Assessing the redistributive impact of higher education tuition fees reforms: the case of Québec

Abstract

In this article, we analyse the redistributive impact of a recent reform of tuition fees in Québec. We adapt Duclos et al.’s (2005) methodology to a generalized Lorenz framework. Many policy analysts argued that maintaining low higher education tuition fees is regressive. We take a look at the empirical validity of this argument using data from Statistics Canada's Survey of Labor and Income Dynamics. We show the importance of using data to validate this argument. The results obtained allow for the conclusion that this redistributive argument is empirically not verified for the province of Québec.

Acknowledgements

We thank Mark Taylor and an anonymous referee for their comments and Mathieu Audet for assistance with the data. We also thank the Regroupement des étudiantes et étudiants de maîtrise, de diplôme et de doctorat de l'Université de Sherbrooke (REMDUS) and the Fédération étudiante universitaire du Québec (FEUQ) for having asked the first author to analyse the question.

Notes

¹See Johannes et al. (2006) for an example of this. This appears to be true even in the context of a developing country.

²See Johannes et al. (2006) for an opposite result in the context of a developing country.

³Following King (1983), assume that y be pre-reform real income assessed using pre-reform prices as reference prices. In this context y is a money-metric indicator of welfare: in King's terminology, it is also called ‘equivalent’ income. As noted by King, the concept of equivalent income function is also used by McKenzie (1956), Samuelson (1974) and Varian (1980).

⁴By the envelope theorem, this is regardless of whether the agent changes his behaviour following the reform. This is because income here is a money-metric indictor of welfare and not nominal income.

⁵This assumption is consistent with the findings of Vermaeten et al. (1994) who show that the overall tax incidence in Canada is proportional to income.

⁶It is, however, possible to assume that there is a subset of indices with high inequality aversion for which this reform will be deemed as regressive by all indices in this subset. See Makdissi and Mussard (2008).