The first model explains 37% of the variance in test score performance, highlight-ing that the forces beyond the control of the sample organizations account for a sizeable portion of their eventual achievements. This confirms the arguments of organization theorists and management scholars about the salience of organizational environments, as well as those of policy analysts who emphasise that the provision of education services often occurs within a fraught and challenging task environ-ment (e.g. Jencks and Phillips 1998; Meier and Bohte 2003). For the control variables, more teacher experience and new teachers exert a negative impact on performance, indicating that the benefits of practical experience “at the chalkface” for education outcomes may eventually turn negative as staff potentially become too set in their ways and resistant to changing their practices. However, surprisingly, teacher salary and class size do not have a significant impact on test scores when controlling for the organizational environment.