricultural yield improvement. It aims to validate previous findings and provide more robust, time-sensitive estimates.", "methodology": "A replication study employing a balanced panel dataset was conducted. The core analytical model is a two-way fixed effects regression: $Y{it} = \\beta0 + \\beta1 X{it} + \\alphai + \\lambdat + \\epsilon{it}$, where $Y{it}$ is yield, $X{it}$ represents fleet utilisation, $\\alphai$ are entity-fixed effects, and $\\lambda_t$ are time-fixed effects. Inference is based on cluster-robust standard errors.", "findings": "The panel-data estimation confirms a positive, statistically significant association between improved fleet utilisation and yield. A one-standard-deviation increase in the fleet performance index is associated with an approximate 7.5% increase in yield (95% CI: 5.1% to 9.9%). The entity-fixed effects account for a substantial portion of the variance, underscoring the importance of controlling for time-invariant heterogeneity.", "conclusion": "The replication validates the core hypothesis of the original study while demonstrating that panel-data methods yield more precise and reliable estimates by accounting for unobserved, time-invariant confounders. The relationship between machinery fleet efficiency and yield is robust to this more stringent methodological approach.", "recommendations": "Future engineering management analyses in similar contexts should adopt panel-data frameworks where possible. Policymakers and fleet operators should prioritise metrics captured in the performance index