Research in a Closed Political Context,COVID,and Across Languages: Methodological Lessons,Messages,and Ideas

Abstract

This study offers a number of contemporary qualitative and mixed methods research lessons regarding cross-language research and research in closed social and political contexts such as post-Soviet countries, especially how such research may be adapted to disruptions like those posed by the COVID pandemic or other disasters. The lessons draw from a study of policy learning among government officials in Kazakhstan which illustrated several methodological strengths and weaknesses and generated a number of methodological recommendations. In particular, while the literature on best practices for research in closed contexts emphasizes deep interactions to develop trust, the COVID pandemic and other disruptions may force researchers to transition to online modalities and constrain options. We suggest strategies to overcome these limitations, including supplementing findings from interviews with content analysis or other multimethod approaches. Additionally, we argue that researchers in post-Soviet and post-colonial contexts must be particularly attuned to the challenges of cross-language research and the combination of local languages with the language of the colonial power. These lessons hold relevance for researchers working in a variety of contexts as they conduct research during times of disasters and geopolitical instability.

Introduction

This paper explores “lessons learned” and methodological insights gained from dissertation research conducted on policy development in Kazakhstan during 2020 and 2021. These “lessons learned” are based on unprecedented circumstances when the COVID pandemic disrupted empirical studies as well as conducting cross-language research in the specific context of a post-Soviet new democracy. Specifically, we highlight four key methodological insights: (1) recommendations for scholars conducting interviews with government officials to consider specific ethical norms of civil servants as they “study up” with this population, (2) complexities of cross-language research and multi-language research in post-Soviet and/or post-colonial contexts that include different languages for personal communication and administrative communication, (3) the constraints that exist conducting research in a relatively closed political and social context, ¹ particularly during disruptions such as those caused by the COVID pandemic that require online modalities, and (4) the potential for content analysis or other multimethod approaches to overcome these challenges. While the specific dissertation study that inspired these lessons explored policy learning among Kazakh government officials, the lessons provide methodological recommendations to scholars from a wide range of social science disciplines and topics. Many scholars conduct research with government bureaucrats, cross-language research, and research in closed political and social contexts and the lessons learned from this dissertation research can support empirical research that is valid and culturally appropriate.

Interviewing government officials can provide data for a wide range of social, economic, and political questions that are fundamental to social science research. While government officials offer a unique and valuable perspective to interviewers, they also exist within the structures and constraints of bureaucracy and bureaucratic norms. Interviewing bureaucrats and government officials is a unique form of “studying up” in which researchers study elites (Aguiar & Schneider, 2016; Nader, 1972) and bureaucrats have professional norms and standards that require confidentiality and adherence to the “party line” (Barnett, 1997; Gusterson, 1997; Nair, 2021; Neumann, 2012). Interviewing government officials is constrained by these ethical norms and these limitations may be exacerbated by disruptions that necessitate the use of online modalities, therefore interviews with government officials must be understood within this context.

Moreover, cross-language research presents a number of challenges for researchers including access and validity concerns. Those challenges may be amplified in multilingual research contexts such as post-Soviet and/or post-colonial nations that utilize an administrative language that may be different from the language used for personal conversations. Our lessons highlight the importance of state and administrative languages, since the use of multiple languages by government officials during an interview is common and carries deeper meanings about the post-Soviet context and public-private relations. These issues may be encountered in other post-Soviet contexts and/or post-colonial contexts.

Research in relatively closed political and social contexts and/or new democracies may also face substantial difficulties related to a lack of institutional trust. In this case, while Kazakhstan is on a pathway to creating democratic institutions and social-economic stability, the closed political and social context constrains the information obtained from interviews and other data sources. Many “best practices” for study in closed contexts recommend in-person research methods and casual, frequent interactions between researchers and participants to build trust and rapport (Ahram & Goode, 2016; Belcher & Martin, 2013; Koch, 2013a; Reny, 2016), but disruptions such as the COVID pandemic or other disasters may make such methods impossible and necessitate online interviewing. The deep contextual knowledge that comes from partnering with local scholars and associates (Caron, Ulrich-Schad, and Lafferty 2019; Schewe et al., 2022; Smith, 2001) becomes even more essential in these situations and supplementing online interviews with content analysis or other multimethod techniques can overcome some of the limitations.

These methodological insights derive from dissertation research beginning in early 2020. At this time, the first author began a research project to understand the policy formation process in Kazakhstan, focusing specifically on development and implementation of environmental policy. Since the research focused heavily on state-civil society relationships and intra-state coalitions, the study was designed to rely heavily on in-person interviews and observation with government officials, bureaucrats, and civil society leaders. The study was designed to implement several of the “best practices” for research in a closed context: interviewing in neutral settings, interviewing in native languages, and building trusting relationships with participants. COVID disrupted all of these plans. Due to the new reality of COVID, face-to-face interviews converted into an online platform and trusted relationships were difficult to build. Suddenly the first author found herself interviewing bureaucrats online while they used their government-owned computers in their government offices. The context was completely different from the environment and research relationships she had planned to create. These new restrictions, however, led to innovation and new insights. In this paper we share some of these insights that may be valuable to other scholars and researchers.

