Keywords

Introduction

How can contacts for elite interviews be established? How can interviews be secured with politicians when they are too busy to answer their emails? Which type of interview questions are likely to be effective for generating new knowledge and information? How can parliamentary ethnography be practically conducted? What needs to be considered in an ethics review for the undertaking of qualitative interviews and ethnography in a parliamentary environment? What are the specificities of interviewing parliamentarians and staff members in the multilingual context of the European Parliament?

This chapter addresses the above questions by focusing on how we gathered data through research interviews and ethnography in the context of the European Parliament. We draw on extant research employing qualitative methods, as well as our own experiences in conducting a large-scale qualitative study in the European Parliament. First, we provide insights into the meaning and practice of an ethical review in the context of researching the gendered practices and policies of the European Parliament’s political groups. At its best, an ethical review can help researchers to clarify a number of practical issues, as well as ensuring that the research is ethically sound. Second, we discuss strategies for the selection of interviewees, strategies for contacting them and obtaining consent for the interviews. We go through the preparatory work required for securing elite interviews and outline our research interview questions and the choices we considered in drafting them. Third, we provide a practical overview of doing ethnography in parliaments. We show how we gained access to the research field, as well as providing a practical explanation of the ethnographic practices we used in the field.

Ethical Review

As with any research project, one of the very first steps was to undergo an ethical review with the funder (European Research Council) and with the host university (Tampere Region Ethics Council). Prior to the task of data gathering, this helped us clarify not just the ethical issues about how to proceed but also the practicalities of producing all the documents we needed for elite interviews and the parliamentary ethnography. Overall, the purpose of any ethical review is to protect the well-being of the research participants and to prevent abuses by scientists (Wassenaar & Mamotte, 2012: 268–269). Importantly, ethical reviews ensure public trust in the integrity of the research process (Bond, 2012: 102).

From the point of view of the ethics review, the first thing was to explain that the interviews with the Members of the European Parliament (MEPs) and parliamentary workers were elite interviews, considered as such because of their privileged position as politicians and workers within the European Parliament. We stated that the interviews were based on voluntary informed consent, confidentiality and anonymity. At the time of the initial contact, and prior to the actual interview, we outlined our commitment to a research procedure whereby the usage of the research data was anonymous, solely for the purposes of this project, its publications and wider dissemination. It was explained by the interviewer to the interviewee that the data would be stored securely and in line with recommended procedures.

In a multilingual setting such as the European Parliament, misunderstandings over language could compromise informed consent if not all the interviewees understood English. We made sure that the interviews we conducted in English were with participants who had high levels of spoken English and comprehension. In some cases, they asked to either see the interview questions beforehand or to bring a personal assistant or a colleague to assist with the interview in the event that clarification was required. These were of course allowed for. We also approached interviewees and conducted interviews in Finnish, French, German, Italian and Polish in which case the information was translated into these languages.

All interviewees were promised full anonymity. An informed consent form, which was signed by the interviewee and interviewer, detailed this issue, though this was only part of establishing informed consent. We have included the form as Appendix 1 as a reference for our project, but we are aware that different universities have their own procedures and templates with regards to ethics reviews. We also drafted the information sheet about the project, see Appendix 2, which detailed the purposes of the research, the conditions of full anonymity, that usage of the research data was solely for the purposes of this project, its publications and dissemination and the secure storage of participant observation notes.

We explained that interviews were to be recorded unless the interviewee requested otherwise, in which case notes would be taken. This arose in the actual interviews, where we had around 5 interviews out of 140 where we took notes instead. We clarified in the ethics review that interviews were to be conducted in public locations chosen by the interviewee or in their offices. The interviewees retained the right to withdraw from the interview process at any time, and to stipulate that the interview data was not to be used. The confidentiality criteria also included a statement that interview content was not to be discussed with any other participants. Given that the interviews might concern gender equality practices and informal institutions that could hinder or advance gender equality within the political groups, we were sensitive to matters of gender discrimination, racism or sexual harassment arising in the interviews (see, e.g. Muasya & Gatumu, 2013). All the interviewers would have information about the European Parliament’s procedures for such cases, as well as contact details of relevant public or voluntary organisation support services which could be offered after the interview.

One of the trickiest ethical questions to address was how to acquire informed consent when carrying out a parliamentary ethnography. The first matter of note was that participant observation would occur in many locations, including plenary meetings of the European Parliament, different public events organised by the political groups, and public committee meetings. We explained that many of these, especially the plenaries and the committee meetings, were public events and as such were commonly televised or recorded. When undertaking participant observation in these places no personal data (e.g. names, nationality) was to be recorded.

The second measure was that we were committed to the principle that our research was to be conducted openly. The European Parliament was to be informed about the ongoing research and an information sheet was to be available to those chairing meetings and all participants. The researchers conducting the participant observation would also offer to discuss the purposes of the research project with the political groups and the committees. To that end, we undertook to organise an opening seminar in the European Parliament in Brussels to disseminate information about the ongoing project and our findings in the final stages of the research. We stated that an open approach to participant observation will also enhance the commitment of the institutions and actors in question. Subsequently, the opening and closing seminars were organised on 30 January 2019 and 7 February 2023, respectively, in the European Parliament in Brussels with the participation of our project researchers, other academics, our expert board members, MEPs and staff.

