Modeling Country-of-Origin Effects in the Car Market: Implications for Pricing
Various studies have applied hedonic price modeling to show that price variation among new cars can be explained by differences in key product characteristics such as horsepower, engine capacity, speed, and safety features (e.g., Reis and Santos Silva 200
- PDF / 177,643 Bytes
- 4 Pages / 595.276 x 790.866 pts Page_size
- 99 Downloads / 188 Views
e insights are of substantial value to marketing managers designing a pricing strategy, particularly when the consumer is confronted with product offerings from multiple COOs differing in their image. Therefore, the novelty of our study resides in both the managerial problem and the research methodology. In our context, COO designates not the country where the car was assembled, but the home country of the brand, which is more relevant from a consumer’s perspective. So, according to Jaffe and Nebenzahl (2006) we focus on brand origin effects, rather than COO effects. METHODOLOGY Although car characteristics are neither produced nor consumed in isolation, hedonic price models postulate that the price of a car reflects the bundle of embodied characteristics valued by some implicit or shadow prices. In empirical studies, these implicit characteristic prices are coefficients that relate prices and attributes in a regression framework. We usually use the semi-logarithmic specification
log Pi
a 0 ¦ E j xij u i
(1)
j
where Pi is the price of the
i product, a0 is a standard regression intercept, E j are regression coefficients, xij is the j
characteristic of the i product, and ui is the error term. As noted earlier, COO conveys information about various aspects of a product that are difficult to quantify, such as reputation and status. Additionally, COO serves as an extrinsic informational cue for consumers’ perceptions and evaluations of a product, and acts as a signal of product quality, influencing consumers’ perceptions of risk. These are particularly so in the car market where a brand’s origin plays a traditionally important role in purchase decisions. To investigate formally the above hypotheses, first, the standard hedonic regression model (1) is extended to include COO effects. Notice that model (1) has a standard regression intercept ( D 0 ) and as many beta coefficients ( E j ) as the number of product attributes. More specifically, let ( D c ) represent the effect of COO c. Then, the hedonic model can be written as
log Pi
D c ¦ E j xij ui
(2)
j
where each COO has its own intercept. Subsequently, we design and implement a disaggregate hedonic price model in which both attribute coefficients and COO effects are allowed to vary over car-type segments. More specifically, let D cs represent the c COO effect within the s cartype segment, and
E js
represent implicit or shadow price of the j characteristic within the s car-type segment. Then, the
aggregate specification (2) can be expressed using the following disaggregate equation.
log Pi
D cs ¦ E js xij ui
(3)
j
Model (3) allows a different constant term (i.e., COO effect) and slope coefficient (i.e., implicit characteristic price) for each car-type segment. Both hedonic regression models (2) and (3) incorporate some type of COO-specific intercepts and are, in essence, panel data models that can be implemented by the fixed–effects technique (Greene 2002). Our estimation dataset includes prices and characteristics of all new passenger car
Data Loading...
