This study examines the dynamic empirical relationship between fiscal resource allocations, secondary school academic performance (NECO SSCE benchmark), socioeconomic constraints, and out-of-school populations in Nigeria over a 26-year period (2000–2025). Employing annual time-series data, the preliminary descriptive statistics reveal significant variability across key indicators, with out-of-school children averaging 13.73 million (SD=3.30 million), fiscal education allocations averaging N586.13 billion (SD=759.98 billion), and academic pass rates (≥5 credits including English and Mathematics) averaging 44.32% (SD=21.91%). Pairwise correlation analysis indicates strong positive associations between education budget allocations and out-of-school children (r=0.8469, p<0.01), alongside moderate correlations with academic performance (r=0.4489, p<0.05), highlighting systemic structural bottlenecks. Standard Dickey-Fuller and Zivot-Andrews unit root tests with structural breaks confirm that key series exhibit second-order non-stationarity (I(2)) driven by regime shifts (notably in 2012, 2017, and 2022), rendering standard Autoregressive Distributed Lag (ARDL) bounds testing invalid.
To resolve this, the study implements the Juselius-Johansen I(2) Cointegration procedure [cite: 1]. The joint trace rank test identifies a single common I(2) stochastic trend with cointegrating rank r=1 and I(1) rank s=1. The normalized long-run polynomial cointegrating vector confirms that higher fiscal allocations (β₂=+0.00396) and improved academic performance (β₃=+0.00972) exert a statistically significant stabilizing effect on long-run educational retention [cite: 1]. Short-run dynamics modeled via a second-differenced Vector Error Correction Model (I(2)-VECM) demonstrate strong overall explanatory power across all equations (R²=84.42% for out-of-school acceleration; R²=70.59% for budgetary acceleration; R²=68.05% for pass rate acceleration). The polynomial speed-of-adjustment coefficient (α₁=−0.2051) establishes that 20.51% of short-run acceleration disequilibria is corrected annually toward long-run equilibrium. Lagged out-of-school growth displays significant negative self-correction (γ=−2.2319, p<0.001), while budgetary growth impacts out-of-school acceleration with a lag (γ=+0.0151, p<0.10), underscoring operational implementation lags [cite: 1]. Post-estimation diagnostic suite tests confirm that the estimated system is robust, non-spurious, and free from specification bias. Policy recommendations emphasize transitioning to a Medium-Term Expenditure Framework (MTEF), institutionalizing an automated Education Stabilization Fund, and linking spending expansions to quality benchmarks.