Thirdly, ethnographic research was also to be conducted in some closed meetings of the political groups or subgroups of the European Parliament. In such cases, we proposed the following procedure: the researcher would negotiate with the person in charge to obtain a permit to undertake participant observation, and amongst other things, explain the ethical principles followed in the research and make the information sheet available to all participants. The leader would inform the group about the ongoing research and the group could potentially discuss it. In such a way, everyone who participated in the meeting would have been informed, and nobody would have been under observation without knowing about the research.

Notwithstanding this comprehensive approach to ethics, full informed consent is difficult to meaningfully achieve in ethnographic research. Specifically, total anonymisation is difficult and if participants provide data on third parties, these third parties cannot give their consent. However, the opening project seminar allowed us to establish a presence as social researchers in the parliament, and to a large degree making the ethnographic research both noticeable and accessible to the participants (Laube, 2021). We also showed the ethnographic observation protocol on one occasion of seeking access to a political group meeting and shared interview research questions with participants in advance. Furthermore, informed consent was not regarded as a one-off agreement. As a project, we frequently reflected on how we could establish and sustain our presence as social researchers of the European Parliament. For example, we maintained an active Twitter account and a frequently updated project website.Footnote 1

In addition to these project-specific ethical principles, we committed not to gather or record any personal data (e.g. names or identity traits) and to follow the duty of not harming participants (e.g. by disrupting political careers). We only kept track of aggregate data that was already publicly available (like nationality or gender) for our own records, and presented it in our publications in general ‘meta’ terms. We undertook to continually evaluate the risk of identification by removing any indirect identifiers (e.g. nationality, occupation, age) prior to the publication of any citations. The ethics review also included a detailed data management plan that described how we would keep transcribed interviews and fieldnotes securely and confidentially on the university drive.

Overall, during the whole research process, we noted the increased importance given to ethical considerations, not just by funding bodies, such as the European Research Council, but also by some high-ranking academic journals, such as the American Political Science Review for whom we were required to submit a full overview, and statements of our ethics review, when our article was accepted for publication.

In recent years, academia has witnessed debates about whether strict ethics reviews are making qualitative research particularly difficult. For instance, Fouché and Chubb (2017) in their literature review on ethics reviews and research involving participants, demonstrated that the criteria used for ethical review have been slow to adapt to the emergent and participatory nature of this research. This has resulted in researchers reporting negative attitudes towards, and experiences with, review boards and ethics review processes. Social science researchers’ negative experiences with ethical reviews can be attributed to ‘time delays involved in obtaining ethics reviews’ due to the infrequent convening of university ethics committees (Wassenaar & Mamotte, 2012: 271). Yet, as we have outlined above, there are strategies for managing the ethical review that can be utilised that we recommend in order to conduct participatory research in an ethical manner.

Finally, the EU General Data Protection Regulation (GDPR) which came into effect in May 2018, obligates any organisation who ‘target’ or collect data related to people in the EU to adhere to specific rules. Because the regulation is admittedly large, far-reaching, and fairly light on specifics, compliance with it can become a daunting prospect, the EU launched a dedicated portal with guidance on compliance (https://gdpr.eu/). Since 2018, GDPR has had an impact on ethical questions for interview-based research or ethnography as well as the scope of scientific choices available to researchers. Within the context of the GDPR and the sensitive status of personal data, there is a concern that academics will increasingly restrict their research choices to safer options. This may involve, for example, reusing anonymised datasets, selecting populations based on expediency rather than theoretical appropriateness or outsourcing fieldwork to professional data collection companies (Molina & Borgatti, 2021: 13). Scholars argue that the social sciences are particularly vulnerable to the intersection of ethics reviews and personal data collection, given the methodologies involved in the research field (Molina & Borgatti, 2021: 13).

Whilst the GDPR puts more demand on obtaining valid consent from research participants and is rooted in ethical principles to safeguard fundamental rights, Molina and Borgatti (2021: 18) call for the establishment of more specialised social science ethics committees to address these ethical dilemmas facing the social sciences in the new digital age. At the outset of our research, the GDPR was just beginning to be implemented and both researchers and universities were not yet aware of how to enact the requirements. In the subsequent years, most institutions have established rules and procedures regarding GDPR and participant research, making it necessary to add GDPR compliance as a step in a comprehensive ethics review.

In conclusion, there is a constant need to commit to high ethics standards, with a very clear picture of what research is actually planned, and to conduct research according ‘to the rules’ established and agreed on with participants at every stage of the process.

Doing Interviews

The clear, principled and relatively straightforward world of the ethics review confronts a messy reality when interviews and ethnography actually begins. This messy reality is marked with constant negotiations around getting agreements for interviews from and actually interviewing, very busy people who can sometimes dedicate only 15–20 minutes in the corridor, to a researcher who wants to talk to them in a silent office for at least an hour to cover all the important interview questions, whilst conforming to the ethics requirements (see Box 4.1).

The initial step prior to beginning the interviews is to imagine the sample—ideally, who do you want to interview and why? For us, it was primarily MEPs, across all genders, all political groups and from different member states that were the main interviewees. We also wanted to interview parliamentary staff who are employed within, and by political groups, which added scholarly and practical benefits to our work and strengthened our research findings. The insights from these interviewees significantly helped our conceptualisation and analysis of power relations, informal institutions and gendered norms. Moreover, we heard the voices of those who have ‘less power’, but might nonetheless exhibit less restraint in talking about what they have observed, vis-à-vis parliamentary relationships, more acutely and/or from a different angle. In scholarly terms, as we argued in Chapter 2, there is a real gap in researching parliamentary staff. We interviewed many Accredited Personal Assistants (APAs) of MEPs, of whom it may be argued, were the most vulnerable category in our research material. Two of our research topics—sexual harassment and racist practices within the parliament—would have very much benefited from research interviews with caterers, cleaners and care-takers in the parliament, who had, unfortunately, fallen outside of our planned interviews and the ethics review.