Literature Review

Our methodological insights build upon scholarship on conducting research in closed or politically restrictive contexts, cross-language research, the role of culture brokers and local partners, and online research methods and mixed methodologies. We briefly review the scholarship in each of these areas to highlight the extensions of our findings.

Research in Closed Contexts

Research in “closed” or politically and/or socially restrictive contexts presents a variety of challenges. Scholars from political science, public administration, and other fields frequently conduct research within closed contexts and have developed a number of recommendations for such research. Our methodological insights build upon these suggestions and extend them for research that also takes place during the ongoing COVID-19 pandemic and political unrest.

An increased level of reflexivity, or examination of how one’s position as a researcher and the research context influences research outcomes, is often espoused as essential for research in closed contexts (Ahram & Goode, 2016; Menga, 2020; Sultana, 2007). In an introduction to a special issue on research in “closed contexts,” Koch (2013a) argues that research in closed environments calls for an even higher level of reflexivity and thoughtfulness than other research. Gerharz (2017) echoes this call for increased reflexivity, saying that research in closed or authoritarian environments “necessitates careful navigation of the self and requires thorough reflection on the consequences of particular moves at the point of decision making” (Gerharz, 2017:1).

The importance of relationship building and the significance of ordinary interactions to build trust with research participants in closed environments is often emphasized. Discussing research in closed contexts, Turner (2013) advocates for the importance of time and patience, emphasizing repeated visits and ordinary interactions with gatekeepers and research participants to establish trust and openness with research participants. Koch (2013b) highlights how in closed contexts, routine interactions with individuals such as taxi drivers can reveal political opinions that otherwise people keep private. Nelson (2013) echoes this emphasis on the importance of ordinary interactions, sharing a story of how a routine interaction with a commercial vehicle driver revealed deep tensions around land and resources in her community of study that she otherwise would have missed. Malekzadeh (2016) advises researchers in closed contexts to “make it boring” and embrace the importance of ordinary, everyday interactions and relationships.

Researching in closed contexts often requires multiple data sources and mixed methods to address limitations in access to information. Many scholars advocate for using documentary analysis, including analyzing materials such as government documents, newspapers, etc., to overcome some of the difficulties of access, safety, and data validity that may exist in closed contexts (Ahram, 2016; Belcher & Martin, 2013; Goode, 2016; Malekzadeh, 2016; Markowitz, 2016). Belcher and Martin (2013) advocate analysis of documents and the “public record” as important data sources in these closed contexts, but caution researchers that even these documents must be considered in context and that the public record and/or press are not neutral. While Goode (2016) recognizes the value that documentary analysis can add to research in closed contexts, he does find that research that actually takes place in the field (ethnography, interviews, etc.) is more likely to show the extent of authoritarianism than documentary analysis. Research in closed contexts may require proxies to study sensitive topics when researchers cannot safely or reliably gain access to data directly addressing their questions (Markowitz, 2016). Reny (2016) also finds that many researchers studying politics in China have utilized qualitative methods to overcome limitations with state sanctioned quantitative data.

“Studying up”, or studying elites and those in power, always presents a number of challenges (Aguiar & Schneider, 2016; Gusterson, 1997; Harvey, 2010; Hughes & Cormode, 1998; Nader, 1972), but those issues may be compounded when the research is conducted in a closed environment. Gentile (2013) reminds researchers that they may face real dangers when studying elites and the government in Post-Soviet nations: “research tends to attract the surveillance of repressive authorities and, more specifically, of the secret services” (Gentile, 2013:427). Gentile argues that researchers should consider the real threat posed by secret services in Post-Soviet contexts, which includes the very real possibility of electronic surveillance and bugs, people being assigned to shadow and/or report on researchers’ actions, and even detention and interrogation. Malekzadeh (2016) encountered a major political and social uprising during his research in Iran and found that he, likewise, faced real threats due to his study of the government. Even when researchers may not face direct threats, studying elites and government officials faces challenges as they limit discussion of “sensitive” topics and/or adhere closely to a predefined government narrative. Discussing strategies to conduct interviews with government officials in closed contexts, Markowitz (2016) says that: “first, the interviewer must put him/herself in the informant’s shoes and conduct the interview accordingly” and: “the interviewer must be tactful yet assertive, and must possess in-depth knowledge of the locality and the subject under discussion” (Markowitz, 2016:904). Markowitz acknowledges that government officials often limit the information that they provide and that they should recognize that each interview may only give a fraction of information. Nair (2021) advocates for a “deep hanging out” when researching government bureaucrats, a method that lies “in the interstices between participant observation and in-depth interviews” (2021:1307).

