Beyond Big Brother: How to Study Tech-Driven Authoritarianism With Restricted Access to State Institutions

Abstract

With the tremendous advancements in Internet, big data analytics, and artificial intelligence, the power and potential of digital technologies has a special appeal to political rulers. How can qualitative researchers explore tech-driven authoritarianism when they have limited access to state institutions? This article addresses this question by arguing for a wider and more nuanced understanding of tech-driven authoritarianism as a state-market complex mediating the political application of digital technologies. Based on my own research on China’s Internet surveillance, I find that the engagement of the private sector, especially technology companies, in authoritarian control creates new opportunities for qualitative researchers to study state power in non-state fields. By reflecting on my experience of field-site choice, gaining access, and informant recruitment, I discuss how thorough preparation in both theory and fieldwork approaches help qualitative investigators develop creative ways of collecting information on tech-driven authoritarianism.

Introduction

With the tremendous advancements in Internet, big data analytics, and artificial intelligence, the power and potential of digital technologies has a special appeal to political rulers (Deibert and Rohozinski, 2011; Noesselt, 2014; Sinpeng, 2020). Social scientists have attempted to understand how illiberal regimes survive the impacts of digital technologies. However, the political context of authoritarian regimes sets various constraints on ethnographers and qualitative investigators (Goode, 2016; Massoud, 2016; Reny, 2016; Rivetti, 2017; Roberts, 2013), especially when the focus of their fieldwork is on sensitive topics such as digital surveillance and Internet control. One of the main challenges facing researchers is the limited access to governments and officials (Janenova, 2019; Reny, 2016), both of which are considered the main units of analysis in the study of authoritarian domination. While the development of quantitative research partially addresses this challenge, the nuts and bolts of tech-driven authoritarianism remain underdeveloped due to a lack of data from traditional ethnographic methods such as participatory observation and interviews.

How can qualitative researchers, especially ethnographers, explore tech-driven authoritarianism when they have limited access to state institutions and officials? A regime, nevertheless, remains a broad structural environment in which investigators face more immediate challenges that can either facilitate or hinder their access to data (Reny, 2016). Despite various structural constraints, the authoritarian nature of a regime does not necessarily hinder all spaces for advancement in social science. Currently, the development of technologies is changing the social reality of authoritarian domination, which might bring new opportunities for ethnographic investigation.

Although it is easier to generate new findings in a politically open setting, authoritarian contexts do not necessarily hinder opportunities for advancement in social science (Reny, 2016). In politically closed contexts, adequate preparation in both theory and fieldwork approaches will help investigators capture these new research opportunities brought by the changing social reality. My general suggestion for scholars who are interested in tech-driven authoritarianism is to challenge the state-centered narrative and to include non-state institutions and non-state actors in the research design. Theoretically, my article calls for a wider and more nuanced understanding of tech-driven authoritarianism as a state-market complex mediating the political impact of digital technologies. The entanglement between state, market, and technology is the main feature of contemporary tech-driven authoritarianism, and ethnography researchers should take note of how non-state sites are becoming the field where digitalized authoritarian power operates. Methodologically, by sharing my dissertation project on China’s Internet surveillance, I argue that the choice of fieldwork approach should be determined by both the theoretical assumption and the social reality defining the researched phenomena. My fieldwork research on China’s industry of Internet surveillance shows that private sectors, especially technology companies, are an unexploited research site where investigators can develop new access-gaining channels and informant-recruitment strategies to collect information on state control.

This article contributes to the literature on addressing the methodological challenge of conducting fieldwork in authoritarian contexts on sensitive topics. It also benefits the discussion of critical qualitative research by developing methodological reflection based on non-western experience. In the next section, I introduce the main methodological problems associated with ethnographic research on tech-driven authoritarianism and then analyze how an expanded theoretical understanding of the studied phenomenon can help address them. Following that, I share my fieldwork experience to show how I rely on non-state fields and non-state actors to explore Internet surveillance in contemporary China.