Secondly, we wanted to include interviewees from different hierarchical positions to get a full sense of the gendered relations and practices of the political groups. We were not concerned with how representative certain statements or views were, which would have involved counting very carefully the number of interviews, and by whom, in which such statements were made, and ensuring that the sample was representative (see Goplerud, 2021). Rather, we sought to establish the qualitative dimensions of issues such as informal decision-making institutions, which underpinned our interest in talking to people in different positions. The highest echelons of power in our interview data were represented by the leadership of the political groups and administrative leadership, namely, the (Vice) Secretary Generals. We also targeted people across committees and in different positions within them, including the powerful positions of chairs, coordinators, rapporteurs and shadow rapporteurs. In addition to the above, we also wanted to ensure that those we interviewed were of different ages and stages of their political careers.

Even though our research was about gender equality, it was important that we did not target only those MEPs and staff who had shown interest in, or commitment to, gender equality during their political or work careers. To be more precise, we wanted to interview ‘everyone’—not just women or feminists—but also men, and those who were not interested in gender equality or even opposed it. For us, this was a way to go beyond the ‘usual suspects’ in gender and politics research, and to locate discourses and practices of gender equality within political groups (see also Elomäki & Ahrens, 2022). This was aided by the parliamentary ethnography that went beyond, and renewed, ‘the usual’ pool of participants in studies of parliaments.

We did, however, (with mixed success, see Box 4.1) want to target some key actors in the setting who would have a greater knowledge about gendered practices and policies than others. These included the coordinators of the political groups in the Committee on Gender Equality and Women’s Rights (known as the FEMM Committee), political group staff working for the FEMM committee, as well as gender mainstreaming administrators from all committees. For the specific policy issues—economic policy, social rights pillar and violence policy—it was essential to interview MEPs and staff who were working on these topics, on relevant files or in relevant committees, to enhance our understanding of policy processes.

Box 4.1 Reflections Post-Interview with Actors Involved in Gender Equality

‘It was a corona interview and I had 20 minutes negotiated between appointments. (…) but because of this short time, it felt that the interview lacked a narrative to it as I had to keep jumping between topics to cover the specific policy questions. Some of her interview performance betrayed a confusion around the proliferation of different structures, initiatives, and actors dealing with gender mainstreaming, especially when she is a usual suspect to be part of the initiative—her staff member was present and reminded her which gender initiative it was and her (non)role in this case: “you were sat in the room with them, but not part of the meeting”. It perhaps shows even more who conducts the labour of being a gender equality actor. Affects and researcher role—the role of laughter in inappropriate situations—finding ignorance of gender simultaneously bizarre and funny, but also serious—this reminds me of Hochschild’s “feeling rules”. Feminists are not supposed to find the dire situation of gender and institutions funny, but sometimes they do’ (EUGenDem research diary 9 Apr 2020).

Strategies for Recruiting Interview Participants

Several recruitment strategies for qualitative research are developed in the literature including probability sampling, maximising range sample, comparison cases and so on (see Weiss, 1994 for more). We combined convenience sampling with snowballing.

For us, the recruitment of participants started with convenience sampling (Weiss, 1994). In our case, this meant reaching out to actors we identified as allies. We understood allies in a broad sense, as anyone we had already made contact with, anyone potentially interested in our research topic or anyone with a similar nationality to ours. In this regard, our feminist and European networks helped greatly. Many of our participants were recruited based on their expertise on a specific issue. For instance, when analysing a policy issue, we targeted MEPs that were members of the committee in charge of it (e.g. MEPs in the Committee on Economic and Monetary Affairs when analysing economic policies). Reaching out to MEPs for an interview by appealing to their expertise, triggers their interest and increases the chances of receiving an answer. This strategy demands necessary and ‘ultra-important’ (Lilleker, 2003) preparatory work to find such expertise and to learn about the participant’s background.

Overall, across the four years the data collection took place, we contacted around 500 people, which resulted in the 140 interviews that formed the core part of our data. Over the course of sending the emails, we learnt the importance of mixing a short official style—including mentioning the funding body and the university affiliation—with some personal details. Personal hooks included stating the reasons why we contacted this person in particular: on someone else’s recommendation (i.e. a colleague in the European Parliament) because they worked on a particular file (i.e. they were an expert) or they held a particular position we were interested to know more about (i.e. a leadership position). Because MEPs and staff are very busy, and not always physically present in the European Parliament’s locations of Brussels and Strasbourg, it was important to mention that the interview could be arranged whenever, and wherever, they saw fit.

Some participants were also recruited by researchers with similar nationalities. Thanks to having a team of six researchers from five different nationalities (Finland, France, Germany, Poland, the UK), we used our respective nationalities strategically to recruit MEPs. In the supranational context of our research, appealing to regionality and local references helped in building connections with MEPs and thus in augmenting our response rate. We found it useful to stress regional similarities in our invitation emails, for instance, by mentioning attendance at the same schools/Universities or similar citizenship, which tends to be particularly decisive for small member states like Nordic or Baltic countries (e.g. by being Finnish). However, we found out that this strategy remains context specific. Whilst it worked well for the Nordic countries, it did not result in any success with Polish or French radical-right groups.

