Do consumers value innovation in agro-food products? A demand system approach

The objective of this study is to calculate attribute price elasticities in order to assess the potential profitability of innovation investments Our hypothesis is that the more inelastic a specific attribute is and the lower degree of substitutability exists among competing attributes, firms could increase prices with a relatively small impact on the quantities demanded and could reduce the Return on Investment (ROI) period. The milk sector has been chosen as a case study as it is one of the most innovative food sectors worldwide. In the milk sector, most of the innovation has been addressed towards two main attributes: health (fat content) and functionality (enriched milk with omega-3, calcium, vitamins…). Additionally, the brand is a key attribute for consumers to make milk choices. The methodological approach is based on the estimation of a demand System. More precisely, two different demand systems have been estimated. The first includes four milk categories: whole, milk, semi-skimmed milk, skimmed milk and enriched milk. In the second, each milk category has been divided into two subcategories depending on if the milk has been sold under a manufacturer brand or under a private label. In the latter case, we are interested not only on the substitution effect among milk attributes but also among brands. In both cases, an incomplete censored Exact Affine Stone Index (EASI) demand system has been estimated as its flexible enough to consider alternative functional forms for the Engle curves. Quality-adjusted prices, endogeneity and heteroskedasticity issue have been considered. The data used in this study relies on homescan data set provided from the KANTAR WORLDPANEL. Only those households reporting milk purchases during, at least, 40 weeks in 2012 have been considered. Our final sample was 838 households. Consumption and expenditure have been aggregated along the year 2012. Results from this study suggest the following. All expenditure elasticities are positive and lower than unity, indicating that fresh milk categories can be considered as necessities. In both models, the expenditure elasticity of the enriched milk is the highest. When we differentiate by brands, the expenditure elasticity of the manufacturer brand is higher than that of the private label for each milk category. Concerning the own price elasticities of milk categories, the demand for all milk categories are price inelastic. Results also show that the functional attribute is the most price inelastic indicating that firms could increase prices as the quantity purchased did not vary significantly. However, when we consider brands, this result has to be interpreted with caution as it only refers to the private label enriched milk. Any price increase of the manufacturer brands in the case of the enriched milk generates a substitution effect towards the private label