The research of this thesis commenced in September 2004 and was originally concerned with the role of manufacturing within the context of increased levels of outsourcing. There was a general concern that the future of manufacturing in Sweden was at risk, and that all new manufacturing jobs would be located in low-cost countries in Eastern Europe and in Asia.
Because previous studies (such as my MSc thesis), had focused on the supply chain management field, it was natural to have a supply chain perspective on manufacturing. This means a more holistic view of manufacturing, where the relations between the focal manufacturing firm and its suppliers and customers are key to maintaining efficient production. When exploring the role of manufacturing in the supply chain, the concept of absorptive capacity seemed a fruitful way of describing the role manufacturing plays in facilitating external integration. Other approaches were certainly also feasible, and the concept of dynamic capabilities (Teece and Pisano, 1994) was indeed also considered. The main weakness of the absorptive capacity literature is the lack of universally agreed definitions, and perhaps another framework would have brought more clarity. Also, the lack of coherent theories about absorptive capacity, particularly in manufacturing, meant that there was great deal of room for interpretation.
The literature contains several studies predicting that both manufacturing capability and external integration should contribute positively to plant performance, but the link between these two was less developed. When it became apparent that those firms that
39
had positive development of the manufacturing department also were more likely to have a strong correlation between the variables concerned with customer/supplier relations and plant performance, an interesting subject for a thesis was established. Studies explaining the observed phenomenon were thereafter reviewed, and the most promising explanation was absorptive capacity (Cohen and Levinthal, 1990). The first phase of the research can thus be described as being a mix between inductive and deductive reasoning. The second element of the quantitative data analysis had a clear deductive approach. During a course in innovation management, I read an illuminating book by Christensen (1997) and curiosity compelled me to test how Christensen’s theories relate to integration. The reading of Christensen’s (1997) book influenced paper 1, which is the main reason the term competitive priorities forms an important part of this thesis. Basing that part of the research on literature focused on manufacturing would have been a more conventional approach to the subject. However, the way Christensen’s (1997) framework is used in this thesis does not deviate much from more traditional approaches to manufacturing strategy. The main difference, and indeed strength, is that Christensen’s (1997) framework is more dynamic compared to manufacturing strategy–based frameworks. A competitive priority is not set in stone, but changes over time, sometimes rapidly, in response to changing technology or customer preferences. Christensen (1997) provides a more detailed description of why competitive priorities change when technologies and markets mature, and what the consequences are.
After the appropriate literature was reviewed, a conceptual framework for investigating the desired phenomenon was devised. The operationalisation of the concepts had to be coherent with the identified literature, but also with the data that had already been collected. This poses some issues concerning validity that will be further discussed later in this chapter. After the appropriate statistical analysis was conducted using the SPSS software package, the results were presented in two conference articles that were accepted at the 2005 and 2006 EurOMA conferences. The second paper of this thesis is identical to the one presented in the 2006 conference, whereas the first paper has been reworked for publication in the International Journal of Manufacturing and Technology Management.
The choice of the survey methodology was also more or less predetermined. When research commenced survey data had already been collected by former PhD student Stefan Westin. Data analysis combined with simultaneous literature studies was thereby the first step on the journey, not the more traditional literature studies – data collection – analysis sequence. There are both advantages as well as disadvantages with the conducted approach. The main advantage concerns the time gained in not having to design, distribute and collect the questionnaire. The main disadvantage is that the survey was not designed in an optimal way for the theory used, which may raise some questions about validity. It would, for example, have been better if manufacturing capabilities had been measured in line with the findings of the case study in paper 5. This paper highlighted some of the most important capabilities needed for successful external integration, which were not the same as the one picked up by the survey. Operationalisation of the constructs had to be done so that they fit with the obtained data. This is a main concern, especially for paper 4, as was also pointed out by reviewers.
The findings of the survey-based papers did not give a satisfyingly clear picture of the studied phenomena, and therefore more in-depth studies of a small number of firms were needed. A case study methodology was chosen because it could fulfil the research criteria
40
within the available time frame. The aim of the first case study, presented in paper 5, was to illustrate manufacturing-related absorptive capacity and to get a better picture of what integration mechanisms and capabilities contribute. A rather short case study, where data was collected mainly through semi-structured interviews, was deemed adequate. By contrast, the aim of paper 2 was to get a more complete picture of the challenges firms face when designing their supply chain. The methodology and the data collected were in this case more inclusive. The opportunity to take part in a preliminary study at Ericsson, and the openness of the managers there, was naturally the main determinant of the chosen case as well as the data collection method.
As mentioned, responsibility for collection of the survey data was originally Stefan Westin’s, but since he quit the university and took no part in the data analysis of the papers, his name is not included in any of papers, although his efforts deserve to be acknowledged in this thesis. The case study data was collected by me. I also conducted all data analysis and writing of papers, with very helpful feedback from colleagues, reviewers and most of all PhD programme supervisors. In acknowledgement of his endless support and substantial insights, the main supervisor, Lars Bengtsson, also appears as co-author of papers 1, 3 and 5.
41