Economic approach to advertising and consumer choice - comparison of advertising elasticities in the smartphone handset market between countries, media and advertisers

摘要

Purpose of the StudyIn my master's thesis I attempt to give the economist's view of advertising and consumer choice. I represent some reasons why economists are interested in researching them. I also do an empirical research to demonstrate the response of sales to advertising. The empirical part of my thesis concentrates on the smartphone handset market which is a typical example of a market with infrequently purchased goods with the frequent introductions of new products. Such market suits well for the discrete choice modeling framework. I concentrate on high-end smartphones the price of which has exceeded €300 at some phase of the life cycle. These are more interesting from the economist's point of view, because they are more differentiated, innovative and less prone to the price competition than the cheaper ones. Non-price competition in the market of differentiated goods can be considered as an umbrella for the paper. MethodologyThe thesis starts with the literature review about economic, both theoretical and empirical, research concerning advertising. I introduce the potential effects of advertising on the market structure and on the utility of a consumer. I cover some consumer choice behavior models and concentrate on logit model which is applied in the empirical part of the paper. In the empirical part of the paper, I compute the advertising elasticities of market share in nine different markets (China, Germany, India, Indonesia, Italy, Russia, Saudi Arabia, United Kingdom, USA), by different media (cinema, online, outdoor, print, tv) and by advertisers (manufacturer, operator/retailer). The model does not fit for computing substitution patterns between goods but it is a simple way of comparison market share responses to advertising between countries, media and advertisers. I use the unbalanced panel data of sales volumes and advertising investments of 20 products representing various brands from January 2011 until November 2012. I run different regressions (pooled OLS, brand fixed effects and product fixed effects) with advertising as a periodical investment and as a depreciating and accumulating stock.FindingsAdvertising elasticities differed somewhat between different regressions, but China and United Kingdom got the highest elasticities in most of the regressions whereas Germany and USA got the lowest. By media, the highest elasticities on the whole sample level got print and tv. That result varied somewhat across countries. Some media got implausible or non-significant coefficients, which is partly explained by the noise and partly by the fact that absolute advertising investments do not capture the advertising exposure as clearly as advertising measured in GRP. The advertising by advertisers yielded neither significant nor plausible results, which is at least partially explained by the measurement error in the data.