As seen in Fig. 3.3 in the previous chapter, we obtained the most interviews based on linguistic and citizenship affinities—we had a high number of British participants, but also Germans, and Finns. It was easier to obtain interviews with participants who spoke English, e.g. we had more Irish interviewees than we could have expected based on the proportion of parliamentarians relative to member states.

For some nationalities, it was not possible to do interviews in English. For example, many Polish or Greek speakers were unwilling or unable to answer in languages other than their own. Some potential interviewees were particularly hesitant because they did not speak English at all and the arrangements for potential translators or interpreters would be too much hassle. To counteract this, we offered to do interviews in participants’ native languages whenever we could. The composition of our team meant that we could offer interviews in English, Finnish, German, French, Italian and Polish. When asked, we also provided the interview questions beforehand in the native language and we also provided a printed interview guide in a clear accessible format, for participants who wanted to read particular interview questions. We also had several experiences, with Italian participants who were happy for interviews to be conducted in English, with one participant’s assistant boasting that the MEP could equally do the interview in Italian, English, Spanish or French. The language politics of the European Parliament suggest an elite English dimension to relationships within the institution context of the willingness and ability to use English structured along North–South and East–West lines (Ringe, 2022).

The above-mentioned convenience sampling strategies increased the likelihood of getting an answer and establishing entry points to the European Parliament. Thereafter, we continued with a snowball sampling strategy by asking for recommendations after each interview to expand our pool of participants (Weiss, 1994). However, as Weiss (1994: 29) pointed out, we were aware of the pitfalls of snowballing strategies, especially the potential to under-represent the experience of those with little or no social contacts or networks.

Not only did we write emails according to the expertise and nationality of potential participants, which in some cases meant translating the email, but we also had to deal with conservative/radical-right populist actors. For the latter, we emphasised ‘equality’ more broadly as a focus of our research about ‘men and women in the European Parliament’—as opposed to ‘gender’, even though these terms did not always reflect the normative foundation of our research. Although we cannot expect policymakers or non-academics to understand theoretical distinctions between different terms, this choice sometimes felt like self-censorship or self-imposed restriction.

As a rule, if a participant was not available for an interview we would ask if they could recommend potential participants in the same way as we did at the end of an interview. Our post-interview note also had a section on follow-ups that included a request for information about possible interviewees or research materials. Beckmann and Hall (2013: 200) present a useful concept of ‘informed probing’ at the end of the interview stage by adopting a more direct approach if familiar names are volunteered, and asking more generally about adjacent participants, for example, ‘what about potential participants in X field?’. According to them, this allows for the building of further connections with the next participant in the invitation email, which they refer to as a ‘name-drop’ to elicit ‘toehold respondents’ (Beckmann & Hall, 2013: 201).

Once emails are drafted and sent, there is no guarantee that they will be answered. Therefore, both patience and persistence were required. We followed up the interviews with phone calls. In the European Parliament, phone numbers are usually displayed on the MEPs’ webpage, and assistants can be contacted by changing one digit only (e.g. a 5 becomes 7). Due to their busy schedules, some elite interview participants prefer to handle demand with rapidity. For example, as MEPs and their staff are highly sought-after by various actors (i.e. lobbyist, researchers, national policymakers, grassroot organisations), they may be more receptive to immediate phone call demands, rather than to emails that can be archived or are easily lost amongst the masses received every day. In addition, when in the parliament, we also used door knocking and simply approached parliamentary actors on the spot.

Because we sought the widest representation of participants in the European Parliament as possible, we developed an activity log in which we documented and kept records of whom we contacted and recruited. The log helped us to ensure our pool of participants reflected a gender balance and fairly represented the range of political views in the Parliament. Each contacted participant was thus registered in the log with information about their gender, associated political groups, and also the date of the first invitation emails sent as well as follow-up emails or phone calls. The log was a key element in the development of a successful recruiting strategy. In this regard, systematically followed up on invitations, shared the workload with members of the team and assigned tasks of contacting some participants to those familiar with specific languages. This log was stored completely separately from the coded data and the metadata, ensuring that no crossover or identification could be made.

Preparing for Interviews: Preparatory Work, Locations and Interview Questions

Elite interviews are often hard-won, short and resource-intensive (including the participants’ time). Therefore, preparatory work for each participant is essential to know their parliamentary and work biography; policy expertise, voting record and public statements. This allows for smoother follow-up questions and ensures that the interview material is not a duplication of publicly available data. Targeted follow-up questions allows researchers to steer the conversation beyond ‘pat answers’ (Chappell, 2020). There can be some risks related to preparatory work, not least over-determining the interview; anticipating answers in advance and ventriloquising participants. It also has workload risks on scholars less acquainted with the fieldsite, its political developments and participants or on early career scholars who feel the need to make expertise visible in the interview (although the benefits of the naive, unthreatening researcher, may be overrated).

At times, the activities of the European Parliament can feel overwhelming. Actors are busy and constantly moving from one city to another. The calendar of the European Parliament, known to all in the ‘Brussels bubble’, gives important information about the organisational structure of the parliament and its activities. The calendar, released in advance each May and voted on in a plenary, has eleven plenary sessions that must be held in Strasbourg. The working months are divided into four weeks; a pink week for Committee work, a blue week for Group work, a red week for Plenary work and a green week for Constituencies (Busby, 2013: 99). Each Friday is reserved for constituency affairs. In that sense, MEPs are constantly expected to travel between constituencies, Brussels and Strasbourg—making their time ‘a precious commodity’ (Busby, 2013: 146) and the calendar an important tool for scheduling interviews. For the parliamentary ethnography we undertook, the weeks when MEPs were in their constituencies—if they were not on delegations—were a good time for conducting interviews with staff and online. Furthermore, the first green week in February was used to contact some UK members who had left the parliament, since the experience of Brexit was still fresh in their minds.

