CAPIRE: Modelling the Impact of Teacher Strikes and Inflation on Student Trajectories in Engineering Education

TL;DR

This paper extends the CAPIRE framework to analyze how macro shocks like teacher strikes and inflation influence engineering students' dropout trajectories, combining empirical data analysis with agent-based simulations.

Contribution

It introduces a macro-shock module into CAPIRE, demonstrating how macroeconomic factors impact student dropout patterns and integrating these insights into predictive and simulation models.

Findings

Macro features improve early-semester dropout prediction.

Strike exposure increases dropout risk two to three semesters later.

Combined shocks cause non-linear increases in dropout levels.

Abstract

This study extends the CAPIRE framework with a macro-shock module to analyse the impact of teacher strikes and inflation on student trajectories in engineering education. Using data from 1,343 students across 15 cohorts (2004-2019) in a public engineering faculty in Argentina, we construct a leak-aware, multilevel feature set that incorporates national inflation indicators, lagged exposure to teacher strikes, and interaction terms between macro shocks and curriculum friction. Random Forest models with cohort-based validation demonstrate that macro features provide stable, non-trivial gains in early-semester dropout prediction (improvement in Macro F1 from 0.73 to 0.78), with inflation volatility at entry and a strike-weighted basic-cycle friction index amongst the most influential variables. Lag analysis reveals that strike exposure exerts its strongest association with dropout two to three semesters after the disruption (OR = 2.34), and that effects are concentrated in early, high-friction semesters. We then embed these empirical patterns into an agent-based model, defining scenarios for inflation-only, strikes-only, and combined crisis. Simulations reproduce three stylised facts: delayed strike effects, basic-cycle vulnerability, and non-linear amplification when inflation and strikes co-occur, with combined shocks generating dropout levels 18-23% higher than the sum of individual effects. We argue that teacher strikes and inflation operate as structurally mediated educational disruptors, acting through curriculum design and financial resilience rather than as isolated events. The framework contributes to multilevel dropout theory by demonstrating how macro-level shocks propagate through institutional structures to shape individual trajectories and provides empirically grounded tools for scenario planning in macroeconomically unstable contexts.