Influence is conceptualised here as control over a policy outcome (Dür & De Bièvre 2007b: 3). What specifically constitutes a policy outcome differs across the policy cycle. At the agenda-setting stage it means an issue that groups have been promoting is raised to the EU agenda (or is kept out of it if this was the intended goal). At the policy formulation and decision-making stage, it implies that adopted policy or decision reflects a desired outcome which interest group lobbying has intended to achieve. At the implementation stage, an actual implementation of policy again reflects a desired outcome, whereas if groups are involved in evaluation, control over a policy outcome would imply that their inputs were reflected in the policy evaluation exercise which
feeds back to the earlier stages (agenda-setting or policy reformulation).
In order to measure interest group influence, I apply methodological triangulation as suggested by interest group researchers (Dür 2008b; Helboe Pedersen 2013): I combine process-tracing and assessment of attributed influence. The combination of two methods helps to correct biases of each of them, at least to a certain degree. A similar model has been used by Arts & Verschuren (1999) who combine assessment of influence by non- state actors (ego-perception) with cross-assessment by policymakers (alter-perception) and researcher’s analysis.
Through process tracing, researchers “try to uncover steps by which causes affect outcomes. With respect to measuring interest groups influence, scholars scrutinize groups’ preferences, their influence attempts, their access to decision-makers, decision- makers’ responses to influence attempts, the degree to which groups’ preferences are reflected in outcomes and groups’ statements of (dis)satisfaction with the outcome” (Dür 2008: 562). The process tracing method is suitable for small-N studies (Cowles 1995; Neuman et al. 2006; Kautto 2007; Joachim & Dembinski 2011) as this is a time consuming exercise which requires good knowledge of a case, including all factors which may have impacted upon a policy outcome. Process tracing also requires thorough data collection to provide accurate and detailed information to uncover all steps of causal process (Dür 2008: 563). For example, a study on tobacco industry lobbying against strict advertisement regulation relied on confidential internal industry documentation which had been disclosed in a legal case (Neuman et al. 2006). Other studies have been based largely on interviews (Kautto 2007; Bugdahn 2008; Joachim & Dembinski 2011). The key advantage of this method is that it helps us to open the black box of policymaking and show how groups lobby. The primary disadvantage however is
that it does not allow for the establishment of the degree of influence, as one can usually only assess whether a group was influential or not. It is also easy to underestimate interest group influence for those actors who did not engage in any active lobbying (Lowery 2013).
To assess the degree of influence, I apply the attributed influence method. This has been applied in both large and small N interest groups studies in which data have been collected mainly through surveys (Egdell & Thompson 1999; Mahoney & Baumgartner 2008; Heinz et al. 1993; Dür & De Bièvre 2007a; Helboe Pedersen 2013; Binderkrantz & Rasmussen 2015). Applying this method, I ask interviewed groups to assess their own influence, namely to what degree they achieved the intended outcomes planned in their strategy (rephrasing the question by Heinz et al. (1993) “To what extent did you achieve your advocacy objectives: all objectives, most, about half, few, or none”). I view influence as the achievement of a group’s objective as a result of the group’s intervention, which implies that a policy outcome should be more in line with the group’s objective than would have been the case had the latter not intervened (Arts & Verschuren 1999: 413). Thus, primarily groups which have mobilised for lobbying are approached for interviews. However, it may still be difficult to distinguish fully goals which were achieved purely due to luck – “the intervention of others, due to external events, or due to autonomous developments” (ibidem) – rather than due to the group’s intervention. By asking groups to explain factors beyond their advocacy success and by introducing cross-assessment by policymakers, I hope to obtain a clearer picture of what has enabled advocacy objectives to be achieved.
The literature speaks of the expansiveness bias evident in expert interviewing, whereby less powerful actors tend to inflate their status while more powerful actors tend to
underreport their influence (Beyers et al. 2014). There is however little evidence that this occurs in practice. Triangulating survey and documentary-based measures of preference attainment, Helboe Pedersen (2013) finds that resourceful groups tend to overstate their influence and activity slightly more than other groups (public interest groups and other types of citizen groups), nevertheless, she finds no clear difference among resourceful groups (e.g. business groups vs. trade unions, for instance). She explains that interest group respondents may lack precise information on their success, providing general evaluations of their levels of influence and suggests posing more concrete questions about specific activities and successes. I take steps to integrate this advice into the process of preparing interview questions and while carrying out the interviews themselves.