With regard to the locations of the interview, several issues had to be considered. We asked participants to propose a location that was both accessible and comfortable for them. This mattered especially with disabled participants where further accessibility requirements could be made such as sign language skills (Evans & Reher, 2022: 700). Many matters were considered when arranging interviews, such as guaranteeing anonymity, the opportunity to record, safety considerations regarding Covid, as well as practicalities like allowing for enough time to get through security.

Our interviews were semi-structured. Semi-structured questionnaires correspond to adopting a conversational format with open-ended questions for in-depth interviews (Soss, 2006). Compared to structured interviews, they allow some leeway to follow-up on whatever is deemed important by the researcher (Brinkmann, 2020). To guide the interviews we developed a joint interview guideline, which we revised after the pilot study and then again two years into the project to reflect new research agendas, such as the impact of the Covid-19 pandemic on the work of the MEPs and staff. Time with the interviewees was always constrained. Consequently, we dropped some interview questions that we felt were already covered, including sexual harassment, a topic on which we published early in the project (Berthet & Kantola, 2021). We had a separate interview guideline for MEPs and staff, and separate interview questions for the topic of Brexit and for specific policy fields, such as economic, social and violence policy.

We opened the interviews with a very short background question. Such questions are often considered important as they can warm up the interviewee, relax the interview situation and build trust. Initially, we had two rather general questions: ‘In which positions, and for how long, have you worked in the European Parliament? Who and whose interests do you see yourself as representing in European level politics?’ After the pilot study, we quickly dropped these due to time constraints for us, and the amount of time it took the MEPs to answer these. Instead, we opened the interviews with a question that we thought took us directly to issues that would provide better insights into our research questions: ‘If you were to advise a new MEP on how to be an effective MEP in your political group, what would you stress? What’s important to understand about the group?’. For us, these questions provided insights about the informal institutions and hidden power structures of the political groups.

Our initial interview guideline included well-balanced parts, and the questions evolved over the course of the project and as our own knowledge increased they became more focused. For example, we dropped a whole section dealing with national politics and elections as we realised that this was not providing insight to our research questions. Instead, most interviews became focused in a very detailed way on the political group level, where insights on the democratic functioning of the political groups in terms of gender equality could be acquired. By the end, we had reduced the interview questions to practically one part relating to this.

When we asked about the MEPs’ everyday involvement with the political group, we were attentive to their descriptions of decision-making in the political group; what they thought makes an effective group leader and whether gender played any role in this. We also asked about the significance of the political group meeting for the interviewee and the relations the MEP had to the group leader, the Secretary General, to other MEPs and staff. Our specific questions about gender equality followed these questions and were formulated as: ‘How would you describe your political group from the point of view of gender equality more generally? What other differences are important?’ and ‘What kind of practices for gender equality do you have within the party group? (quotas, work-life-balance, etc.)’. The question alluding to ‘what other differences are important’ was designed to seek insights about other bases of inequality than gender, including race and ethnicity, disability or sexual orientation. We were interested in asking the question openly to see what basis of inequality the interviewees would see as most relevant. However, on reflection, we did think that we should have guided the interview more strongly in relation to intersectionality and inequalities other than gender, including racism (see, e.g. Kantola et al., 2023).

We had specific interview questions about policy-making in the political groups which included: ‘How does policy-making work in your group? What are your possibilities for influencing group priorities and policies? Have you actively put yourself forward to some positions and failed/succeeded? What had an impact on your success/failure? How do you persuade people within the group, to whom do you go?’ We also asked about conflict and informal socialising within the political groups—always including the question of whether gender played a role in either. We also asked about the MEP’s working relationships with other political groups and any obstructive behaviour between them. In addition to these policy-related questions, we asked about speaking in the plenary, and what made this effective, or conversely, hampered it. To complement these, we had a list of specific questions about the impact of Covid-19 on MEPs and political group work on Brexit, Europarties, codes of conduct and sexual harassment within the groups.

With these questions and follow-up enquiries, we were tapping into the practices and informal institutions of the political groups: how they worked in relation to power; informal politics; norms; practices; gender. One of our findings revealed that it was not easy for many interviewees to talk about such matters, whilst discussing specific policies and policy content came more easily and far more naturally to them. This relates to several important points for research: gendered hierarchies and informal institutions are difficult to study because they are invisible and embedded in the very power structures which are difficult to perceive and to talk about. Parliamentarians are generally concerned about being re-elected, which is realised on the basis of the policy work that they do, not for making their political groups or parliaments work better and in a more democratic way. This is also where the value of qualitative research reveals itself most intuitively: gathering a data set of qualitative interviews where several interviewees try to talk about the same topic, will generate multiple qualitative insights about things that are difficult to talk about for individual interviewees.

This ought not obscure the fact that elite interviews present many challenges. Beyond those related to recruiting participants, scholars pointed to the difficulties of accessing knowledge during interviews with political elites ​​(Holmes et al., 2019). Many discussed the power dynamics at stake between the interviewer and the political interviewee. This includes, for instance, the possibility of receiving a ‘political talk’ since ‘politicians are used to evading difficult questions’ (Cowley, 2021: 3). Katharine Sarikakis convincingly shows that even if the researcher asks the questions, a reversal of power hierarchy can function between the researcher and the participants (2003: 423). In our case, such reversal occurred when participants asked us, for instance, to define concepts like racism. Another challenge includes the lack of trust between interviewee and interviewer during ‘fly in fly out’ interviews (Chappell, 2020: 135) that only ethnographic methods can provide a balance (Brown, 2018; Miller, 2022).