The majority of scholarship making recommendations for research in closed contexts assumes that the researcher is an “outsider” to the context (Menga, 2020). What research in authoritarian or closed contexts might look like for those who are “insiders” to the nation-state or community is largely unexplored; we address this gap with this paper examining how insider status affects openness for researchers. While many scholars have examined best practices for research in closed contexts, the realities of this research during the ongoing COVID-19 pandemic and current political and social unrest raises unique challenges that we explore in this paper. In particular, the kind of ordinary interactions and relationship building often advocated for researchers in closed contexts is extremely restricted for research within the COVID context. Additionally, while interviews face a number of challenges in authoritarian settings, when COVID forces those interviews to be conducted online or over the phone they may face additional obstacles. We explore these issues to extend recommendations for contemporary research in closed contexts.

Cross-Language Research

Cross-language research occurs when the researcher(s) and participants speak different languages and/or when the research is conducted in one language(s) and then translated into another (Squires, 2009). Cross-language research raises a number of challenges, including issues of validity and reliability as research is translated and interpreted, which involves both literal translation and contextual interpretation (Temple, 2006). “In fact the process of translating the written or spoken word from one language to another is one of reconstructing meaning rather than uncovering it” (Temple & Young, 2004:161). Regarding validity, cross-language researchers must address challenges such as when languages do not have direct equivalencies for words or concepts or when direct translations have different contextual meanings in the two languages (Pan et al., 2019).

Translators “are active producers of research”, not “neutral transmitters of messages” (Temple, 2002:845–46). Because “there is no one correct way of translating” (Temple, 2002:846), translators apply their own logic, understanding, and experience to the translation process to choose the specific words and rhetoric used in the translation. Squires (2009) identifies 14 specific recommendations for cross language research that are broadly recommended throughout the cross-language methods literature, grouping them into four categories: conceptual equivalence (criteria 1–3), translator credentials (criteria 4–6), translator role (criteria 7–10), and methods (criteria 11–14). Based on her study, translators’ credentials such as a level of language competence of the researchers is one of the often neglected recommendations (Squires, 2009:283). Squires (2009) finds that very few, only six of the 40 empirical studies examined, fulfill all of these criteria. Instead, most studies exhibited some, but not all, of the best practices, demonstrating the need for improved practices in cross-language research.

When a researcher themselves is multi/bilingual, cross-language research still presents challenges. “Bilingual researchers are seen as unproblematically acting as bridges between communities just because they are bilingual” (Temple, 2006:1), however, bilingual researchers still bring their perspectives and influences to the research. Indeed, “bilingual researchers come to the research carrying their own perspectives and, possibly, shared histories with the communities they are working with” (Temple, 2006:2). Bilingual researchers play a “double role”, acting as both researcher and translator in their own research (Shklarov, 2007). These complexities of bilingual and cross-language research must be considered as an extension of reflexivity and positionality. We extend this literature by examining a case of multilingual research and the context of a multilingual case in which the language used in government administration – Russian – may be different from the language used in personal communication – Kazakh.

Culture Brokers and Community Partners

Research that is cross-language is often also cross-cultural, involving not only linguistic translation but bridging across distinct cultural contexts. Cross-cultural research often relies on culture brokers, or cultural “insiders” to facilitate relationships between researchers and participants (Smith, 2001). Culture brokers may play the role of “cultural mediation” (Miklavcic & LeBlanc, 2014) between researchers and participants from distinct cultural groups, even if researchers and participants share a language. Culture brokers help provide access and recruitment of participants (Eide & Allen, 2005), help researchers establish cultural competence (Jezewski 2016), and generally help mediate between culturally distinct researchers and participants and reduce conflict and misunderstanding (Geertz, 1960; Kaufert & Koolage, 1984; Press, 1969; Wolf, 1956). Culture brokers are often also translators or interpreters, but researchers may rely on culture brokers even when they are bilingual (Eide & Allen, 2005).

Beyond the role of culture brokers, scholars increasingly recognize the value of community partners for research. Attempting to overcome concerns around access, validity, and exploitation, scholars have explored a variety of approaches to actively engage community partners in research (Chambers, 1994; Halseth et al., 2016; Jay, 2010; Schewe et al., 2022; Smith, 2012). Fully participatory research methods engage community partners as active research partners at all stages of research (Vaughn & Jacquez., 2020). These participatory methods are focused on empowering community members and producing research that is relevant to the needs and desires of participants rather than solely scholarly production (Vaughn & Jacquez., 2020). Community-engaged research may not involve the full spectrum of participatory methods (Halseth et al., 2016), but builds upon this desire to conduct research that is relevant to community members and avoid the pitfalls of “parasitic” or exploitative research (Cable, 2012). Additionally, community partners can increase the validity of research, helping to ensure that researchers accurately capture the concepts at hand (Caron et al., 2019; Schewe et al., 2022). We extend this literature by examining the potential of an “insider” researcher to overcome some of the challenges of cross-cultural research, particularly in a closed political context, as well as some of the difficulties of doing relevant and non-parasitic research within the constraints of the COVID pandemic.

