In this paper, we show how results in state administered university entrance standardized exams in Portugal are sensitive to the electoral cycle. Using individual data for all senior high school students taking national exams between the years 2004 and 2018, we reveal that whenever exams occur in years of upcoming legislative elections, the average senior high-school student grade increases by 1.17 and 0.63—on a 0 to 20 scale, in the core disciplines of Mathematics and Portuguese, respectively. Furthermore, the probability of passing the exams, that is, of getting more than 9.5 out of 20, increases by ten and eight percentage points in those same exams. Neither the quality of students nor the leniency of grading in these years explain this rise in pre-election exam grades. Robustness tests using early legislative elections, held, exceptionally, before national exams, and therefore not anticipated by those who arguably manipulate the exams, show that there was no inflation of exam scores in those years. Furthermore, we highlight that exam inflation can be a powerful electoral tool, with a unit increase in the municipality average score in Mathematics and Portuguese exams associated with a 1.46 percentage point rise in the vote share of the sitting government.

在本文中，我们展示了葡萄牙国家管理的大学入学标准考试的结果如何对选举周期敏感。使用2004年至2018年间所有参加国家考试的所有高中学生的个人数据，我们发现，在即将举行的立法选举的几年中，每当进行考试时，高中学生的平均成绩就会分别提高1.17和0.63。 20等级，分别位于数学和葡萄牙语的核心学科。此外，通过这些考试的概率，即20项考试中超过9.5项的概率，在相同的考试中增加了10个百分点和8个百分点。这些年来，学生的素质和宽大的成绩都不能解释选举前考试成绩的上升。使用早期立法选举进行的稳健性测试，参加国家考试之前，因此可以说是那些可以操纵考试的人所无法预料的，​​那表明那几年考试分数没有增加。此外，我们强调考试通货膨胀可能是一种有力的选举工具，数学和葡萄牙语考试的市政平均得分每增加一个单位，而现任政府的投票份额就会增加1.46个百分点。