School Effectiveness, Teaching Effectiveness and School Climate: A Critical Review of Leadership as a Mediated and Contested Influence

School effectiveness, teaching effectiveness and school climate developed as three partly separate research traditions, each with its own conceptual vocabulary, preferred designs and characteristic outcome measures. Leadership is routinely invoked as the factor that binds them together, yet the empirical warrant for that claim varies sharply according to how leadership is measured and how the counterfactual is constructed. This review critically examines the evidence linking school leadership to teaching quality, organisational conditions and student outcomes, with particular attention to the persistent gap between estimates derived from perceptual survey research and estimates derived from administrative-data designs. Literature was identified through open scholarly search platforms, bibliographic metadata registries and institutional repositories, supplemented by backward and forward citation searching, and appraised for design adequacy, measurement quality, treatment of confounding and transparency. Four conclusions emerge with reasonable confidence. Leadership effects on student achievement are predominantly indirect, operating through teaching quality, staffing decisions and organisational conditions. Estimates from cross-sectional perceptual studies are systematically larger than those from designs that exploit leader mobility, and the difference is better explained by measurement and identification properties than by substantive disagreement. School climate functions simultaneously as a mediator, an outcome and a measurement artefact, and much of the literature does not distinguish these roles. The evidence base is geographically and methodologically concentrated, with a strong reliance on principal or teacher self-report from a small number of high-income systems and, increasingly, from a small number of middle-income systems using the same instruments. Priorities for future work include registered longitudinal designs capable of separating leader effects from school effects, measurement work that establishes whether climate instruments capture a school-level property rather than aggregated individual perception, and contextually grounded studies in low-resource systems where the transferability of dominant leadership models remains untested. Claims that leadership is second only to teaching in its influence on learning should be treated as a plausible ordering rather than a settled quantitative finding.