Online Interviews

The use of online interviewing methods has been widely explored in qualitative and mixed methods literature. Overall, comparisons between face-to-face interviews and online video interviews have found many points of equivalence and similarity between the methods (Gruber et al., 2008; Lo Iacono et al., 2016; Miller & Dickson, 2001; Nehls et al., 2015; Shapka et al., 2016). Online interviews can provide access to a wider range of participants than may be accessible in face-to-face methods and can help eliminate or reduce geographic restrictions on access (Lo Iacono et al., 2016; Miller & Dickson, 2001; Nehls et al., 2015). Online interviews may be particularly well-suited to access participants who are busy (O’Connor & Madge, 2003). Conclusions on the depth of data produced by online interviews and potential for bias as compared to face-to-face methods are more mixed. Joinson (2001) and Hanna et al. (2005) found that online interviews elicited greater self-disclosure than other methods, while others have found that participants may be more guarded in online interviews (Shapka et al., 2016). Generally, however, scholars have found that online interviewing methods can provide valid and in-depth responses from participants (Miller & Dickson, 2001; Nehls et al., 2015).

Challenges of online interviewing include ensuring adequate technology, which may make computer-mediated methods unsuitable for some participant populations such as the elderly or others with limited technological access (Nehls et al., 2015; O’Connor & Madge, 2003). Importantly, establishing rapport with participants in online interviews can be challenging. Scholars emphasize the importance of open communication in advance of the interviews, self-disclosure by the researcher, and active listening to establish and maintain rapport (Janghorban et al., 2014; Lo Iacono et al., 2016; Nehls et al., 2015; O’Connor & Madge, 2003). While online interviewing can raise privacy concerns for some participants (Nehls et al., 2015), participants may actually feel more comfortable given that they can control the setting of the interview directly (Gruber et al., 2008; Janghorban et al., 2014; Nehls et al., 2015). In this paper we examine the impacts of a rapid “pivot” to online interviewing required by the COVID pandemic, the specific constraints of online interviewing in a closed political context, and strategies to overcome these challenges.

Mixed Methods to Overcome Limitations

Mixed methods research (MMR) is research that involves data collection and analysis from more than one research method simultaneously or in sequence (Creswell & Clark, 2010). Often, MMR is used to describe research that combines qualitative methods of data collection with quantitative methods of data collection. However, scholars increasingly recognize the value of multiple and mixed methods, even if those methods may be multiple qualitative and/or quantitative methods (Leech & Onwuegbuzie, 2007). For instance, as reviewed above, several scholars advocate for the use of secondary data and/or content analysis to overcome limitations of research within closed contexts (Belcher & Martin, 2013; Goode, 2016; Markowitz, 2016; Reny, 2016).

We extend this literature by examining how use of multiple methods may help researchers overcome the constraints of research during COVID or other disruptions, an approach that was explored by several scholars during the pandemic (Meixner & Spitzner, 2022). In particular, this research faced challenges interviewing bureaucrats who may be reticent to share, further constraints because of the closed political climate, and then further yet challenges posed by conducting online interviews because of COVID. Our lessons highlight the advantages and disadvantages of online interviews and highlight the potential of mixed methods approaches to deal with the limitations imposed by the pandemic or other disruptions.

Case Background

Kazakhstan is a country in Central Asia that gained independence after the collapse of the Soviet Union. Bordering Russia in the North, China in the Southeast, and with Central Asian countries in the South, including Uzbekistan, Kyrgyzstan, and Turkmenistan, the geographical area of Kazakhstan is 2.724 million square kilometers, which is the largest in Central Asia and the ninth-largest territory in the world (“The Republic of Kazakhstan”, 2022). According to the Constitution, the Republic of Kazakhstan is a unitary state with a presidential form of government (“Constitution of the Republic of Kazakhstan”, 2022). However, despite Kazakhstan’s peaceful democratic transition when compared to struggles such as the civil war in Tajikistan from 1991 to 1997 or religious violence in Uzbekistan, more recent unrest in Kazakhstan in January 2022 reveals hidden complexities of Kazakh civil society.

Kazakhstan, like most post-Soviet nations, is dominated by both the Russian language and the local language of Kazakh. Although the current number of ethnic Kazakhs is relatively high compared to numbers before the country’s independence in 1991, the dominance of the Russian language in everyday life is obvious and persistent. According to the Constitution of Kazakhstan, the Russian language can be officially used in governmental institutions along with the Kazakh language as a state language (“Constitution of the Republic of Kazakhstan”, 2022, Article 7). In most cases, the Russian language is characterized as the dominant language, particularly for civil servants and government communication.

