Volume 45 | Number 6p1 | December 2010

Abstract List

David Bradley Wright


To test whether physicians' provision of charity care depends on their hourly wage.

Data Sources

Secondary data from four rounds of the Community Tracking Study (CTS) Physician Survey (1996–2005). Data are nationally representative of nonfederal office‐ and hospital‐based physicians spending at least 20 hours per week on patient care.

Study Design

A two‐part model with site‐level fixed effects, time trend variables, and site–year interactions is used to model the relationship between physicians' hourly wage and both their decision to provide any charity care and the amount of charity care provided. Salaried and nonsalaried physicians are modeled separately.

Data Collection/Extraction Methods

Data from each round of the CTS were merged into a single cross‐sectional file with 38,087 physician‐year observations.

Principal Findings

The association between physician's hourly wage and the likelihood of providing charity care is positive for salaried physicians and negative for nonsalaried physicians. Among physicians providing any charity care, hourly wage is positively associated with the amount of charity care provided regardless of salaried status. Practice characteristics are also significant.


The financial considerations of salaried physicians differ significantly from those of nonsalaried physicians in the decision to provide charity care, but factor similarly into the amount of charity care provided.