par Gothelf, Nicolas ;Herbaux, Denis ;Verardi, Vincenzo
Référence The Economics of Art and Culture (2009-07-11: Steyr, Autriche)
Publication Non publié, s.d.
Poster de conférence
Résumé : This paper focuses on cultural goods such as entertainment goods, especially theme parks. We have built a dataset regrouping the largest parks in Europe and looked at the relation between the costs and the quality of a theme park, thereby trying to estimate if a theme park deserves its attendance with respect to its characteristics such as quality and price. We tackle the issue that both the price and the attendance are highly correlated with each other guessing that in fact both the price and the attendance level can be explained by the sum of the characteristics of these parks. So we run two separate hedonic robust regressions on price and attendance to find out if there are outliers in terms of these two variables. To do this, we identify outliers in the space of explanatory variables using the Mahalanobis distance. From the detection, we downweight the outliers and the system we regress allows for the possibility of a link between both equations (SURE). Moreover, we propose a new typology for this cultural good.
This paper focuses on cultural goods such as entertainment goods, especially theme parks. We have built a dataset regrouping the largest parks in Europe and looked at the relation between the costs and the quality of a theme park, thereby trying to estimate if a theme park deserves its attendance with respect to its characteristics such as quality and price. We tackle the issue that both the price and the attendance are highly correlated with each other guessing that in fact both the price and the attendance level can be explained by the sum of the characteristics of these parks. So we run two separate hedonic robust regressions on price and attendance to find out if there are outliers in terms of these two variables. To do this, we identify outliers in the space of explanatory variables using the Mahalanobis distance. From the detection, we downweight the outliers and the system we regress allows for the possibility of a link between both equations (SURE). Moreover, we propose a new typology for this cultural good.