Methods

This article draws from a study conducted by the first author using online interviews with governmental administrators, independent experts, scholars, representatives from NGOs, and activists located in Nur-Sultan, the Akmola region, Almaty, Kokshetau, Shchuchinsk, and Ust-Kamenogorsk, Kazakhstan. While the research was originally designed around face-to-face interviews and observation to begin in May 2020, COVID forced a transition to online methods.

The data was collected as a part of dissertation research focused on environmental policy development in Kazakhstan (Kazbekova, 2022). Between May 2020 and September 2021, the first author conducted in-depth, semi-structured online interviews with 43 participants: 20 governmental administrators, six independent experts, four scholars, and 13 representatives of NGOs, and activists. Drawing upon her previous network, including initial contacts, she used snowball sampling for these interviews. The first author also has extensive prior experience conducting semi-structured interviews with German embassies’ representatives in Vienna, Astana, and the Goethe Institute in Amsterdam, as a part of her research for a previous PhD in International Relations.

Interview questions focused on environmental policy development, specific projects including ‘green economy’ transition segments, current goals and beliefs, and the relationship between stakeholders and policy change. She transcribed and deductively analyzed transcripts and notes using NVivo software (“QSR International” 2021), using the Advocacy Coalition Framework (ACF) (Weible & Sabatier, 2017, p.151). Online interviews were conducted through various online platforms such as Zoom, WhatsApp, Skype video, and audio calls based on participants’ preferences. Interviews were conducted in Russian but often transitioned from Russian to Kazakh during a conversation. Interviews were transcribed in the original language of the interview and then translated to English following full transcription. ² The duration of the interview lasted between 15 minutes to 3 hours, depending on the availability of each participant and readiness to share their thoughts. Interviews were coded by the first author using an iterative coding approach in which the first stage of coding was deductive utilizing the ACF concepts (Kazbekova, 2022, p.85; McKibben et al., 2020) and second round coding using open-ended discourse analysis to allow themes to emerge in vivo (Charmaz, 2006; Kazbekova, 2022, p.86). All research received human subjects IRB approval from xxxx University (study number 20–035).

Additionally, research included content analysis of the daily newspaper Kazakhstanskaya Pravda. Kazakhstanskaya Pravda was chosen for being the most popular and widely read newspaper since 1920. Kazakhstanskaya Pravda is the national, state-sponsored newspaper focused on official news related to all spheres of social, economic, and political life. The first author searched the online archives of Kazakhstanskaya Pravda from 1991–1999 for any mention of environmental policy. This resulted in a collection of 13 articles explicitly focused on environmental policy initiatives. Articles were coded in NVivo using the same codebook used to analyze online interviews. Interviews, coding, and content analysis were conducted solely by the first author.

The methodological “lessons learned” were derived from analysis of the first author’s personal experiences and research memos. Following each interview, she wrote a memo describing the context of the interview and experience, including such details as whether the camera was on and the perceived rapport and comfort with the participant. Follow-up analysis and discussion between authors led to the development of key “lessons learned.”

It is essential to recognize the researcher’s positionality and her background as a civil servant in Kazakhstan, working in the state court and central government as well as graduating from the Academy of Public Administration under the President of the Republic of Kazakhstan. This personal background helped to provide access to participants and also having insights and understanding of the ethics and norms of civil servants allowed her to create a friendly environment to address questions and get answers. However, being a scholar at an American university influences her positionality and the perception of her among participants as not fully an “insider”. Thus, it provided a deeper understanding of the limitations for outsider scholars conducting research in countries such as Kazakhstan.

Findings

Based on experiences conducting dissertation research in Kazakhstan, we identify several practical recommendations for scholars to conduct research with government officials, in closed political contexts, cross-language research, and/or during disruptive events such as COVID.

Consideration of Specific Ethical Norms of Government Administrators

Government officials are representatives of the current government who advocate, support, and follow the country’s political course. While bureaucrats might share their personal views, opinions, and expertise during an interview, they generally do not undermine specific issues related to a ministry or a department by discussing implementation failures or shortcomings. Researchers should not expect from these representatives strong criticism or disputes and must recognize that the absence of such critique is not equivalent to a positive assessment. Bureaucrats often begin their responses with support of current strategies or with general information highlighting advantages of the present political course, for example (R ³ 2):

The Ministry of Ecology, Geology, and Natural Resources is a newly established ministry with its strategic goals and tasks. We often face challenges in our daily routine by realizing programs, ³ strategies, and plans. Still, it is a common practice, and we understand that we, as civil servants, need to find the right solution to solve issues and provide a good quality of our work. While the implementation of strategic plans of the Ministry is our daily responsibility, the performance of the country’s strategic path and political course, announced by the Administration of the President, is one of our high priorities as well. Of course, we fully support new political policies, reforms, and all government initiatives and concepts announced by the President.

