Method
This paper uses a quantitative content analysis of press articles to analyze and describe the case of the historical emergence and evolution of the cleantech investment category. Similar methods have found increasing prominence in organizational research recently (e.g. Phillips, Lawrence, & Hardy, 2004; Ventresca & Mohr, 2002; Wuthnow, 1989). For example, research on the historic shifting composition of actors and frames in corporate environmentalism (Hoffman, 1999) and the construction of market categories in the computer workstation market (Kennedy, 2005). McGrath (1998) used media patterns to analyze technology s-curves and technology cycles within the electric vehicle battery industry. Furthermore, the evolution of the U.S. satellite radio was researched over sixteen years based on a multitude of data sources including a large sample of newspaper articles. Within the textual documents changing thematic frames explained the differentiation in three different phases - emergence, commercialization, and early growth (Navis & Glynn, 2010). A further category emergence paper looks at the market for modern Indian art. A discourse analysis of diverse textual documents including newspaper articles shows the growing legitimization of the market category through a shared rhetoric (Khaire & Wadhwani, 2010).
The advantages of computer aided content analysis are in the depth of the assessment of a broad sample of textual documents (e.g. newspaper articles, excerpts). Additionally supporting is the independence from databases, which, especially in the field of VC investments, usually present a restricting factor in regard to data quality and availability. Resulting, quantitative content analysis enables a more detailed industry analysis than the assessment of classical venture databases (Berelson, 1952; Elo & Kyngas, 2007; R. P. Weber, 1990). WordStat 6.0 by Provalis Research is a content analysis and text mining software for unstructured textual documents. It is used to analyze the created database based on a predefined dictionary, which is a collection of words, structured in several different subgroups. The software counts the words according to the structure defined in the dictionary (Krippendorff, 2012; Neuendorf, 2002).
As a foundation for structuring the cleantech sector dictionary, the taxonomy developed by the Cleantech Group2 was used. The Cleantech Group is one of the leading market intelligence companies in the field of clean technologies and is widely seen as influential over the establishment of the term “cleantech”. The Cleantech Group’s definition of the sector spans 13 categories encompassing several different industries and technologies. We built the dictionary according to these categories, and introduced two additional overview categories to observe general word groups related to “cleantech” and “ecology”. These 15 different categories were applied to the quarterly structured articles from the 17 years from 1995 to 2011. The analyses permit us to give attention to included themes and technologies throughout the research period. Below, the relative importance of the cleantech category in entirety is contrasted to all VC mentioned. Specific sub-categories are reported in comparison to the cleantech category.
A description of the early years of the cleantech industry is found in O’Rourke (2009). Her analysis shows the important linkage of cleantech to the VC sector and defines it as a venture category. The analysis reports that 74% percent of all articles mentioning “cleantech” include mention of “venture capital”. The analysis used in our study differs from O’Rourke (2009) - by looking at all articles mentioning “venture capital” and the terms associated with cleantech. “Greentech [Cleantech] VC investing has received little attention in the scholarly press, but enormous attention in the popular press” (Kenney, 2011b, p. 218). This paper modifies Venture Capital Life Cycle Model so far only used in a national context (Avnimelech et al., 2004; Avnimelech & Teubal, 2006).
To understand the detailed analysis of investment data this paper builds on the quantitative content analysis of a dataset of 84,259 articles reporting on “venture capital”. Applying the assembled cleantech dictionary allows for fine-grained analysis of trending categories and themes at certain points in time. In order to investigate emergence it is necessary to have data predating emergence; for investment categories this implies the need to cover a time not observed by traditional financial databases or added to them post hoc (Woolley, 2011). [1]
- [1] http://www.cleantech.com/
