Statistical discrimination, also referred to as weblining, is a form of discrimination without prejudice. Data collection and analysis of variables like annual household income, purchase history, and internet activity result in individuals being grouped based on demographics and predicted behavior. Often expected characteristics are inaccurate and assign people to undesirable social affiliations. This act of grouping consumers into a generic profile has direct effects on the individual’s purchasing power. Additionally, customer rankings determine the services offered and the fees associated with those services.
Smart Marketing advocates generally make two claims. The first claim is that by collecting data from monitoring consumer activity online, companies can better serve the consumer’s interests. Secondly, sophisticated web analytics avoid marketing discrimination that dis-serve the individual consumer. While these claims appear to be on the level to end social prejudices, many believe more subtle acts of excluding classes of consumers from the marketplace will replace blatant discrimination.
On the one hand, marketing strategies run cost analysis to determine if generating advertisements to segments of a customer base will produce revenue enough to make this process worthwhile. On the other hand, individuals are reduced to a unit of analysis, and systematic discrimination is detrimental to equality. This behavioral marketing further divides societies and deepens the inequalities between social groups.
Redlining is the term used to describe discriminatory practices banks and insurance companies used to deny services to inner-city residents. Anti-redlining policies include the Fair Housing Act of 1968 and the Community Reinvestment Act of 1977. These policies stopped much of the overt discrimination practiced by institutions throughout the 1950s, 1960s, and 1970s. However, discrimination is almost impossible to detect when not blatant and seemingly done without prejudice.
Currently, no policies address the issues related to statistical discrimination because it is not blatant racially motivated treatment of a targeted group. The collection and analysis of big data are deemed how all future decisions will be based. Without recourse, as the amount of data increases, the risks of it being used as a form of discrimination increase.
What are your thoughts on this topic?