Although demographics, cohort, and contextual factors are expected to influence physician supply at the intensive margin, much of the literature has examined the demographics and very limited cohort analysis is undertaken. This paper employs a cross-classified fixed-effects methodology to examine the importance of age, period and cohort, and contextual factors in explaining the declining work hours of Canadian family physicians. We define cohorts with five-year intervals according to year of graduation from medical school. Contrary to the previous literature, we find no evidence of reduced hours of work provided by the new cohorts of physicians. Compared to the 1995-99 cohort, older male cohorts perform similar total hours of work per week except those who graduated in the 1960's while older female cohorts consistently perform fewer total work hours in the range of 3-10 h per week. Consistent with the literature, it is found that female and older physicians provide fewer hours of work compared to the male and younger counterparts, respectively. Although there has been a decline in total hours of work for all physicians in the range of 2-3 h per week in each period, we find a large decline in direct patient care hours (about 4-6 h) and a marginal increase in indirect patient care (about 2-4 h) over the period. Having children less than 6 years and children aged 6-15 years in the physician's family reduce the work hours of female family physicians by about 7 h and 3 h, respectively. A number of other contextual factors influence work hours of physicians in the expected direction.
Copyright © 2011 Elsevier Ltd. All rights reserved.