Another bureaucrat from a local department also mentions that “in the civil service, you have to expect a huge number of tasks and requirements but not having enough time and resources to address and complete them. However, it doesn’t mean the political course is wrong and ineffective” (R6). This quote demonstrates one common strategy during interviews which is that a critique, in this case that administrators face a large volume of work in civil service, is partnered with an affirmative and positive statement.

Experienced bureaucrats divide their answers between their personal views and views as a representative of the ministry. For instance, one interviewee said (R19):

I think the transition to a green economy is critical because our environment is getting worse, the growing issues among them are air pollution, lack of access to clean water in regions and villages, and an increase in common allergic diseases in our country. However, I personally believe that the transition to a green economy will take a long time.

Researchers interviewing bureaucrats must recognize these norms and understand their interviews in context or risk misinterpreting interview data. The absence of critique must be understood as not equivalent to support.

Acknowledgment of the Complexities of Multi-Language Research

Although Russian was the primary language of communication between interviewees and the scholar, there were cases when interviewees switched to Kazakh during the conversation. This transition to Kazakh occurred when participants wanted to differentiate their personal views, concerns, disagreement, or critique from their professional governmental role. For example, a participant from a regional department expressed his arguments in Kazakh. The participant turned off the camera, and I could not see the respondent’s face. However, I ⁴ heard from the respondent’s voice and his emotional speech in Kazakh the graveness of his concern about corruption and poverty (R20):

[translated from the Kazakh] The Green economy transition, the EXPO-2017 exhibition in Astana, and other international events are nice actions that help us, as a country, be recognized internationally. However, all the money and budget spent on these events could be spent on opening new kindergartens and constructing a better logistic infrastructure, particularly in villages and other regions of Kazakhstan, by providing access to clean drinking water that some places still do not have.

Several key changes alerted me to the significance of the statement: first, the respondent turned off his camera, second, he switched to Kazakh after conducting the rest of the interview in Russian, and, third, he dropped the volume of his voice to a low whisper with an emotional tenor. Another participant from the same regional department underlined their disappointments of the budget distribution by switching to Kazakh in mentioning that in order to organize the EXPO-2017 event in Astana, the government spent a lot of money for a 3-month event instead of using this money for opening social facilities or rebuilding roads in regions.

In another example of the language-switching, a local representative switched to Kazakh when discussing corruption at different levels in the civil service. (R21): “[translated from the Kazakh] Corruption in the system and poverty among Kazakh people in the regions, who do not have enough enthusiasm to care about the environment, are based on uncertainty about their future economic and social survival.” The participant articulated his frustration in Kazakh about the current environmental policy implemented by the government and citizens’ neglect and irresponsibility regarding a cavalier attitude to the environment.

While Western scholars might have good enough language skills in Russian and Kazakh, this case demonstrates the complexity of multi-language research and deep context that the language provided. This depth of analysis was possible because of a deep level of knowledge by a local scholar. Although non-local scholars may have strong language skills, the reason a participant switched to their native language may only be recognizable by native speakers with deep contextual knowledge.

Potential of Local Collaborators to Overcome Limitations Researching in Closed Political Contexts

This research also revealed many challenges of research in a closed political and social environment. Participants may have fears of being pressured by administration, fired, or, in the worst scenario, administratively or criminally convicted, and researchers must be attuned to these fears. At the beginning of each interview, it was essential for me to ensure a participant’s confidentiality and safety, including basic requirements addressed in the IRB (Review of International Research) processes. Although the consent forms’ aim is to provide confidentiality for a participant and a researcher, it does not offer a friendly environment at the beginning of the interview. Moreover, oral consent forms are more acceptable than written forms among government representatives because they keep their confidentiality secure, ensuring they cannot be punished or convicted later. This practice has roots in the historical patterns under the Soviet regime and different forms of prosecution that lead participants to prefer staying anonymous and safe.

Based on this research experience, essential elements that provide the foundation of trust between a participant and a scholar are face-to-face eye contact, an emotionally friendly atmosphere, and scholars’ appreciation of participants’ time and effort. I created an emotionally friendly atmosphere with actions such as providing flexibility to a participant to use or not to use a camera, choosing his/her language preferences, stating appreciation for the time and effort of participants, and, most importantly, emphasizing shared background and experiences between participants and myself to build comfort and fruitful conversation.

A coherent description of a study’s objectives, significance, and future contributions alongside personal experience and communication competencies, are necessary to build confidence between a participant and a scholar. During the interview, I acknowledged that a brief description of the study is not enough to build confidence between a scholar and a participant. However, sharing my experience as a civil servant in the past helped me create space and trust for an interview to occur. For instance, one participant wanted to be critical of hiring and promotion practices but paused midway through his answer. He referred to my experience as a former civil servant by mentioning cases of corruption that are still taking place in society (R9):

Unfortunately, the government system does not accept intelligent people with new ideas and approaches. Only reliable people are in need, supporting the old system of governance mainly based on nepotism, family relations, and being in favor. Today’s servants who match these requirements can be ministers tomorrow. [Long pause] This is one of the issues why people quit civil service. You are aware of these cases, I am sure.

