Age Effects and Information Shocks: A Study of the Impact of Education Policy on Student Outcomes

This thesis studies the impacts of school entry policy and information revelation on student outcomes using a sample of students from the province of British Columbia (BC), Canada. The questions examined by the first two essays arise from a policy used by many industrialized countries, whereby students born within a 1-year time span all begin school at the same time. This policy creates large differences in age among students in the same class, which are thought to affect their academic performance along a number of dimensions. In the first essay, I contribute to the literature by establishing the persistence in test score differentials among students in the same class who differ in age. I show that in grade 4 older students outperform younger students by a large margin in numeracy, reading and writing, an effect that persists to a lesser magnitude until grade 10. The persistence is strongest for the writing skill, and it is also much stronger for girls than for boys. The strength of the test score differential in grade 10 suggests that the effects of age could have more lasting effects on cognitive and labour market outcomes. In the second essay, I take a closer look at how age affects outcomes, by disentangling the entry age effect from the test age effect. Nearly all studies in this literature interpret age-related differences in student outcomes as the result of entry age, but because students who enter later are also older at every stage in compulsory schooling, the entry age effect has not been separated from the test age effect. Using a set of students entering school at the time of BC's dual entry experiment, I show that test age is largely responsible for age-related differences in the probability of repeating grade 3, and entry age is largely responsible for age-related differences in grade 10 numeracy and reading scores. I show further that having an extra year of schooling reduces the likelihood that a student repeats grade 3, but has a negligible impact on grade 10 test scores. Both the entry age and test age effects are stronger for boys than they are for girls. The final essay examines whether school choices change when parents are exposed to a new source of information on school quality. I model the effect of new information on choices using a simple expected utility framework and show that parents will use the new information to make different choices if they do not perceive it to be too noisy and if they have poor prior information on school quality. Furthermore, they make increasing use of the new information as more observations become available, since it becomes a more accurate predictor of true quality. Using the sudden release of BC's new standardized testing regime, I then study whether there is empirical support for the model. I show that the likelihood of switching out of a school increases when a school performs worse on the test, and that enrollment into kindergarten responds positively to increases in test scores. The response becomes stronger when more test score observations are available. Finally, I show variance in the response among parents living in less-educated neighbourhoods and among those who do not speak English at home, suggesting that prior information does play a role in the information use.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.