Cheating reduces the signaling value of examinations. It shifts the focus of teachers and students away from learning. Combating widespread cheating is difficult as students, teachers, and bureaucrats all benefit from high reported grades. We evaluate the impact of computer-based testing (CBT), a policy taken by the Indonesian government to reduce widespread cheating in high stakes national examination. Exploiting the phased roll-out of the program from 2015 to 2019, we find that test score declined dramatically, by 0.4 standard deviations, after the introduction of CBT. Schools with response patterns that indicate cheating prior to CBT adoption experienced a steeper decline. The effect is similar between schools with and without access to a computer lab, indicating that the reduction in the opportunity to cheat is the main reason for the test score decline. We find evidence of spillover from CBT, as schools that still used paper-based exam also cheated less and scored lower. This could be driven by CBT making cheating less socially permissible, or making the logistics of cheating harder due to a lower demand and supply for answer keys. Despite the fact that the examination remains high stakes, we find no evidence of any examination score improvements after three years of CBT implementation.
An earlier version of this paper is published in the RISE working paper series under the title From Cheating to Learning: An Evaluation of Fraud Prevention on National Exams in Indonesia.