A Quantitative Analysis of Racial/Ethnic Disparities in Health Care Usage: The Roles of Perceptions of Care & Expected Longevity

Abstract

Countless studies have documented the existence of disparities byrace and ethnicity in the health care system in the United States.However, not many have sought to explicitly quantify the relativeimportance of each of the proposed determining factors. Thisdissertation thus sought to identify the major causes of disparitiesin health care usage by employing an adaptation of theBlinder/Oaxaca decomposition method originally used in theliterature on wage differentials.Specifically, this study sought to investigate the roles that racialdifferences in patients' perceptions of care and expected longevity(as they relate to expected gains) might play in the observeddisparities in usage. Perceptions of care measures (trust andfear/dislike) were used to investigate the importance of thedoctor-patient relationship in determining usage disparities throughits effect on expected gains. Expected longevity was used as thelength of time that someone expects to live would affect how willingthey are to invest in anything with payoffs derived in the future.Thus, if expected net payoffs/gains of utilizing health care arepositive, the individual would use care.Results show that while expected longevity was found to positivelyaffect whether or not someone utilizes care, it was not acontributing factor to the racial divide in usage. However, whenexploring the role of perceptions of care, while the differences inutilization rates between whites and minorities were due mainly tounobserved factors peculiar to each racial/ethnic group, perceptionsof care disparities also played a major role in addition tosocioeconomic and access factors. These results, while encouraging the need for further analysis into what drives usage disparities by race, point out that the doctor-patient relationship (apart from SES and access issues) should also be considered as a possible contributor to racial/ethnic disparities in usage.