ISER Discussion Paper, Institute of Social and Economic Research, Osaka University 845
Using a natural experiment designed by the Italian national test administrator (INVALSI) to monitor test procedures in Italian primary schools, this paper shows that the presence of an external examiner who monitors test procedures has both a direct and an indirect effect on the measured performance of monitored classes and schools. The direct effect is the difference in the test performance between classes of the same school with and without external examiners. The indirect effect is instead the difference in performance between un-monitored classes in a school with an external examiner and un-monitored classes in schools without external monitoring. We find that having an external examiner in the class reduces the proportion of correct answers by 5.5 to 8.6 percent compared to classes in schools with no external monitor. Comparing results of un-monitored classes across schools with and without external examiners, we estimate indirect effects that range between 1.2 and 1.9 percent. The size of the overall effect of external supervision varies significantly across regions and it is higher in Southern Italy.