• Keine Ergebnisse gefunden

3.4.1 Data collection and preparation

Our enquiry into accounting standard setting internals is premised on a content analysis of IASB meeting audio playbacks. MP3-audio recordings of board meetings, which are accom-panied by further content such as Q&A webcasts and podcast summaries,11F11F10 are publicly avail-able since 2006 from the IFRS Foundation’s website for completed and ongoing standard set-ting projects. Whereas the latter are prepared to complement written information (e.g., high-level summaries, comment letter submissions, and press releases), IASB meeting audio play-backs are unedited recordings of IASB meetings and echo board debates and tentative deci-sions, accordingly.

Data were collected in two steps. First, we downloaded the recordings of all 21 IASB meetings that took place after the end of the DP comment period (November 2008) until the release of the amended standard (June 2011) from the PEB project section of the IFRS web-site (duration of 29:27:01). As presented in TABLE 2, 14 board meetings relate to the due process phase between DP and ED (16:45:51)12F12F11 and seven meetings to the phase between ED and the amended standard (13:30:13). Second, we prepared transcripts for all 21 IASB meet-ing audio playbacks in their original language, i.e., English13F13F12.

For the content analysis, we limit our sample to the 14 board meetings between DP and ED (16:45:51). We do not include meetings prior to the release of the DP because we aim to observe board discussions under explicit consideration of constituents’ comment letter feed-back. In compiling the DP, the IASB collected and presented different approaches without being selective, whereas, in the post-ED phase prior (tentative) decisions and ED proposals

10 The availability of IASB meeting audio playbacks grounds in the due process core principle of transparency as formulated in the IFRS Foundation’s Due Process Handbook (IASC Foundation, 2006; IFRS Foundation, 2013b). Accordingly, meetings of the IASB or the Interpretations Committee are generally open to the public (outside observers are either allowed to attend meetings in person or via a live-webcast after registering online). Moreover, all meetings are recorded and provided online (IFRS Foundation, 2013b: 3.2). As of No-vember 2013, the IFRS Foundation website contains approximately 800 hours (33 days) of IASB meeting audio playbacks.

11 The period further involves a meeting of six IASB members and the Analyst Representatives Group (later renamed into Capital Markets Advisory Committee) on 24 February 2010 (0:51:04). Although a recording is available, we are unable to transcribe it due to poor audio quality.

12 We follow the basic transcription principles of Mergenthaler & Stinson (1992). Accordingly, we maintain syntactical errors, eye dialect, interruptions and duplication of phrases, but smooth out hesitation vowels for the purpose of readability. We mark incomprehensible and unarticulated phrases with square brackets. For parsimony, we abstain from indicating linguistic style (e.g., pitch, intonation, accentuation) in our transcripts, but mark nonverbal communication (e.g., laughter). Any statement is marked with a time label [h:mm:ss] and a speaker label (initials). Because speakers are not formally introduced, we identify them by the use of first names throughout all meetings. We fail to identify two members of the technical staff (U1 and U2) who ap-pear on only a limited number of occasions.

were largely confirmed14F14F13. Accordingly, we do not include board meetings between the ED and the amended standard in our sample but resort to the respective transcripts to track major decisions that were made within the sample period.

TABLE 2

IASB meetings on the amendment of IAS 19 (2011)

Phase Date (dd/mm/yyyy) Duration (h:mm:ss)

Before DP (not included) various 13:30:13

TABLE 2 displays date and duration of 21 audio playbacks of IASB meetings on the amendment of IAS 19 (2011) between November 2008 and February 2011, arranged by due process phases. The DP to ED-phase includes 14 IASB meetings that took place after the end of the DP comment period until release of the ED (16:45:51, sample period). The ED to standard-phase covers seven board meetings that were held after the end of the ED comment period until release of the amended standard (12:41:00).

No meetings took place during the comment periods. Our content analysis is limited to the DP to ED-phase (sample period) and does not embrace board meetings prior to the DP and after the ED. Through-out the paper we use the British date format dd/mm/yyyy.

_________________

a Meeting audio playback provided in two MP3-files.

b Original duration of the recording is 1:41:21. The meeting closes with a coffee break of 0:29:40 that was accidentally recorded (not included).

TABLE 3 presents all sample meeting attendees arranged by board members (Panel A), tech-nical staff (Panel B) and senior staff (Panel C).

13 Agenda Paper 5B (November 2011) illustrates that nearly all major ED proposals were confirmed by the board in the post-ED phase.

TABLE 3 Participants of the sample IASB meetings on the amendment of IAS 19 (2011) Panel A: Board members Initials Name Origin Period of service Previous employerNative speaker DTSir David Tweedie (Chairman)UK01/2001–06/2011UK Accounting Standards BoardYes TJ Thomas E Jones (Vice-Chairman) USA 01/2001–06/2009 Citigroup Yes MB Mary E Barth USA 01/2001–06/2009 Stanford University Yes SC Stephen Cooper UK08/2007UBS Yes PDPhilippe DanjouFrance11/2006Autori des Marchés Financiers No JE Jan EngströmSweden05/2004VolvoNo PF Patrick Finnegan USA 07/2009CFA Institute Yes RGRobert P Garnett South Africa01/2001–06/2010Anglo American Corp.Yes GG Gilbert Gelard France 01/2001–06/2010 KPMG No PK Prabhakar Kalavacherla India 01/2009KPMGYes JL James J Leisenring USA 01/2001–06/2010 FASBYes PMC Patricia McConnell USA 07/2009–06/2014 Bear Stearns & Co. Yes WMG Warren McGregor Australia 01/2001–06/2011 Australian Accounting Research Foundation Yes AGAmaro Luiz de Oliveira Gomes Brazil 07/2009Central Bank of Brazil No JS John T SmithUSA 09/2002–06/2012 Deloitte Yes TY Tatsumi Yamada Japan 01/2001–06/2011 PricewaterhouseCoopers (affiliate) No WZ Wei-Guo Zhang China 07/2007China Securities Regulatory Commission No Panel B: Technical staff Initials Name Position AMGAnne McGeachin Senior Project Manager MKManuel Kapsis Assistant Technical Manager AP Andrea PrydeTechnical Principal U1 Unknown 1 U2 Unknown 2 (Continued)