Furthermore, to mitigate the bias of self-estimation and the fact that advocacy groups may find it difficult to attribute policy outcomes to their lobbying efforts, I also complement their answers by the cross-assessment by policymakers (who are asked to name lobby groups which have been most influential in their advocacy and which have provided most valuable input). Policymakers may still be tempted to systematically overrate the influence of highly visible actors (the attractiveness bias, see Beyers et al. 2014). But careful phrasing of interview questions (for example, eschewing potentially sensitive words as suggested by Beyers at al. (2014)) and cross-validation of interview findings with interest groups and policymakers should help to mitigate the measurement bias.
Another limitation of the attributed influence method is that it measures perceived influence as opposed to actual influence (Dür 2008: 566). Such a measurement is termed ‘subjective’ as it relies “on policy actors’ subjective assessments of the extent to
which they actually attained what they wanted” (Bernhagen et al. 2014: 204, italics in original). While interest group scholars increasingly measure influence as preference attainment spatially, arguing that it is more ‘objective’ (ibidem), this method cannot be applied here in full. First, it is suitable when lobbying takes place through one venue and through one strategy (for example, the Commission and participation in online consultations during the policy formulation stage as in Klüver (2013)), while this study is exploratory in the sense that it attempts to examine who interest groups lobby when attempting to influence EU foreign policy, and how they do so. Moreover, many EU foreign policy documents are not publicly available, consultations with interest groups are not formalised and there is no analogy to the Commission in the legislative making process which is responsible for drafting and possesses knowledge on the positions of decision-making actors and interest groups. It is thus extremely difficult to find a valid primary source for positional data necessary to measure influence spatially.
Given the lack of publicly available (or disclosed) information on foreign policy decision-making and the dominance of informal communication between advocates and policymakers in this area, the study draws on interviews as a source of primary data on aspects of the political process, lobbying and influence. Whereas data collection through interviews introduces a number of biases which are difficult to eliminate completely, I follow the advice of experienced interest group scholars who believe “that a judiciously prepared and well-conducted set of interviews can contribute significantly to research on political mobilization, strategies and policy influence” (Beyers et al. 2014: 177). The primary data have been collected through 161 semi-structured interviews with interest groups representatives, policymakers and expert observers conducted by the author between May 2013 and January 2015 in Brussels, a number of EU member states
and the neighbouring countries covered by each case. The interviews were conducted face-to-face, and only in a few cases was interviewing done via Skype or telephone. The average interview length was 60 minutes. All interviewees were offered the principle of anonymity in order to allow for more open and frank answers. In order to further enhance rapport with the interviewees, interviews were not voice-recorded, instead written notes were taken down and typed soon after the interview.
My data collection strategy has involved interviewing representatives of all key groups advocating on the case and all key policymakers (representatives of key member states in the Council, key MEPs or their staff, Commission and EEAS officials responsible for the dossier and contacts with outside actors). Initial potential interviewees were identified through background research based on the analysis of secondary (specialist literature on each case) and primary resources (documents produced by interest groups, their websites, media reports, interviews with independent experts in each policy area). Further potential interviewees were added to the list through the ‘snowball effect’ (by asking groups to name other groups who lobbied on the same issue and policymakers whom they target and by asking policymakers to name other policy actors and advocacy groups).
In addition to interviews, documentary analysis (policy documents where publicly available, minutes of parliamentary debates, position papers and reports produced by advocacy groups) and analysis of media reports (English language online media specialised in EU politics, such as EUObserver and EurActiv, and national media covering the countries in focus) were employed. These data were analysed in preparation for interviews and afterwards to corroborate interview findings.
Conclusion
This chapter has introduced main concepts, theoretical assumptions, research design and methodology. In this thesis, I apply the definition of influence which has been predominantly used in the interest group literature and which regards influence as ‘control over policy outcomes’. When assessing interest group influence, I also follow the rationalist assumptions that actors have clear and stable preferences that do not change in the course of interaction with other actors and act to maximise their utility. Such conceptualisation of influence mostly refers to the first and second ‘faces’ of power which are decision-making and controlling the agenda, with the second ‘face’ being more difficult to observe. Despite possible criticism of such a definition of influence on theoretical grounds, an advantage of this conceptualisation is that it is accompanied by a developed methodology for measuring influence which has been widely used and referenced in the literature on interest groups both in the EU and the US. Moreover, without downplaying the importance of socialisation underlined by the constructivist views on lobbying in EU foreign policy, I argue that the rationalist account has the potential to provide a more complete picture not only of interest group influence, but also the reasons behind lobbying strategies and tactics.