Doing fieldwork requires much data management and paperwork for both the researcher and the participant. Not only signing the consent form but also carrying around confidential paperwork. Efficiency is key, otherwise, it can put additional burdens on participants. There can be problems if participants do not sign the consent form immediately, which creates considerable additional work to print, sign and scan. In order to keep all of this safe and anonymous, we scanned consent forms to the server and then destroyed their hard copies.

Some Additional Hurdles: ‘Stressbourg’ and Covid-19

Strasbourg provided both an opportunity and a challenge for conducting qualitative research. In terms of opportunities, the long hours and informal community in Strasbourg created opportunities for dialogue in the margins of meetings and spaces to build rapport. The change in atmosphere is palpable and the culture is more contained. Parliamentary actors all stayed in hotels, rather than returning home, were more dressed up, drank Crément d’Alsace, and frequented bars and restaurants and thus there were plenty of opportunities to speak and develop a rapport in a distinct way. Arguably, we gained greater credibility as an ‘insider’ at the margins of meetings, and in the practice of travelling to Strasbourg alongside political group staff, interpreters and members of the parliamentary administration in a carpool. Moreover, in Strasbourg, political group meetings are held in the evenings, rather than in the morning or in the afternoon in Brussels, which creates later days and a different atmosphere. Political group leaders’ press conferences are all held in Strasbourg and are open to attend.

Strasbourg also provides its own distinct challenges. Busby (2013: 99), for example, has noted how some assistants referred to the Strasbourg week as ‘Stressbourg’, reflecting the long hours and hectic schedules. Assistants would often not schedule interviews with themselves, or with MEPs, for that week because many pre-booked meetings would be running simultaneously, such as with lobbyists, group meetings or the plenary. To counter this, we were able to do ethnography in a political group meeting. Helped by a senior contact, we were able to attend a political group meeting organised by the Left Group (also known as GUE/NGL), which was not recorded and published online, but was marginally more accessible than other political group meetings (Miller, 2022).

Brussels, by contrast, provided very different experiences. Some less active MEPs, or those who were not in positions of seniority in their political groups, were reluctant to host researchers, saying ‘there’s not much on this week’, for example, during some committee weeks.

Whilst the fieldwork was multi-sited in Brussels and Strasbourg, the geographically dispersed and transnational nature of the MEPs, meant that they were often present in national capitals and cities of team members who could approach them for interviews. For example, the Left Groups (also known as GUE/NGL) held their study days in Helsinki, whilst some of our interviews were conducted in London and Berlin depending on the locations of both the MEPs and our team researchers. After the onset of Covid-19, this changed yet again, and our telephone and online interviews were conducted with participants either at home or in their ‘local’ offices in member states.

Like every aspect of normal daily activities, our data collection endeavour was equally transformed by the Covid-19 pandemic and accompanying restrictions. Due to the latter affecting parliamentary activities and access to the Parliament, we had to modify our recruitment and interview strategies. A major impact on our data collection was the closing of the Parliament as a whole, which by definition ended the ethnography.

When the Covid pandemic had just begun, various changes impacted our activities. The status of the Visitor’s pass became ambiguous, and clarification had to be continuously sought about its status. Project resources and data collection very quickly had to be balanced against safety for others in the parliament as well as for the researchers. In the earliest days, events were attended, such as a Press Conference given by David Sassoli, then President of the European Parliament, about the openness of the parliament who alluded to how the key democratic functions of the institution would be retained. Eventually, however, meetings became closed to everyone except MEPs, and interviews were conducted in person and then eventually online, or by telephone call. Exceptionally, the parliamentary calendar was updated and trips to Strasbourg were cancelled. This meant that travel arrangements had to be undone at the last minute, and our ‘exit’ from the ethnography was unconventional, not least because traditional thanks could not be given, but also due to the field site itself, which became inaccessible to its members who had moved their work online.

As a project, we were forced to reassess our logistics and switch to online and phone call interviews. This generated advantages and disadvantages. Amongst the advantages, online and phone call interviews allowed for acting spontaneously and quickly, instead of limiting the schedule to when the researcher is visiting the location (in our case, Brussels and Strasbourg). These interviews are also easier to record and the presence of the dictaphone is less obvious. However, amongst the disadvantages, the interviewee can end meetings more easily, either due to bad connections (real or imagined) or to distraction from homeschooled children. In addition, these interviews create a distance which makes it more difficult to build a connection, to create a normal conversation or to observe body language. A further practical disadvantage is the difficulty to get consent forms signed and returned.

In relation to plenary debates, the Covid-19 pandemic significantly changed the setting, with debates being conducted within a hybrid format. In these circumstances, only a few MEPs were present in the European Parliament, and most contributed from their homes in Brussels or in their home countries. It was quite odd to see the MEPs in relaxed clothing, in their living rooms or separate office rooms, even in their kitchens.

