Guillaume Bérard, Edward Levavasseur

phd seminar

Guillaume Bérard, Edward Levavasseur


IBD Salle 16

Îlot Bernard du Bois - Salle 16

5-9 boulevard Maurice Bourdet
13001 Marseille

Tuesday, February 19, 2019| 12:30pm to 2:00pm

Océane Piétri: oceane.pietri[at]
Morgan Raux: morgan.raux[at]
Laura Sénécal: laura.senecal[at]


Guillaume Bérard
A real estate tax all in one: yield, progressivity and feasibility

We are studying the feasibility of a renovated large property tax on the housing wealth of households, that would replace all existing taxes on real estate, and in particular : the property taxes, the IFI (wealth tax), the real estate transfer taxes and the real estate income received by landlords. The value of the net housing wealth would be taxed at the rate of 1% up to €1.3 million (IFI’s threshold) and up to 1.5% beyond. Tax revenue would be shared between the state and the local governments with an equalization fund. After presenting the reasons that legitimize such a reform proposal, we make a first assessment of the tax yield and of the profile of the tax burden according to income, family size and age, using the Household Wealth Survey 2014-2015. The inverse correlation between income and real estate assets for a part of the population makes the goal of creating it a large yield tax relatively difficult, even if the expected tax revenue is equivalent to the lost revenue from the taxes replaced.

Edward Levavasseur
Do educational systems limit social reproduction?

This paper seeks to assess how systems of compulsory education are able to limit social reproduction. Our data shows that inequality of opportunity increases in all countries, as children from different socio-economic backgrounds go from grade 4 to grade 10. This paper considers that educational systems with the smallest increase in inequality of opportunity between grades 4 and 10, are those which limit social reproduction the most. Thus all countries are ranked based on the magnitude of this increase. However, due to educational systems having a non-linear effects on their pupil’s cognitive skills, the use of a standard difference-in-differences (DID) approach would lead to biased results. Instead, we construct non-parametric counterfactual distributions of children from one educational system, had they been placed in another educational system. This paper's approach relies on a variant of Athey and Imbens' change-in-changes (CIC), adding conditionality on family background. Relying on TIMSS 2003 and PISA 2009 data, educational systems are then ranked using dominance criteria on the extent to which inequality of opportunity increases, as children climb the educational grade years.