Methodological challenges of performing qualitative research on tech-driven authoritarianism

Collecting data in authoritarian contexts itself is challenging, but doing ethnographic research on politically sensitive topics is even more difficult. One of the primary barriers that authoritarian regimes generate for ethnographers is the problem of access (Glasius et al., 2017; Janenova, 2019; Koch, 2013; Reny, 2016). In authoritarian contexts, subjects often remain inaccessible to researchers due to various structural constraints. For example, authoritarian governments or local authorities may limit access to research subjects by exercising their gatekeeping powers (Ahram and Goode, 2016; Reny, 2016). Evidence has shown that the Chinese government has denied visas to scholars conducting sensitive research. ¹ Even when researchers are permitted to conduct fieldwork in the interested countries, they can also be rejected due to local communities’ low trust in outsiders and the political fear brought about by state scrutiny (Glasius et al., 2017; Koch, 2013; Nelson, 2013).

There is a growing body of literature on the experience of conducting fieldwork in authoritarian contexts (Glasius et al., 2017; Hsiung 2015; Li 2021; Roberts 2013; Shenton and Hayter 2004; Valenzuela-Fuentes 2019), but few studies have specified the challenges scholars meet when studying the use of digital technologies for authoritarian control. The rise of tech-driven authoritarianism, ² the use of digital technologies (e.g. Internet, social media, smartphone, big data, and artificial intelligence) in illiberal regimes for political control, has become a cutting-edge theme in the study of contemporary authoritarian politics. In this field, considering that most popular topics such as Internet control or digital surveillance are politically sensitive, the access issue still bothers most scholars in this field. Furthermore, in a politically sensitive context, it is not easy to access authoritarian rulers and interview/observe them about how they use digital technology to govern people.

Compared with the governor, the access to the governed is relatively easier. An alternative path for ethnographers and qualitative researchers is to study those who are the subject of authoritarian control. Sociologists and political scientists have traditionally studied authoritarian power by investigating how the civil society is shaped or repressed (Fu, 2016; Spires, 2011; Teets, 2014). To learn about the technological engagement of authoritarian domination, scholars often choose citizens or organizations as the unit of analysis to see how their activities are interrupted by tech-driven control practices (Han, 2018; Liu and Graham, 2021; Smeltzer, 2012). Although the impact of technology on civil society is an important topic in the study of tech-driven authoritarianism (Diamond, 2010; Farrell, 2012; Koc-Michalska and Lilleker, 2017; Schlaeger and Jiang, 2014), when data are produced from the side of civil society, scholars are answering questions about how civil society is impacted by digital technologies rather than questions about how the state is using technology to govern the people. What remains unclear is the motives of the concrete actors and the mechanisms underpinning the authoritarian regimes’ use of the internet, big data, social media, AI, and other digital technologies.

Tech-driven authoritarianism as a state-market complex

Does the methodological challenge determine the ethnographic studies’ silence on the topic of tech-driven authoritarianism? The answer is no. The digitalization of authoritarian power is not only arousing new scholarly interest but also producing new research opportunities. To capitalize on these new opportunities, researchers need to be well-prepared in both theory and methods. In this article, I argue for a wider and more nuanced understanding of digital authoritarianism as a state-market complex mediating the political application of technological developments. Although authoritarian states are still the main actor promoting most tech-driven control methods, my article highlights that the role of non-state engagement deserves more methodological attention.

My understanding of tech-driven authoritarianism is inspired by surveillance studies, an emerging field in which scholars investigate the social process of personal information being gathered and influenced (Lyon, Haggerty and Ball, 2014). Although the state is widely considered one of the most established agents of surveillance, a growing body of literature has shown that commercial interests and for-profit organizations are also among the main drivers of the global expansion of digital surveillance (Lyon, 2014; Wood and Ball, 2013). By studying the expanding security industry and surveillance market, scholars have recognized that the state has entered into a partnership with the private sector (commercial suppliers of surveillance technologies) in order to exercise that power (Hayes, 2014).

