Audit studies were developed by fair housing organizations in the late 1970’slooking to uncover discrimination inhousing markets.
The technique makes it possible to gauge discrimination in the actand for this reason has become a popular tool of researchers in the field. The audittechnique attempts to overcome thesignificant problem of confounding that plagued earlier discrimination research bystriving to eliminate as much non-racialdissimilarity between subjects as possible.
The weak points of an audit studydepend largely upon the type of contact thatis being utilized. One general problem isthat by nature, audit studies are not double- blind, so auditors may introduceidiosyncratic qualities to the experiment thatfavor a certain endpoint.
Further,individual discrimination observed in anaudit study may not serve well as anindicator of market discrimination. These problems are addressable, however, throughcareful research and the utilization of a wide body of audit data. Generally, however,audit studies are characterized as quasi-experimental and give the researcher a gooddeal of power in assessing the types of treatment that different races face in themarket.
Since its first use in ’79, the auditmethodology has been widely employed inan attempt to discern broadly the presence of discrimination. In 1987, George Galster collected fifty separate audit studies fromdifferent housing markets around the UnitedStates. In his analysis of the data, Galster used a conservative measure of racialdiscrimination and concluded that AfricanAmericans faced about a 50% chance of discrimination in the U.S. housing market.
A more recent study of car dealerships also detected the presence of discrimination in the market, this timeagainst both African Americans and women.The data, collected by Ayres and Siegelmanfrom their study of over 300 paired audits atnew car dealers, showed, at a statisticallysignificant level, that African Americans andwomen were quoted higher final prices thanwhite men. Based upon these findings, theresearchers attempted to identify the reasonsthat dealers would be inclined todiscriminate. Their conclusion cited the possible presence of non-economic tastes for discrimination such as traditional animus or bigotry and the use of statistical inference(by the car dealers) to generalize about a buyer's reservation price based on his/her gender or race. The researchers note that car sellers were probably inclined to infer a buyer’s reservation price through preconceived (race and gender linked)notions of consumer knowledgeability, theability to search, and openness to prolonged bargaining.
The astronomical time and monetarycosts associated with running audit studiesthat rely on personal contact has led socialscientists to pursue alternative audit studydesigns. Furthermore, some social scientistshave speculated that racial discriminationmight take place before the applicant and thediscriminator actually meet face to face.The discriminator may rely on auditory cuesfrom the applicant’s voice before visuallyobserving race. Discriminating over the phone might be a more efficient method of screening applicants that could allow thediscriminator to avoid the discomfort of rejecting the applicant in person.To test this hypothesis, socialscientists have increasingly used the tactic of phone auditing. However, in instanceswhere telephone audits are the contactmethod of choice, the study is weakened bythe fact that auditors cannot convey their race over the telephone with absolutecertainty. Further, since the audit methodnecessarily stops a business transaction at anearly stage, it cannot gauge the presence of discrimination in later stages of the business2