Doing Ethnography

Ethnographic research within parliaments offers many advantages in the exploration of how the functions of parliaments are entangled (Crewe, 2021), especially the symbolic dimensions of parliaments such as their architectures (Verge, 2022: 1053), and how political work is differently carried out (Crewe, 2015). Ethnography reveals multiple interpretations of the same event or process, such as parliamentary work at the time of Brexit (Kantola & Miller, 2022); institutional hierarchies (Lewicki, 2017); how working worlds in parliaments are reproduced through everyday gendered relations (Miller, 2021) and what concepts such as gender, gender hierarchies, gendered relations and their redress mean in and for everyday lives in parliamentary settings (Miller, 2022). This short section shows how ethnography may be fruitfully done in one parliament. Other parliaments may differ: some are far smaller, which makes some practices more difficult, since the researcher might become more visible and collaborations with other parties might be more visible, whilst others may be larger, making a finer-grained analysis more complex.

As discussed above, our research design consisted of a pilot study to shadow nine MEPs from five political groups and a longer stay as a study visitor. This ethnography took an explicitly political focus, centring on activities, rules and practices of the political groups of the European Parliament in the so-called 8th and 9th Parliaments (the legislatures of 2014–2019 and 2019–2023). Whilst an interpretative approach was taken to the European Parliament (Geddes & Rhodes, 2018; Miller, 2021), ‘aspects of power central to politics, such as competition, conflict and interests’ (Firat, 2019: 16) were also attended to.

Here we present some strategies and tradeoffs for those wishing to produce ethnographic data on parliaments. First, there are different entry routes into parliaments where permission needs to be sought to conduct research. It should be noted that a parliament is not a unified entity, they consist of many different actors with different jurisdictions, and so any permission to enter needs to be reflexively considered, especially with regard to issues of informed consent. One entry route we used was the two-month European Parliamentary Research Service placement. This was good for establishing a base to make contacts and request attendance at meetings, though access through this route provides no guarantee of gaining a presence in the myriad political worlds of parliament (Niemi, 2010: 107–113). Prior to the placement, the researcher can use personal and academic networks to identify participants who might be able to cooperate with the research and request to attend meetings.

Another route is through parliamentary leadership, political leadership or administrative leadership. In the European Parliament, this would be through the President, Vice-President, Bureau; political group leaders and the Secretary Generals of Political Groups. We did attempt this, although it proved to be our least successful strategy for several reasons. The Parliamentary and political leadership is likely to be more defensive in order to preserve the reputation of ‘the good institution’. We quickly found out that trying to negotiate access at this level involves a meeting, the production of materials (beyond the information sheet) and an explanation of what the research is interested in. There may also be reciprocal responsibilities tied to such an opening, for example, presenting a final report or giving evidence to a committee. Entry through the parliamentary leadership may deliver cooperative benefits, but when feeding back research findings to different groups in the hierarchy, field members might feel they are being watched by the parliamentary leadership, leaving researchers in the role of reinforcing hierarchical structures of control. Finally, there is the inherent risk of institutional patronage: these leaderships might change and concomitantly jeopardise the continuation of the research.

The third, and the most effective point of entry for us, was through individual MEPs’ offices. Here, an MEP might share their calendar and find a time in the placement for an interview. Gaining access with an MEP might be more successful (in our case) if they are a feminist MEP, have a shared policy interest or have expressed an attitude or experience in public sympathetic to the research project. Strategic positionality (Reyes, 2018) matters here, and the nature of what will be ‘strategic’ will vary by parliaments. As discussed above, nationality and language played a significant role in the European Parliament, meaning a multilingual team was advantageous in negotiating access. Political party activism and regional links also helped.

Persistence with contacting the relevant office was important, and so too was stressing our flexibility. One MEP wanted to cancel a shadowing placement on the same day that we were flying to Brussels, due to the fact that there was ‘not much interesting’ happening. This was potentially problematic because we did not have access via the European Parliamentary Research Service at the time, and thus a significant part of the trip would be cancelled and resources wasted. In the end, we emphasised how we were interested in observing all aspects of the parliamentary calendar and that we would be able to use the opportunity to conduct interviews. Luckily, we also had a shadowing placement scheduled later on that week. The lesson learned was to make sure we scheduled multiple activities whilst in Brussels. It is fortuitous for those seeking to conduct parliamentary ethnographies, that MEPs hire several assistants. This can increase the capacity for hosting a parliamentary shadowing placement, though inevitably places strains on staff working days. In smaller parliaments (Niemi, 2010: 105), this opportunity is squeezed as parliamentarians have smaller staffing budgets with which to employ multiple assistants.

Arguably, the key to entry is transparency. Formally, the information sheet and the email approaching the person needed to grant access and acquire ‘informed consent’ were important. Informally, more dynamic conceptions of consent are likely to involve discussing ideas and first impressions with parliamentary actors to (1) develop ideas and (2) ensure that the participants do not forget why you are there. In some ethnographic research of parties, all elements of transparency have not been followed, such as revealing party identification when reporting the findings (Bellè, 2016). With regard to informed consent, parliamentary researchers need to be pragmatic. MEPs were busy, consequently securing informed consent at the outset of a busy shadowing day was sometimes difficult, if we were attending a breakfast meeting, or being taken to see them in committee. Our strategy was to send all the project details well in advance and try to be very clear with the assistants, so in theory, at least, the MEP and their staff can be legitimately assumed to have read all documents about the project. In practice, we recognise that this is a grey area.

Remaining consistent with the principle that parliaments are publicly elected places and should be accountable, we were acutely aware that some participants may be in a position of vulnerability, especially staff. The pragmatic realities of informed consent, as well as the normative imperatives of being a critical researcher around social justice agendas matter; together they combine to impart a professional responsibility to the next cohort of researchers. We ensured the principles of no harm and anonymity by having a trusted colleague read the work and check that no one is identifiable or compromised.