My deep local and professional knowledge supported this kind of disclosure, which likely would not have been available to a foreign scholar. For foreign scholars, collaboration with local partners might address these challenges as local partners are more familiar with the social and political context. For example, this interviewee again referenced that my insider status made them willing to share and discuss controversial issues (R8):

You know that people in villages, located in regions far from big cities, mostly have concerns about basic needs such as food and a paycheck for the next month. (Emphasis added)

Local scholars and community partners can help provide enough information for foreign academics to understand the accurate picture of the country’s situation and to place results in context, particularly in a closed political setting, and may also make participants more comfortable. Although knowing the native language of a country to conduct research is a requirement that scholars must always be mindful of, language proficiency is not enough to overcome cultural differences, particularly in closed political contexts.

Using Content Analysis to Overcome Limitations

This study demonstrates mostly beneficial outcomes from conducting online interviews during the COVID pandemic. Although at the beginning of the study the research was focused on face-to-face interaction with government administrators, the online format was largely successful and provided some unique opportunities. Firstly, the online interviews broadened access to more participants and still allowed me to utilize my personal and professional network to solicit participants. Although there were people who refused to participate in online interviews that used video calls, WhatsApp and Skype audio calls offered an alternative to those participants and was a convenient mode of communication widely supported by participants. The uniqueness of the COVID pandemic and worldwide isolation provided options for conducting interviews for participants and scholars. The flexible availability of participants due to working from home, particularly at the beginning of lockdown, or opportunities to use a comfortable space to be online with the camera off or on, and time preferences were advantages for both participants and scholars. Surprisingly, arranging an online interview was not problematic and was convenient for most participants, even considering the time zone difference. A key convenience was the ability to choose an appropriate place for a participant, whether at home, during break time, in the office, or sitting in the car and using their cell phones.

Regardless of format, it is always a challenge for a researcher to build sufficient trust or a comfortable environment for a participant during both in-person or online interview processes. While online interviews allowed me to continue my research, issues such as a poor internet connection presented minor technical problems. There were cases when a participant turned the camera off during an interview due to a poor internet connection. There were also some cases during interviews where participants had time limits to talk or less confidentiality in a crowded space during their lunch breaks or in their office. As a result, some interviews were postponed or rescheduled several times.

Content analysis as a second research method helped to overcome the limitations in building rapport with participants and getting access to participants in online interviews. Mainly, the analysis of publications provided an essential historical background of critical events that was essential to understand and interpret interviews. It was not feasible for interview questions to cover the entire time frame that was relevant to the study nor for interview participants to provide information on the full context of their discussion, particularly given the constraints of online interviewing that limited the length of interviews and ability to have side conversations and observations.

For example, content analysis revealed three key themes that were not provided by interview data: (1) the existence and perspective of multiple stakeholders beyond government agents, (2) historical context describing an early struggle amongst stakeholders to create a shared vision for environmental policy, and (3) context regarding attempts to engage the international community on environmental policy. Firstly, the analysis of early publications showed the perspectives on environmental issues from a wide range of various stakeholders, including international organizations, scholars, and activists, regarding agenda-setting and decision-making. The perspectives of these stakeholders were not easily available via interviews and would otherwise have been absent from the analysis. Secondly, the newspaper content revealed that multiple stakeholders criticized the government and underlined the governmental ignorance regarding ecological tragedies happening in the past and actively neglected in the present. Given the extreme reluctance of bureaucrats to offer critical perspectives and the constraints of interviews in a closed political context, this data regarding disagreement and dissension would not otherwise have been captured. Finally, the newspaper content described how Kazakh policy did and did not align with international priorities, a perspective that again was not present in interviews given the internal focus of bureaucrats and the reluctance to discuss international politics because of fears from being in a closed political context.

Discussion and Conclusion

Our findings highlight a number of methodological challenges and strategies to overcome those challenges in doing research in closed political contexts, multilingual post-colonial settings, and when faced with disruptions such as COVID or other disasters. Firstly, we found that government bureaucrats have ethical norms that require formal and affirmative language to be used and an unwillingness to discuss critiques; researchers must understand this context to interpret results. Additionally, we find that in the unique post-Soviet context individuals often move between communication in Russian and Kazakh, and that this language-shifting reflects a boundary between professional and personal statements. We also find that the closed political context may make individuals wary to express their views and that a local partner familiar with the context is essential to fully understand this context. Also, the COVID pandemic limited the ability to conduct face-to-face interviews and required transferring them online, adding some limitations of online modalities particularly regarding building trust and rapport with participants. Content analysis was successful in supplementing the interviews to overcome limitations and help understand broader historical and political context that was difficult to obtain during online interviews.

