Longitudinal Statistical Analysis of Child Schooling Outcomes in North West Nigeria Using LSMS Panel Data

Child schooling outcomes in North West Nigeria exhibit significant variation across time and households, and longitudinal statistical analysis using the World Bank Living Standards Measurement Study panel data provides superior causal insight into determinants compared to cross-sectional approaches. This study conducts a longitudinal statistical analysis of child schooling outcomes in North West Nigeria using the 2015 to 2016 and 2018 to 2019 LSMS-ISA panel waves covering Kano, Katsina, and Kebbi States. The panel covers 1,840 children aged 6 to 17 years observed across both waves. Outcomes analysed include school enrolment, school attendance rate, and grade progression. Fixed effects logistic regression for enrolment and random effects Tobit regression for attendance rate controlled for unobserved household heterogeneity. Correlated random effects models applied Mundlak correction for time-varying regressors. School enrolment increased from 52 to 61 percent over the panel period. Fixed effects logistic regression showed that household income shocks (OR 0.54, p = 0.003), father's educational attainment (OR 1.62, p < 0.001), and school distance (OR 0.71 per km, p = 0.008) were significant within-household determinants of enrolment change. Girls showed a significantly larger enrolment benefit from new school construction (OR 2.41) than boys (OR 1.38, interaction p = 0.014). Income shocks reduced attendance by 11 percentage points by Tobit estimation. The study provides longitudinal causal evidence for North West Nigeria schooling policy and recommends conditional cash transfer programmes linked to school attendance. Keywords: longitudinal analysis, child schooling, LSMS panel data, North West Nigeria, fixed effects