A five-year hedonic price breakdown for desktop personal computer attributes in Brazil.
| Autor | Fouto, Nuno Manoel Martins Dias |
| Cargo | Report |
Introduction
The aim of this article is to study the value of the different attributes that compose the market price of desktop personal computers. We employed hedonic regression to obtain the specific weight of each main pricing characteristic. The relevant attributes associated to a certain product may be related to its physical characteristics, complementary services or products, the manner and conditions under which is sold, subjective image aspects, etc. Determining the relative importance of these characteristics allows companies to define their strategic position more adequately, bearing in mind the possibilities of meeting demand. In this analytical approach, products are seen from several dimensions, transcending the traditional approach strictly tied to price and quantity variables.
Hedonic pricing methods are reasonably well known to econometric studies, although they receive little mention in marketing research textbooks. Hedonic analysis uses the prices practiced in product transactions as a dependent variable, and corresponding attributes as independent variables.
The desktop personal computer market is supplied by companies that offer heterogeneous, vertically-differentiated products. Personal computers first reached the market in the mid-1970s. The industry grew quickly and became dominated by a small number of large-scale companies. In the 1990s, however, a large number of smaller companies entered the market, making the industry highly competitive.
A desktop personal computer may be identified according to characteristics such as: processing performance; processor brand, hard drive and random memory capacity and access interface, whether it has CD and DVD drives, screen size and technology and display adapter technology, expansion devices, communication devices, the number of input/output ports, main dimensions, sound devices, security features, BIOS, operating system and additional software, warranty and environmental specifications.
This article has been organized into five sections. The first presents a review of the literature regarding the evolution of the hedonic pricing concept and its applications. The second section describes the analytical model employed in evaluating the attributes of personal computers. Sections three and four, respectively, present the data and the study's results. The final part, section five, presents general conclusions, an outline of the study's limitations and possible extensions.
REVIEW OF THE LITERATURE
Microeconomic theory bases its analysis of individuals' choice processes on the fact that the consumption of goods and services provides varying levels of satisfaction. The expression hedonic analysis comes from this perspective. Etymologically, the word hedonic is derived from the Greek hedonikos, meaning pleasure. Such a designation therefore calls to mind the idea of usefulness or satisfaction inherent to the attributes that compose the offer of a good or service.
The method known as hedonic pricing was introduced in the mid-20th century to handle product quality issues. Only more recently, however--in the 1960s--did it gain notoriety, when it was used in the United States Consumer Price Index [CPI] (Hulten, 2002). Schultze and Mackie (2002) considered hedonics to be "the most promising technique for explicitly adjusting observed prices to account for changing product quality" (p. 122). In price indexes, hedonic regressions are used to estimate the value of specific bundles of individual characteristics that, when considered as a single set, form goods or services.
By estimating hedonic functions, where prices are broken down into their constituent attributes, one may therefore separate pure price changes from changes in the quality of the attributes considered. The coefficients of characteristic or attribute variables in hedonic equations represent average marginal implicit prices for each relevant attribute/characteristic (Bartik, 1987; Epple, 1987; Rosen, 1974). One may say that properly valuated attributes denote the consumers' structure of preferences by associating price variations to the type and intensity of the main characteristics (Freeman, 1993).
Several studies have employed this analysis method. The first of these studies was that on the vegetables market conducted by Frederick V. Waugh and mentioned by Berndt (1991). Court (1939) later pioneered the use of the adjective hedonic, suggesting the use of the coefficients of regressing automobile prices on their characteristics in the construction of price indexes. Houthakker (1952), in turn, introduced the concept of quality as a set of distinct variables to be considered concomitantly with the quantities consumed. He defines a quality price considering the price differential according to different attribute combinations. Lancaster (1966, 1971) and Gorman (1980) then adapted the concept of a utility map from a new analytical perspective. In these studies, alternatively to the traditional view of consumer theory, where individuals choose between quantities of products, choices are based on attributes and their respective intensities. Griliches, however, was the first to point out that interesting studies could be accomplished with hedonic pricing models. Building from the ideas put forth by Court (1939), Griliches (1961) proposed the use of hedonic pricing as a way to attenuate the issue of new product launches when constructing price indexes. As new products frequently offer more characteristics desired by consumers, the difference between their prices and the prices of their older counterparts cannot be attributed solely to inflation for the periods before and after the entry of the new products into the market. Another of Griliches' lines of research concerned the use of production and input indexes to measure technological change. Economic models of the time showed most production growth to be a result of technological evolution, measured by the residues of their equations. The relative importance of these residues led him once more towards hedonic regression, in a study of the problem of measuring change in quality, carried out for the National Board of Economic Research [NBER] in 1961 (Griliches, 1971). Court and Griliches suggested the estimation of a surface that would relate prices to characteristics. This estimated surface would be employed in obtaining estimates of product prices adjusted, according to their quantities, to a set of characteristics. This would allow estimates of price changes in differentiated products, adjusted to quality, to be obtained.
Hedonic price functions may therefore be seen as empirical representations of the relationship between prices and characteristics of goods sold in markets whose products are relatively differentiated. The term hedonic method means that a hedonic function is applied to economic measurement,
P = h(c) (1)
where P represents, in a cross section of prices of goods and services, one price [p.sub.ijt] for each model or variety 'j' of the good or service 'i' available at a time 't'. The matrix c has one characteristics row for each model (Triplett, 1990).
A reasonable number of papers on hedonic pricing followed Griliches' work, with a theoretical focus on examining the relationships between price and characteristics: from the demand point of view (Muellbauer, 1974); from the supply point of view (Ohta, 1975); or generated by equilibrium in differentiated product markets (Anderson, Palma, & Thisse, 1989; Berry, Levinsohn, & Pakes, 1995; Feenstra, 1995; Rosen, 1974).
The model published by Rosen (1974) is considered to have been the first to theoretically relate the hedonic function to the utility function and the production function. Rosen's paper elicited several others, which advanced theoretical discussion of important issues, such as the identification problem (Bajari & Benkard, 2001; Bartik, 1987; Brown & Rosen, 1982; Epple, 1987; Kahn & Lang, 1988).
According to Rosen (1974), characteristics are the real arguments of the utility function. Therefore:
Q = Q(c, Z) (2)
where Q is the utility (or scalar production) and Z is a vector of other homogeneous goods (or productive inputs). For the sake of simplicity, Triplett...
Para continuar a ler
PEÇA SUA AVALIAÇÃOCOPYRIGHT GALE, Cengage Learning. All rights reserved.
