How much do consumers know about the quality of products? Evidence from the diaper market
- PDF / 881,693 Bytes
- 29 Pages / 439.37 x 666.142 pts Page_size
- 47 Downloads / 234 Views
How much do consumers know about the quality of products? Evidence from the diaper market Andrew T. Ching1 · Tülin Erdem2 · Michael P. Keane3 Received: 21 June 2018 / Revised: 7 July 2019 / Accepted: 2 September 2019 © Japanese Economic Association 2020
Abstract To measure the extent of incomplete information about brand qualities faced by consumers, recent research in marketing and economics has extended traditional static choice models to explicitly allow for consumer learning. These models tend to be complicated and make stringent assumptions such as Bayesian updating. In this paper, we provide a simpler alternative method to measure how much consumers know about the quality of quasi-durable products. Our key insight is that for products that depreciate over time and require repeated purchases, individuals’ observed inter-purchase spells provide another measure of brand qualities in terms of durability. This is simply because the higher the durability, the longer a product can last in general, and hence its observed inter-purchase spells should also be longer. Based on this argument, we propose an empirical framework to estimate both the perceived brand quality (based on revealed preference data) and brand durability (based on brand-specific inter-purchase spells) and apply it to a scanner panel dataset for diapers. Our estimates allow us to compare these two measures of qualities and infer the extent of incomplete information faced by parents. With our results, we can address questions such as: Do parents make the right choice in the diapers category? Can they save some money by switching from a national brand to a store brand or the other way around? How much savings can they get? Keywords Incomplete information · Product quality · Efficiency unit · Quasidurable goods · Brand choice · Inter-purchase spells · Inventory
* Andrew T. Ching [email protected] Extended author information available on the last page of the article
13
Vol.:(0123456789)
The Japanese Economic Review
JEL Classification C12 · C33 · C35 · C41 · D12 · D80 · E21 · L68 · M21 · M31
1 Introduction The pioneering work by McFadden (1974) and Guadagni and Little (1983) has led to not only a large academic literature in studying brand choice, but also many marketing applications in practice (Guadagni and Little 2008). The majority of applied brand choice studies estimate the quality of brands by assuming that consumers have complete information and use revealed preference data to recover the vertically differentiated quality levels. Nevertheless, it is common that consumers do not know the quality of all brands available to them and just choose a small subset of brands that they are familiar with, even though the actual quality of other brands could be higher. For instance, even though some store brands have very high quality, many consumers may fail to try them due to uncertainty about their quality. If we ignore this potential incomplete information problem, estimating static brand choice models may lead to biased inference in the tr
Data Loading...