TABLE 3Continued Participants of the sample IASB meetings on the amendment of IAS 19 (2011) Panel C: Senior staff Initials Name Position PC Peter Clark Director of Research GFGavin Francis Director of Capital Markets WU Wayne Upton Director of International Activities TABLE 3 displays the 25 participants of the 14 sample IASB meetings on the amendment of IAS 19 (2011) between November 2008 and February 2010 (DP to ED). At- tendees are arranged by board members (Panel A), technical staff (Panel B) and senior staff (Panel C). Origin, service period and previous employer of board members were adopted from Botzem (2012) and Deloitte (2015). We consider an IASB member a native speaker if English is national or official language in his or her country of origin. Information on the positions of technical and senior staff members was collected from the IFRS Foundation website.

We further collect all observer notes on the 14 sample board meetings. These papers are prepared by technical staff and provided to board members prior to a meeting (IFRS Founda-tion, 2013b). They contain staff recommendations, executive summaries and illustrative ex-amples. We include them in our analysis to illustrate staff proposals and supplement discus-sions that are not self-explanatory using attendees’ statements (e.g., regarding numerical ex-amples) or the progress of meetings.

3.4.2 Content analysis and inductive coding

Content analysis is a research method that aims to structure and describe the content of com-munication by means of abstraction and simplification (Berelson, 1952; Dey, 1993; Flick, 2014; Holsti, 1969; Krippendorff, 2004; Mayring, 2010; Neuendorf, 2012).Whereas content analysis may imply purely quantitative analyses (e.g., word counts), due to its mainly inter-pretative and contextualising character, it is generally considered a qualitative approach. The basic methodical idea of content analysis is to abstract from the original complexity of com-munication by arranging pieces of the same or similar meaning into definite categories. The outcome is a more accessible category (or, synonymously, code) framework that groups relat-ed observations and that may be interpretrelat-ed on its own, allowing for inferences to be drawn that extend beyond the original data (e.g., Elo & Kyngäs, 2008; Krippendorff, 2004). Accord-ingly, the definition and assignment of codes—ultimately guided by the researcher’s analyti-cal aim—are methodianalyti-cal core issues in content analysis. Data categorisation (or, coding) may be performed in two different but not mutually exclusive ways. First, codes may be ex ante deduced from related theoretical considerations, comparable observations or extant evidence and then assigned to the data at hand (deductive content analysis). In that regard, deductive coding allows for the identification and re-search of established concepts in a novel data con-text. Second, codes may be directly processed from the material itself (inductive content anal-ysis). The inductive approach is grounded in gradually identifying, refining and validating categories that seem most salient with respect to the researcher’s purpose of analysis. Induc-tive content analysis thus aims at the most naturalistic description of the communication at hand (Mayring, 2010).

Because there is no prior study that applies content analysis to IASB meeting recordings comprehensively (to the best of our knowledge), we consider inductive content analysis the most promising approach to start with. Methodically, inductive content analysis involves an open coding process (Glaser & Strauss, 1967; Strauss & Corbin, 1990). Accordingly, we jointly read through the transcript data and assigned preliminary categories to board and staff

members’ individual statements (or parts of them). In a second round, categories were refined by clustering codes of similar meaning and dropping redundant ones, which resulted in a final set of 205 categories. Consistent with our research interest in the aforementioned internal as-pects of IASB standard setting, we arranged them by four main categories:

1) Project elements: The first main category involves any project characteristics, standard elements and proposals that were discussed in the sample of IASB meetings. Subcate-gories relate to scope and timing of the project, definitional issues (e.g., contribution-based promises), recognition, measurement and presentation of pensions, disclosures and several quick-fix issues.

2) Arguments: The second main category reflects arguments that were brought forward in discussing and justifying project elements. We classify these into three subgroups.

First, we identify conceptual arguments that were commonly used in different con-texts, e.g., usefulness. Second, we cluster specialised arguments that were closely tied to individual project elements. Finally, we identify arguments of consistency that point to any related IAS/IFRS or to ongoing projects that were taken into account while drafting the amendments to IAS 19 (2011).

3) References: Our third main category links individual statements to information sources whenever they were explicitly revealed by participants (e.g., citing comment letters).

4) Governance: The last main category codes data that relate to any technical and organi-sational matters in our sample meetings. The respective subcategories reflect board votes and voting results, internal policies and requests for input on unresolved issues.

APPENDIX A provides a detailed breakdown of the category set, including category defini-tions and further explanadefini-tions. Inductive coding yields 1,993 codings in total. As we address different aspects of board meetings, individual statements may be coded with more than one category.