Cohort Turnover and Operational Performance: The July Phenomenon in Teaching Hospitals

We consider the impact of cohort turnover—the planned simultaneous exit of a large number of experienced employees and a similarly sized entry of new workers—on operational performance in the context of teaching hospitals. Specifically, we examine the impact of the annual July turnover of residents in U.S. teaching hospitals on the average length of hospital stay and mortality rate in teaching hospitals relative to a control group of non-teaching hospitals. Despite the anticipated nature of the cohort turnover and the supervisory structures that exist in teaching hospitals, the annual July turnover of residents results in a longer average length of stay (i.e., increased resource utilization) for both minor and major teaching hospitals and higher mortality rates (i.e., decreased quality) for major teaching hospitals, relative to a control group of non-teaching hospitals. In major teaching hospitals in particular, we find evidence of an anticipation effect that presents as a gradual decrease in performance beginning several months before the actual cohort turnover. We identify higher overall quality of nursing and increased intensity of potential quality assurance as managerial levers for mitigating this decrease in hospital operational performance. ∗Corresponding author. Email: hsong@hbs.edu. Address: Wyss House, Soldiers Field Road, Boston, MA 02163. This is a significantly revised version of an earlier working paper, "Cohort Turnover and Productivity: The July Phenomenon in Teaching Hospitals," Harvard Business School Working Paper No. 14-113 (May 2014). We thank Jay Bhattacharya, David Cutler, Arnold Epstein, William Evans, LeRoi Hicks, Barbara McNeil, Paul Oyer, Douglas Staiger, and seminar participants at the University of Pennsylvania, Harvard Business School, Stanford Graduate School of Business, Kellogg School of Management, Fisher College of Business, Yale University, the 14th Annual Health Economics Conference, the National Bureau of Economic Research’s Summer Institute, the 2013 INFORMS Healthcare Conference, the 2013 INFORMS Annual Meeting, and the 2015 POMS Annual Conference for helpful comments on an earlier version of this manuscript. We also thank Juliana Pakes for information regarding the Charlson Index and Sarah Woolverton for expert coding assistance. We acknowledge financial support from the Division of Research and Faculty Development at the Harvard Business School.