In surveillance studies, the exploration of digital surveillance is not based on a priori understanding that surveillance is political behavior monopolized by the state. Rather, it suggests that surveillance is a complex social process in which the close connection between the market and the state mediates the technological application in the collection and governance of human information (Bennett, 2015; Fuchs, 2012; Hou 2020. Ethnographic investigators in this field not only explore the institutionalized surveillance systems that emerged in state institutions such as the police system or national security agencies but also focus on those data collection practices utilized in commercial agencies and even citizens’ everyday life (Brayne, 2017; Green and Zurawski, 2015; Lyon, 2007; Mac Giollabhuí et al., 2016; Van Doorn, 2017). All these findings present the truth that the current expansion of digital surveillance is underpinned by multiple agents including governments, corporations, and even ordinary citizens.

Conversely, the popular understanding of tech-driven authoritarianism, including the surveillance practices that exist in authoritarian contexts, is dominated by an Orwellian narrative: tech-driven authoritarianism refers to digitalized repressive moves solely practiced by authoritarian governments (Jiang and Fu, 2018; Zeng, 2016). The political application of digital technologies is narrowly considered an indicator of an antagonistic state-society relation under those regimes (Hou, 2017). Although this narrative represents a critical element of the phenomenon, it leads to the limited theoretical imagination we use to guide empirical studies. For qualitative scholars, if digital authoritarianism only refers to tech-driven state moves, the limited access to governments and officials almost indicates their silence in this field. In addition, the Big Brother narrative will probably be reproduced due to the lack of new qualitative evidence displaying the new actors, mechanisms, and motivations underpinning the encounter between digital innovation and authoritarian domination.

An expanded understanding of tech-driven authoritarianism can inform the empirical research on authoritarian domination in this digital era. Evidence shows that, with the help of the market mechanism and private sectors, technological innovation can be transformed into control devices or repressive tools for state control (Brayne 2017; Dubbeld 2005; Han, 2015; Hou 2020). Taking citizens’ usage of social media as an example, not only the control space, such as social media platforms, is operated by for-profit organizations but also the execution of concrete online control practices, such as Internet surveillance or propaganda, is carried out in the market sphere or by for-profit actors (Deibert and Rohozinski 2011; Han 2018; Vendil Pallin 2017). Although the state-centered explanation remains the main framework to use to understand tech-driven authoritarianism, the high engagement of market actors deserves more methodological attention, as it creates new space for researchers to explore state control in non-state fields.

The understanding of tech-driven authoritarianism as a state-market complex also inspires researchers to gain knowledge from both state agencies and non-state agencies. As mentioned by previous literature, state is not a fixed entity but a set of processes and practices, the transcendence and coherence of authoritarian state power should not be treated as a taken-for- granted fact, but as a precarious achievement (Cho 2011; Ferguson and Gupta, 2002). To unpack the state-market complex, researchers should go beyond a state-centered analysis and focus on how state and non-market forces encounter in the operation of authoritarian domination. On the one hand, researchers should still consider the information gained from the ‘state-side’ of authoritarianism critical to understand tech-driven authoritarianism. By interviewing officials, analyzing official documents, and studying the division of state institutions, researchers can investigate the rationality of government bureaucracies and the motivation of political actors in authoritarian contexts. On the other hand, by extending the field site to non-state institutions, especially the market sphere, researchers achieve the opportunity to further explore how market forces (etc., market competition, entrepreneurial endeavors, profit motivation) drive the political application of digital technologies. To be more specific, through interviewing industry actors, visiting technology companies, and reviewing the documents published by market actors, researchers can not only obtain the chance to cross-check the data gained from governmental sources but also achieve new knowledge about how non-state actors make sense of state control and the process during which market actors actively align their business with the state.

