6. Limitations and further research
Several limitations can be identified for this study. First, the industry-specificity and relatively small sample of ICT MNEs limits the generalizability of the findings presented in Section 4 for a broader set of ‘firms-in-industries’ (Geels, 2014) in response to the emergence of smart cities. Second, the lack of specific firm-level data on revenues and sales for their activities in smart cities makes it difficult to determine the extent of firm-specific investments in this market, and assess its strategic importance for the firm. Similarly, the lack of firm reporting on key figures at the level of specificity of cities is a limitation to explore firm strategies at the subnational level. While the GaWC inventory, which was adopted instead, provides insight into the presence of a firm in a particular city, it does not give an accurate picture of the scope of firm activities at that level. A third limitation stems from the use of documentation published by the focal firms, which is inherently a form of self-representation, often meant for reputational purposes. By triangulating firm information with other sources, including semistructured interviews and publications from reputable third parties where possible, an attempt was made to redress this limitation. However, lack of possibilities to check company statements is an issue.
For future research, it would be fruitful to explore which intra-MNE knowledge sharing mechanisms are most effective in leveraging nonlocation-bound FSAs throughout the MNE network. The interviews with IBM, Cisco, and Accenture all confirmed that the transfer of explicit knowledge between locations occurred between subsidiaries, and enabled them to draw on resources and capabilities developed from multiple smart city engagements globally. The transfer of tacit knowledge is far more complex, however, given that it is embedded in the routines of individuals, and therefore difficult to transfer through information systems. Hence, gaining insight into effective mechanisms for knowledge transfer between subsidiaries with the MNE network would be worthwhile. This is intertwined with the capacity of MNEs to leverage non-location-bound FSAs beyond a specific local context, as emerged from the analysis of these MNEs. The spatial heterogeneity of each urban environment (Beugelsdijk et al., 2010; Beugelsdijk and Mudambi, 2013), and the need for local responsiveness on the part of the MNE in relation to environmental and social issues (Kolk, 2010; Kolk and Margineantu, 2009), has made ICT firms and smart city technologies an interesting initial research context at the subnational level. Nevertheless, there are many questions, related to the actual importance and relevance as well as the implementation, beyond that what is stated by companies verbally and in writing.
