Frame analysis

Frame analysis was used to answer research questions about the perceptions of the state administration regarding the case study conflicts and conflict management in state forestry. Both the written data and the interviews were used to identify and to reconstruct the frames (step , Figure 0). Written and public documents by the administration, such as the management plans, policy documents as well as press and Internet material were treated as a source for the collective frames of the organisations (Metsähallitus, ministries), while the interviews were treated as the representations of the frames of the individuals within the organisations. These two categories were kept separate throughout the analysis in order to allow the comparison between the collective and individual frames.

For the sake of anonymity and simplicity, all interviewees are in the text referred to as he, despite their gender. The choice of pronoun ‘he’ is based on the fact that an overriding majority of the informants are men, and calling all of them ‘she’ would imply a very different reality than the one found in the state forest administration. When the interviews are quoted in the text, the quotes are only in English (translations by the author) in order to protect the anonymity of the interviewees. The interviewees are referred to with codes that indicate whether the interviewee is from Metsähallitus Forestry Division (FD) or Natural Heritage Services (NHS) and whether (s)he is working in Kainuu (K), Inari (I) or at the head office in Tikkurila (T). The ministries are also indicated (MAF; MOE).

The interviews have a reference number according to the order they were carried out. For example:

NHSK6 = a person working for Natural Heritage Services in Kainuu, reference number 6

FDT0 = a person working for Forestry Division in Tikkurila, reference number 0

MAF8 = a person working at the Ministry of Agriculture and Forestry, reference number 8

Because my interest concerns both the content of the frames (question ) as well as on the ways the frames are used to support certain practices and courses of action (5), the analysis has included both of these elements. My approach was inspired by Peuhkuri (00) and Lewicki & Gray (Lewicki & Gray 00, 7; Gray 00, 0) who combine the analysis of the content of the frames with what Peuhkuri calls discursive frame analysis.

Classic content analysis comprises techniques for reducing texts to unit-by-variable matrix and analysing that matrix quantitatively for producing objective, systematic and quantitative description of the content of communication (Ryan & Bernhard 000, 785–

786). In more general sense, however, the term content analysis is used to refer to a wide range of methodological approaches that aim to produce a condensed description of the contents and structure of communication in a text or an interview (Tuomi & Sarajärvi 00; Peuhkuri 00, 5). I have used the latter approach to identify and describe the frames of the forest administration. This part of the analysis emphasises the elements of frames as organisers of experience, whereas the analysis of the dynamics of framing (below) focus on the bias-for-action element of frames.

Thus the third step of the analysis was to systematically construct and describe the structure and content of the frames. In the analysis I focused on four types of frames:

conflict frames, conflict management frames, organisational identity frames, and characterisation frames. In order to identify them, the following questions were asked in the case of each interview:

• What is the conflict about?

• Who are the key opposing parties in the conflict?

• Who is the “we” in these conflicts?

• Who are “they” in these conflicts?

• How should these conflicts be addressed? What would be the way forward?

• What would be a good solution to this conflict? Why?

These questions formed the main categories (Hirsjärvi & Hurme 000, 8) of the frame analysis. The answers to them came in many forms. Typically they were interpretations and reflections on what the past events were about, and why they happened the way they did. Based on the answers of the informants, and on the written data, I formulated subcategories for each of the main questions listed above and indicated for each statement in which interview or written material it could be found in. The purpose of combining research questions and main categories derived from theory, with subcategories from the empirical material, was to secure sensitivity to the material while maintaining theoretical relevance (Dey 99, 96).

Rather than looking for specific key words or other linguistic units, my analysis was based on looking for broader “storylines about what is to be comprehended” (Fisher 997). As both Fisher (997) and Nieminen (99) maintain, frames are not necessarily literally outlined in the text. Rather, to mention some elements is to recall the whole set. Like Nieminen (99), I differentiated in the analysis the frames from individual interpretations of an event based on that frame. For instance, a statement “there are other user groups we need to think of” is not a frame but an interpretation produced by a frame (balancing of local interests -frame).

Having gone through all the interviews once and having created new sub-categories along the way, I went through the material again with all the sub-categories in mind, and completed the analyses by putting it in the form of a table. The connection between the raw data and the table was made by indicating either the page in the written material where

the argument supporting the category could be found, or, in the case of the interviews by sorting the material out based on the categories with the help of Atlas.ti software for qualitative analysis. In the end, by looking at the categories derived from the material, it was possible to define different types of conflict frames, identity frames, characterisation frames and conflict management frames. When writing the analysis I returned to the original data, in addition to referring to the summaries produced in table format or in Atlas.ti.

I then approached the bias-for-action component of framing in a similar fashion to Peuhkuri (00) and Lewicki et al. (00), by analysing how the frames support certain courses of action over others (step ). Nieminen (99) calls this the “politics of framing”.

This step of the analysis functioned also as a test on the accuracy of the previous stage.

Johnston (995, 5 in Fisher 997) maintains that

“The final test of whether a […] frame has been correctly described is if these reconstructions help the analyst to understand why individual participants and social movement organisations act the way they do.”

Schön & Rein (99, ) identify a number of methodological challenges related to the fact that it may be difficult to tell what frame really underlies an institutional actor’s policy position:

• The rhetorical frames the actors use may be different from the frames implicit in their actions

• The same course of action may be consistent with quite different policy frames

• The same frames may lead to different courses of action

• The meaning of a policy made at the central governmental level may be transformed at the local levels at the stage of policy implementation

• It may be difficult to distinguish between conflicts within a frame and conflicts that cut across frames

• It may be difficult to distinguish between real and potential shifts of frame

According to Schön & Rein (99, 6) these challenges may be overcome by “carefully nuanced observations and analyses of the processes by which policy utterances and actions evolve over time and at different levels of policy-making.” They maintain that any given construction of a frame can be tested against relevant data – for example, the texts of policy debates or the routines of policy-making process (Schön & Rein 99, 6).

I have attempted to follow this advice by looking at the frames at different levels of the state forest administration, by analysing the practices, and by covering a time period of several years through the written material in both case studies. Nonetheless, there is an inherent theoretical difficulty with frame analysis, namely that frames must be constructed by someone. Those who construct frames do not do so from positions of unassailable frame–neutrality (Schön & Rein 99, 6). I have tried to address this unavoidable challenge by outlining my method as openly as possible and by openly reflecting on my position as a researcher (see Chapter 7.5).