Public Employee Accountability: An Empirical Examination of a Nomological Network

Abstract

While the field of public management has long emphasized the importance of holding public employees accountable for their actions, our understanding of how best to do so has been hindered by a dearth of empirical research on this topic. Recent literature has called for more examination of individual subjective experiences of being accountable—employee accountability—in order to research how macro accountability systems make a difference to individuals. To validate the employee accountability construct in the public domain, we collect data from two separate samples using a multidimensional scale to measure public employee accountability. We analyze these data to explore the nomological network of public employee accountability by assessing its relationship with both a logical antecedent and expected outcomes. Based on findings that validate the nomological network, we discuss how this research contributes to the literature on accountability and possible directions for future research.

Keywords:

• Accountability system
• employee accountability
• individual accountability
• nomological validity
• public employee accountability

Notes

1 A confirmatory factor analysis (CFA) of data from this study sample verified the five-dimensional higher order factor model of the scale of reflective indicators capturing measurement errors (the results are available on request).

2 In line with previous literature (e.g., Byrne, 2016), we conducted Bayesian Structural Equation Modelling (BSEM) to treat our variables as categorical. The Maximum Likelihood (ML) estimation method is recommended when there are five or more categories, sample size is small, and category thresholds are approximately symmetric on non-normally distributed data whereas categorical least squares methodology is recommended when data sets contain variables with fewer than five categories (e.g. Rhemtulla et al., 2012); with six to seven categories, results are similar across methods for many conditions. Results of BSEM were not different from our findings generated using ML. Thus, we present the ML results below in order to show all measures including various indicators of overall fit. Results of the BSEM analysis are available on request.

3 All detailed statistical information produced to analyze the potential CMB issue is available on request.

4 We checked differences in chi-square fit between models with and without the structural paths and also models collapsing the measurement model into single observed constructs with a fixed error variance. However, the model alterations do not lead to significantly different results but instead reach the same conclusions. As such, model fit measures was not confounded (details available on request).

Additional information

Funding

This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2018S1A3A2075609).

Notes on contributors

Yousueng Han

Yousueng Han is an Assistant Professor in the Department of Global Public Administration at Yonsei University. He received a Ph.D. in Public Affairs from the Paul H. O’Neill School of Public and Environmental Affairs at Indiana University-Bloomington. His research and teaching interests include public management, citizen participation, and governance.

Peter J. Robertson

Peter Robertson is an Associate Professor in the Sol Price School of Public Policy at the University of Southern California. His research and teaching focus on the development of collaborative organizational and governance systems that enhance the quality of life for people, their communities, and the natural environment.