There'd still be a correlation between spending and academic scores regardless of the actual causative mechanism.