Once inside the parliament, more bottom-up research access can be found through parliamentary researchers, Trade Union members and staff in the parliamentary administration. Immersion when inside the parliament can, to some extent, reject or at least ameliorate the affects of established gatekeepers. They might also value your research insights to be fed back to them. However, it should be borne in mind that staff might lose their influence due to staffing changes and rotations of responsibility.

Ethnographic Practices

The choice of each research practice within Parliament may vary by research stage and by priorities. In terms of the degree of control over fieldwork, timing can be considered important with regard to the types of practices that are engaged in at each stage. An observation schedule is a loose itinerary of activities a researcher may wish to follow in a parliamentary ethnography, using the aforementioned parliamentary calendar. Planning can sometimes be accomplished. Nadia Brown (2014), for example, conducted a ‘focused ethnography’ of the Maryland legislature which is akin to Chappell et al.’s ‘rapid ethnography’ (2017). In a different example, Niemi’s (2010: 75) study of the parliamentary administration in the Finnish Parliament, Eduskunta, was heavily guided by her exit from the field, which involved much data generation. In terms of the ability to plan in the European Parliament, timing and access in the electoral cycle and also unpredictable situations, in this case, Brexit deadlines and Covid, affected the planning of our ethnography.

We used three ethnographic research practices: shadowing, meeting ethnography and hanging out. First, shadowing is the practice of accompanying actors throughout their daily work lives, and was useful at the beginning of the research stage (Bussell, 2020: 471), to provide a generalised overview of the kinds of activities that MEPs participated in vis-à-vis their group. In general, shadowing helps guide a researcher in a field at the beginning. Emma Crewe uses an alternative term to shadowing and talks about ‘following closely’ her local MP. She describes one of the challenges she faced by saying that she ‘had the perpetual feeling that I was in the wrong place at the wrong time’ (2015: 5). We shadowed nine MEPs from five different political groups over a period of half a day to three days. We followed throughout their day and took notes on how they related to their political group. This also gave us important access to political group meetings.

Secondly, meetings are ubiquitous in the European Parliament, and meeting ethnography is developing conceptually as an analytically distinct practice that explores what meetings do (Brown et al., 2017; Sandler & Thedvall, 2017). The type of meetings attended may depend on the type of parliament and the form of activities. For the purpose of our research project, the political group meetings, working group meetings and national party delegation meetings were of interest. However, with the Covid-19 pandemic other meetings became important too, such as President Sassoli’s Press Conference on European Parliament’s Covid-19 restrictions. We were able to observe and record in field notes how the group's ‘top brass’ were present for the Press Conference. We also recorded how attendees were filming the meeting, and how people were walking into the room later than the scheduled time, almost as if they knew the real routines and rhythms of the meeting.

Although some meetings were open in the parliament, in order to access others, we often asked an MEP’s office if we could accompany them. To prepare, we asked in advance what the substantive content and political context of the meeting was like, and either the MEP or the assistant would sometimes provide a briefing. A chair, for example, allowed us into a pre-trilogue meeting. Prioritisation is equally important. Political groups hold their meetings in the same week, often at the same time, therefore researchers must prioritise and ask themselves, what is more important, group representation, or how a particular group, National Party Delegation or actor are going to behave if they come under fire? Such deliberations are not always easy.

Thirdly, in elite settings, the ethnographic practice of hanging out offers an alternative possibility for immersion. This is a more diffused and dialogic practice than meeting ethnography. Hanging out requires three facets: ‘a period of continuous residence amid members of a field, engage[ment] in informal, ludic and sociable interactions sited outside or at the side-lines of members’ professional habitats and participa[tion] in activities where striking and sustaining rapport is as important as the goals of the research’ (Nair, 2021: 10). Densely clustered institutional spaces, such as the European Quarter, provide several immersive opportunities (Lewicki, 2017; Nair, 2021: 23). Being present also allows the researcher to recruit non-traditional participants. Hanging out, however, also creates demands—dealing with a posture of openness, transparency and reciprocity means the obligation to go to meetings or events that one is invited to. In terms of the research data that is achieved, hanging out provides better access to dissenting voices. For example, we observed dissenting voices from the civil society participants in the European Parliament's events on anti-racism (Kantola et al., 2023).

Finally, it is worth noting that other practices, such as interviews, can be conducted during the ethnographic fieldwork in high intensity. Nadia Brown, for example, conducted 49 interviews over a period of just nine days (2014: 185). This may reflect on the type of fieldwork relationship, trust and time with interlocutors that has been established. Similarly, Emma Crewe compared UK MPs to UK Lords. She suggested ‘peers are perfect informants: leisurely, candid and reflective. MPs are the opposite in every possible way’ (Crewe, 2015: 5).

Conclusion

This chapter has discussed different aspects of gathering data through research interviews and parliamentary ethnography. Both the amount of preparatory work and the need to be creative, persistent and flexible were evident in our deliberations. The preparatory work included passing an ethical review, preparing forms, making decisions about whom to interview, considering different strategies to approach the interviewees, and doing background research about their political careers and work interests, all of which are undertaken prior to the interview or the ethnographic practice of shadowing. This required a constant willingness to evaluate the chosen strategies, interview questions and the ability to be reflexive when some choices were not bringing the desired outcomes. This chapter has strongly demonstrated the time-consuming and labour-intensive character of data gathering. Data analysis, which we turn in the next chapter, is no different in that sense.