This thesis focuses on the role of institutional factors in determining lobbying strategies and interest group influence. From the literature on lobbying in the EU, we know that some EU institutions are lobbied more than others due to their powers in the policy process and openness to input from interest groups. However, the literature mainly draws on the roles of EU institutions in policy areas under the Community method in which the Commission sets the agenda and the Council and, increasingly, the Parliament co-decide within the structured legislative process, resulting in a clear policy outcome –
a piece of EU legislation. In many areas of EU foreign policy, policymaking is still dominated by intergovernmental arrangements. The most lobbied EU institutions – the Commission and the Parliament – have little role to play, whereas the members states control the agenda, decide through consensus and implement the resulting decisions. Moreover, there are specific EU foreign policy institutions set up to assist in intergovernmental cooperation and to ensure inter-institutional coherence in managing the EU’s external action.
Few insights on how interest groups interact with institutions under the intergovernmental method were presented in the Chapter 1. From the case study of advocacy on a CSDP policy issue (Joachim & Dembinski 2011) we learnt that the intergovernmental set-up does not prevent group involvement, but affect their advocacy strategies and success. Non-state actors act at both EU and member state levels and their advocacy is facilitated by the presence of EU level norms and powerful allies among the decision-makers. They can on occasion be successful in achieving their advocacy goals (ibidem). But one case study presents a limited story. It does not allow us to examine whether interest groups’ strategies and influence have been determined by institutional factors as opposed to, for instance, issue characteristics. Thus, I will conduct a comparative study to shed more light on the issue of whether variation in the institutional conditions of lobbying leads to varying advocacy strategies and influence. I will test four hypotheses:
H1: On the continuum between the Community method and the intergovernmental method of decision-making in EU foreign policy, interests groups are likely to achieve higher degrees of influence for an issue which lies closer toward the Community method; and are likely to achieve lower degrees of influence for an
issue which is situated closer to the intergovernmental extreme.
H2: If interest groups have allies among key decision-makers, they are likely to achieve higher degrees of influence.
H3: The intergovernmental method of policymaking mostly favours groups that defend the status quo rather than promoting change.
H4: Groups tend to adopt strategies targeting actors at both EU and member state level, even when policymaking is closer to the intergovernmental method.
Key independent and dependent variables are summarised in Table 3. Three cases have been selected for this study to present a variation on the key explanatory variable – type of policy regime. Selection criteria have been applied to ensure that cases deviate as little as possible on issue-level characteristics that can also be responsible for variation of advocacy strategies and influence. The impact of group-level characteristics (e.g. material and nonmaterial resources) will be controlled through a within-case analysis. Aside from testing the main hypotheses through the cross-case analysis, I use the within-case analysis to generate more insight into groups’ strategies and influence on EU foreign policy, including key factors which condition it.
Table 3. Key variables
Independent variables Dependent variables Policy regime
(Community method◄-►Intergovernmental method)
Allies among key decision-makers (presence/absence)
Degree of influence (advocacy objectives achieved)
Type of influence (status quo vs. policy change)
Type of strategy (multi-level vs. national-level)
measure interest group influence. The latter method grasps influence as it is perceived by groups themselves (self-assessment) and by policymakers (cross-assessment). The former method helps to establish a broader picture of the policy process in which decision-making and advocacy efforts take place and to analyse the effect of institutional conditions on lobbying as well as other conditions that may determine policy outcomes.
Given the scarcity of the literature and limited number of publicly accessible policy documents from which positions of decision-makers and groups can be extracted, the study draws extensively from 161 semi-structured interviews with interest group representatives, policymakers and expert observers conducted in Brussels, EU member state capitals and third countries. Analysis of EU policy documents and documents produced by groups and media articles allows us to supplement and cross-check where possible information provided during interviews. The strength of a methodology relying on interviews is that it allows us to ask interest groups directly about their influence and how they perceive and assess it as well as to see how interest groups’ advocacy is seen by policymakers. The weakness is that it introduces inevitable biases on how facts are presented and interpreted by interviewees.