Our findings extend several existing methodological literature regarding research with government officials, cross-language research, and research in closed political contexts and underscore the recommendation to collaborate with local partners who hold deep contextual knowledge. Studying “up” presents a number of challenges (Aguiar & Schneider, 2016) and studying bureaucrats, in particular, requires attenuation to specific norms to understand their statements (Nair, 2021). We confirm Markowitz’s (2016:904) guidance that tactful discussions based on knowledge of a subject during interviews were essential to maximize the limited information provided by officials. We extend this by identifying common strategies in which mild critique from bureaucrats was sandwiched with official position statements and by identifying the language-switching that was used in this post-Soviet context. Additionally, we find that the professional history of the lead author as a government bureaucrat herself and her positionality as a local scholar and native speaker were essential to correctly interpret participants’ interviews. Together, our lessons learned recommend that scholars be attentive to the professional norms of bureaucrats and collaborate with knowledgeable local partners.

While this study is an example of cross-language research, our findings also speak to the complexity of multilingual and post-colonial settings such as Kazakhstan in which professional and personal languages may differ. As Temple mentions (2002), during translation it was essential to establish a contextual interpretation of conversations with participants by translating each transcript from one language and converting this meaning into another. We extend this by identifying that participants used language-switching between Russian and Kazakh to differentiate between their official professional statements and personal opinions; this deeper context would likely have been missed by a researcher who was not a native Kazakh speaker and deeply familiar with the specific context. Researchers in post-Soviet and other post-colonial contexts should be attentive to language-switching such as this and its meaning and recognize the potential value of a local partner to help identify these contextual meanings.

There are many challenges in conducting research in closed contexts, particularly during the pandemic. We agree with Turner (2013), who stresses the importance of time and patience for scholars conducting research in closed contexts, emphasizing the effectiveness of repeated visits and ordinary interactions in building trust and openness with participants (Turner, 2013). We also confirm advice from Koch (2013b), Nelson (2013), and Malekzadeh (2016) that small talk and ordinary conversations with various representatives from a community could provide additional data and essential findings that offer stories and political opinions as a big picture in closed context (Koch, 2013b; Nelson, 2013; Malekzadeh, 2016), options that were not possible due to COVID. Since this study was conducted online, we confirm that a lack of in-person interaction with participants and the possibility to return to places and people was one of the most critical limitations. Given the constraints of online research, our experience demonstrates that researchers must work even harder to create a comfortable and secure environment for participants and that this was greatly supported by the researcher’s shared professional background and local knowledge. During disruptions like those caused by COVID or other disasters it is perhaps even more essential to partner with community brokers or other local partners.

Finally, given the limitations posed by COVID and online interviews, we used content analysis to try to provide some of this broader context, as suggested by Goode (2016). The addition of content analysis provided both data and conclusions that were not possible to explore with online interviews alone and helped to address some of the limitations of research with bureaucrats and online interviews in a closed political context. Content analysis provided historical and social context for the information shared in interviews, broadened the perspectives available to the researcher, and presented evidence of disagreement and dissension that was not accessible in interviews. Researchers in closed contexts, working with bureaucrats, and/or conducting online interviews should consider content analysis or other methodologies to broaden data.

While all subjects agreed to participate in online interviews, each interview began with an explanation of the topic of the research, the scholar’s affiliation, aims, and research questions of the study. Thus, based on IRB policy, the explanation of participants’ rights, privacy, and confidentiality was provided and multiple assurances to participants about their anonymity. Importantly, participants demonstrated a higher level of trust based on the first author’s shared professional experience and national identity. However, most participants required multiple oral confirmations that their confidentiality would be protected, their names would not be recorded anywhere, and no official statuses or titles would be shared.

Although COVID pushed academia to mobilize online platforms for data collection, these platforms will prevail and remain in the future as a convenient and safe way to connect with participants. Being flexible and prepared for rapid transition and changes by having additional research methodological strategies such as content analysis is one of this study’s key lessons as is the importance of deep local knowledge and partnership. While often explored separately, the intersection of online methods and closed political contexts has not been widely explored in the social science methods literature. The COVID pandemic forced virtually all research into online modalities and based on the many studies conducted during the pandemic, a comparative study of online research between closed and liberal contexts would be interesting to examine. Additionally, a fruitful avenue for future research would be comparative studies examining the practice and meaning of language-switching in post-Soviet and post-colonial contexts. Doing so would be critical to elaborate on unique aspects of multilingual post-colonial context and their impact on research.

While the insights we offer were derived from specific dissertation research on policy formation in Kazakhstan, they hold relevance for the many researchers who conduct research with government officials, in multilingual post-Soviet and post-colonial contexts, in closed political and social contexts, and whose research may be disrupted by major disasters or challenges.