Werth, Bogdan (2010) Uncertainty in IT Outsourcing of Large Financial Institutions, Doctoral Thesis, Manchester Metropolitan University Business School, Manchester, United Kingdom.
According to the recent industry reports, relatively few European banks have benefited from information technology outsourcing to the extent originally anticipated by these institutions. Despite the setbacks, most intended to increase outsourcing activities in pursuit of reduced labour costs, specialized skills, process expertise, superior technical resources, and enable a better focus on their core business. Whether or not these reasons are justified, outsourcing often has unexpected impacts, both positive and negative, on the operation of an organisation.
Despite the lack of a commonly accepted interpretation of outsourcing outcomes, the practice of outsourcing in modern corporations is proliferating. This trend has forced both practitioners and academics to theorize and speculate on the underlying momentum towards outsourcing. So far the information technology outsourcing literature has mainly focused on the transaction itself, without investigating the strategic characteristics of the organization leaving out of sight a built-in social framework of the firm or social context it is located in. This trend captures the widely held perception that organisational members makeoutsourcing decisions based upon an economic rationale and regard social factors as negligible in an overall picture of outsourcing.
Agent-based modelling provides us with a tool for examining the implications of various social and organizational choices within organisations. This thesis examines how this can be done in the case of choosing to outsource (or not) the information technology needs within large banking organisations. This thesis introduces an alternative approach to the investigation of economic theories by means of evidence-based agent-based social simulation. It suggests and seeks an alternative metric for asset specificity with a more qualitative flavour. Williamson’s transaction cost framework is replicated with autonomous, heterogeneous agents as actors of the modelled outsourcing process that can act according to changes in the environment they are located in. Social structures emerge from the interaction and information exchange between individuals in the market. The models succeed in producing a qualitative definition of the term asset specificity.
The modelling and the fieldwork were carried out by the same party with data engineering involving stakeholder interactions. A constructive and modular approach to model design was adopted. The rules for the agents’ behaviour were derived partly from the relevant reports and partly from qualitative insight into the modelled target system. The first research stage involved gathering of relevant data in collaboration with industry partners. Based upon data collected from the fieldwork a prototype declarative agent-based coarse grained model was developed. Building a “mock-up” model first was intended in order to point out data requirements, and help to determine which data is important and which can be dismissed. This model allowed the identification of gaps in the current data set that need to be filled in order to develop a more informative model. In subsequent research stages additional interviews were conducted in order to fill gaps, which became evident through the analysis of the mock-up model. Exploration and validation of the models’ results happened in a constant feedback-loop, together with domain experts. The constant cross validation with stakeholders and domain experts facilitated the development of a simulation that was credible for practitioners, who came to participate in the research with sceptical views. This particular case study is presented as an example of a generalised approach that can be used to examine different aspects of organisational change.