The Effects of Managers’ Purposeful Performance Information Use on American Hospital Performance

Abstract

There is a growing body of research on the impact of purposeful performance information (PPI) use on organizational performance (Kroll, 2015a), but questions remain about how a top manager’s operating context may affect the relationship between PPI use and performance. In this study, we examine the relationship between top managers’ PPI use and performance, and cross-sector differences in the relationship between PPI use and performance. We test our hypotheses by combining survey data of nearly 1,000 top managers from public, private, and nonprofit American hospitals with indirect (client satisfaction), direct (patient readmission rates), and perceptual (managers’ self-assessment) performance measures, while controlling for task. We move beyond overall performance information use by exploring the relationship between PPI use and performance in four specific decision areas. We observe a fair amount of homogeneity across the three sectors but also find salient differences regarding how PPI use is linked to different performance indicators. While public hospitals primarily use PPI to meet community needs, nonprofit and for-profit hospitals use PPI to improve patient outcomes. The article concludes with a discussion of the findings within the performance management and cross-sector differences literatures.

Keywords:

• cross-sector differences
• hospitals
• performance management

Notes

Notes

1 See Johansen and Zhu (2014) and Zhu et al. (2015) for more detailed statistical assessment of survey response rates by key organizational features (hospital size, service specialization, organizational ownership, etc.) and survey nonresponses.

2 Common source bias would be more of a concern if all we used were perceptual measures. However, taken together with objective and client satisfaction performance indicators, our results from all five performance measures provide support for our hypotheses. We performed two tests for common source bias between the Meets Community Needs variable and the four PPI use indicators. The principal components analysis is below the 50% threshold (proportion of component 1 = 0.446). The Cronbach’s α score is 0.6692, which while a bit high, still falls within the acceptable range.

3 Despite different modes of data collection (e.g., qualitative interviews, large-N surveys, or mixed-methods strategy), scholars often rely on managers’ self-reports on how, and how often, they use performance information. These kinds of self-reported measures may be prone to social desirability bias: managers may over-report their activity levels in order to make themselves appear more socially desirable. We recognize this as a limitation of our empirical measures of PPI use. However, given the fluid nature of activities in performance information use, this is still a legitimate and common approach to collecting large-N data regarding performance information use and other managerial activities.

4 A fifth performance information use measure was created, PI use index, which is a factor score of the four purposeful performance information use measures. This measure allows us to more easily compare findings from more general PI use measures to the findings presented here, particularly when examining cross-sector differences. A summary of the results using the PI use index measure based on the full sample and by sector is available in the Appendix.

5 All states, except Delaware, generate valid survey responses. For empirical models, in which we pool all the hospitals together, we check model results by replacing the categorical ownership variable with two dummy variables for public and nonprofit hospitals. We find very similar results and draw the same substantive conclusions regarding how PPI affects hospital performance.

6 The online Appendix can be accessed from https://ling-zhu.com/research-2.

Additional information

Notes on contributors

Taehee Kim

Taehee Kim is assistant professor public administration at Seoul National University of Science and Technology, South Korea. Her research focuses on public management, performance management, and human resources management.

Morgen Johansen

Morgen Johansen is professor of public administration at the University of Hawaii at Mānoa in Honolulu. Her research interests include cross-sector differences in managerial behavior, gender and racial diversity, and the relationship between managerial behaviors and performance. She is the author of “Social Equity in the Asia-Pacific Region.”

Ling Zhu

Ling Zhu is associate professor of political science at the University of Houston. Her research interests include the political economy of welfare policy, social inequality in health care and outcomes, the management and governance of public service networks, and research methodology.