Program committee

With growing concern about childhood obesity and associated health risks, several countries are considering banning food advertising directed to children. In 1980, the Canadian province of Quebec imposed a ban on all TV advertising to children under the age of 13. In this paper, we look at whether the advertising ban affected consumer food choice in Quebec. To the best of our knowledge this will be the first study to explore the effect of the Quebec ban on consumption. Using data from the Canadian household expenditure survey and Canada Foodex surveys from 1984 to 1992, we ask whether expenditure at fast food restaurants and on junk food is lower in those groups affected by the ban than those without. Following Goldberg (1990) we control for mother tongue as native English speakers have access to more sources of media from outside the province. Unlike Goldberg, we use a difference-in-difference (DiD) and match estimation methodology across several dimensions, holding family characteristics such as income, education, and immigrant status constant. First, we test whether consumption of traditionally highly advertised fast food and breakfast cereal consumption among Frenchspeaking families with kids in Quebec differs more markedly than consumption of English-speaking families in Quebec compared to their English and French-speaking counterparts in Ontario. Under the assumption that other non-English speaking families will also have less access to media from outside the province, we also compare consumption of those families whose first language is neither French nor English to Anglophones in the two provinces. Further, we anticipate that the ban primarily affects the expenditure of families with children, and estimate the effect of the ban by testing the difference between households with children and those without in Quebec versus Ontario. Our second approach is to estimate the treatment effect by matching households with similar characteristics across Ontario and Quebec. We find consistent results for fast food across all types of difference-in-difference and matching estimation. Results for breakfast cereal are not as robust using DiD technique, the difference in expenditure being significant only in the difference between families whose mother tongue is neither French nor English compared to English-speakers in Quebec and Ontario. But using matching technique the result turns out to be significant but at magnitude of the difference is much smaller than fast-food. Although we cannot test the effect of the ban directly due to lack of preand postban expenditure data, we provide myriad of evidence by comparing expenditure cross sectionally and over time indicating that the ban had an effect on the expenditure on highly-advertised products like fast-food and breakfast cereals.