Causality analysis and human resources management policy determinant in Balanced Scorecard: A system dynamics approach

The most widely adopted Performance Measurement Systems (PMS) is the BSC which is a framework for translating organisation strategy into a set of achievable performance indicators. The uniqueness of BSC that differentiates it from other PMS is the strategy map that has a unidirectional causality which follows the bottom-up approach. However, BSC contains certain shortfall despite its apparent popularity. The criticisms on the causality in BSC have been widely discussed as to whether they are based on statistical testing, logic or assumptions. Well-developed causal models are valuable for improving business performance, predicting and decision-making to foresee how action affects future performance. Therefore, the relationships between measures should have notion of causality. In seeking empirical evidence of causality linkages in BSC, a theoretical framework which consist of 10 propositions based on the Service Profit Chain (SPC) theory were developed and tested using the statistical causality analysis; the Granger causality tests on the 45 time series data point extracted from the Business Performance Review report. Results of the study show insufficiency of well-established causality models as only 40% of the causal linkages were supported by the data.;However, in situations of insufficient result based on historical data, it was suggested to use expert knowledge. Therefore, the Delphi method was conducted in this study as a mode of primary data collection to achieve consensus on the causality existence among the 15 selected expert persons within their domain of expertise by utilising a series of questionnaires through the calculation of mean and standard deviation. Interview sessions working with unstructured questionnaire were conducted in the first round whilst the second and third round questionnaires were distributed through emails. It was found that only 20% of the propositions were not supported by the expert knowledge. Results from the both methods indicated the existence of bidirectional causality between employee engagement and revenue and there were interaction among the causal linkages which demonstrate significant dynamic and complexity environment. However, the impact of cause and effect is difficult to observe and determine in such environment. Taking this into consideration, a follow-up computer modelling and simulation using System Dynamics (SD) was adopted to observe the behaviour of performance and to identify and analyse how alternative policies of Human Resource Management (HRM) impact the performance of the business in the said environment.;The sensitivity and extreme condition tests were conducted to ensure the robustness and validity of the model as a policy analysis tool. With the existence of interaction and causality linkages among measures, the simulation results indicate that 90% of the tested propositions on the causality were supported by SD computer modelling. In terms of policy analysis, it was found that the upsurge in employee training hours would increase employee engagement, productivity and revenue except customer satisfaction and profit. In addition, organisations should design a strategic plan to overcome the increase in number of employees leaving which would decrease employee engagement, productivity, customer satisfaction, revenue and profit. In general, the study largely contributes to the SPC theory by narrowing the research gap in identifying the dynamic relationship (bidirectional relationships) between measures. Literally, this study contributes to the extant literatures on the areas of PMS by analysing the cause and effect relationships among measures through econometrics statistical analysis, expert knowledge and computer modelling and simulation approach.