Following the burst of the dot-com boom, new information technologies were created and revolutionized how companies tailor prices to individual consumers. In an influential study conducted on car retailing titled, "Consumer Information and Price Discrimination: Does the Internet Affect the Pricing of New Cars to Women and Minorities?”, economist Fiona Morton and her colleagues seek to answer one major question: “whether the increased difficulty in accurately assessing a consumer's willingness to pay on the Internet and consumer's ease in finding information affects race and gender discrimination in car retailing” Scott Morton et al. (2003).
To achieve this, the authors collect their data from the MRI, a major supplier of marketing research information, spanning from January 1, 1999 to February 1, 1999. MRI's data is uniquely suited for several reasons. First, the dataset contains approximately 700,000 individual purchases by consumers who reside in communities with varying demographics. Second, the data offers timely, high frequencies on the make, model, trim level of car, and financing in over 352 U.S. dealerships. And third, it allows us to extract as the profitability of the car as well as the consumer to the dealership.
Scott Morton et al. (2003) find that internet provides car buyers a certain degree of anonymity to consumers, thereby “eliminating variation in new car prices that arise from individuals from individual characteristics associated with race and ethnicity”. When they do not control for demographics, their study reveals that African-American and Hispanic consumers who use the Internet pay 2.2% less for their car than their counterparts who purchase the car offline. Overall, the results imply that the Internet benefits those individuals whose observable traits, such as income, education and search costs, operate a disadvantage in the bargaining process.
One important limitation in addressing this question is that it fails identify consumers who may have used an online referral service other than Autobytel.com. That is, their study focuses on only one internet referral service, autobytel.com, and as such cannot obtain a pure reading on the effects of different search or purchasing activities. This may potentially bias results, causing the study to suffer from lack of internal validity. With respect to measures of race and gender, many cars are officially purchased by two people, however, the dataset only records the first name on this registration, making it impossible to tell who bargained for the car. And lastly since the dataset was obtained during the dot-com boom, more recent empirical research is needed to understand how current price discrimination is evolving and how precisely consumer personal information is used.
From a policy standpoint, Scott Morton et al. (2003) paper contributes to the body of literature by raising questions regarding the extent to which “fairness” prevails in the offline car retail industry. More specifically, the results better inform us on how dealer practices, such as race premiums, result from “disparate impact”, rather than “disparate treatment”. In addition, the paper places greater concern on the welfare of those affected by the “digital divide”, that is the very people that could obtain the highest benefit to internet access more often than not, have less access to it. As such, this paper provides a good starting point for dialogue and policy reevaluation on subsidizing internet service to those who are unable to afford it or find it difficult to accessing.
The paper models autobytel.com's differential pricing method in the following way: the car prices are individually negotiated, thereby presenting an opportunity for price discrimination in the market. That same car will then sell for different prices because supply and demand changes over time and consumers differ in preferences. Under this model, minorities and women seem to pay higher prices because on average they face higher search costs. The advent of the internet gave rise to Autobytel.com, which in turn reduced these same search costs and provided more reliable, quick information to consumers, decreasing the effect of information asymmetry. More subtly, also, this type of site gives consumers leverage, as now they can easily arbitrage across products or markets, rendering them more elastic consumers. In a car market where prices are negotiated (rather than fixed), this will drive down the consumer surplus for minority groups. Thus, the internet acts as a medium to modify this price discrimination.
In conclusion, the paper provides a solid empirical argument for why certain individuals operate at a disadvantage in the bargaining process and thus are more likely to use the internet for these transactions. One noteworthy feature of this paper is the timing of the data. Following the burst of the dot-com boom in 2000, new information technologies were created and revolutionized the ways through which companies tailored prices to individual consumers in the 21st century. If there was ever a case of how and why this happened, this paper, in my opinion, makes a compelling one.
...(download the rest of the essay above)