Studying tech-driven authoritarianism with limited access to state agencies

My methodological reflection originates from my own dissertation project, which focuses on Internet control in contemporary China. In the popular narrative, Internet control is widely considered a state action used by the Chinese regime to establish its regulation over cyberspace (Creemers, 2017; Qiang, 2019). According to the current literature, China has established a comprehensive control regime with a full set of organizational, institutional, and administrative tools (Han, 2018). However, a narrow focus on the state sets theoretical limits on both researchers’ research design and fieldwork approaches. Due to the sensitive nature of the topic, it is almost impossible for most scholars to gain access to state agencies to observe how digital control is operated in the field. Even though quantitative researchers contribute to capturing the digital path of the state through social media data and experiment data (Distelhorst and Hou, 2017; King et al., 2013; Meng et al., 2017), their research regards the state as a unified entity and therefore takes the risk of obscuring the dynamics, patterns, and struggles of authoritarian domination. ³

To understand how Internet control is operated in the front line, rather than seeking answers from these inaccessible state agencies, I focus on non-state actors: I explore how private sectors are involved in China’s Internet surveillance. While reading media reports on how selling surveillance services to government clients is becoming a big business in China, ⁴ I realized that this subject might be a new opportunity to conduct a qualitative study on digital control from the market sphere. By visiting companies that provide surveillance services to governments and interviewing non-state actors working in or with this surveillance market, I was able to gather rich information showing how state-market collaboration is becoming the main feature of China’s digital control.

Field-site choice: investigating state surveillance in the state-market complex

Choosing an appropriate study site and gaining access to it is a key part of qualitative studies (Thummapol et al., 2019). If state institutions stay relatively inaccessible to researchers, what are alternative sites for fieldwork? The choice of fields involves a combination of careful planning; academic knowledge; practical research experience; researcher knowledge, skills, and commitment; and luck (Kondowe and Booyens, 2014). Although state-market collaborations create more space for investigating tech-driven authoritarianism in non-state fields, scholars need to deliberately consider how they can transform newly emerged opportunities into workable fieldwork plans.

Like most novice scholars, the first challenge I had to address is the selection of the field site. Site choice is based on both the researcher’s theoretical concern and their pragmatic consideration (Angrosino, 2007). For sociologists, conducting theory-driven qualitative work usually means their research objects are derived from theoretical knowledge and questions. Therefore, the field of sociological research cannot be found somewhere out there but is constructed by the researcher (Nadai and Maeder, 2013). When I was preparing my research proposal, although I had already known the surveillance industry and its role in state surveillance would be my research focus, I did not have a clear idea about where I should go for my fieldwork. This newly emerged industry has not been developed enough so that scholars could easily identify a clearly bounded space or group for information collection. As one of my research targets is to present the ecology of this industry and provide a more trackable path for future research, I need to find appropriate research site(s) that will not only allow me to address my theoretical puzzle but, even more importantly, will also be accessible enough for scholars who cannot visit Internet-control agencies of the state.

To do this, my fieldwork adopts a multi-sited approach to identify concrete sites where researchers can collect information about tech-driven authoritarianism from non-state institutions. Multi-sited fieldwork refers to the study of social phenomena that cannot be accounted for by focusing on a single site, and it requires researchers to track complex social and cultural phenomena by following people, connections, associations, and relationships across space (Belanger-Vincent, 2011; Marcus, 1995). In a theory-driven research project, researchers must rely on theoretical clarification to identify concrete locales where the practices and interactions s/he is interested in can actually be observed (Nadai and Maeder, 2013). As my focus on the surveillance industry is based on my theoretical understanding of digital authoritarianism as a state-market complex, I identify the following three types of sites where I can observe the state-market collaboration that is allowing Internet surveillance to develop into a business.

The first type of site refers to corporate bodies that provide control services to authoritarian regimes. If governments are the main client groups in the surveillance industry, the supplier must be the other critical actor constituting the industrialized state-market complex. In my case, for-profit organizations such as technology companies, media companies, and think tanks are all included under this category. These organizations rely on their technological and intelligence resources to produce user-friendly products such as surveillance software, analysis reports, and Internet censorship platforms. The Chinese government relies on these products to govern citizens’ online expression. Visiting these companies not only helps observe how these Internet-control services operate but also sheds light on the labor division underpinning state surveillance in the frontline. I therefore call for more methodological attention to the private actors (especially technology companies) that are engaged in the operation of tech-driven authoritarianism. Performing ethnographic observation there can help scholars obtain rich information about the production process through which technological development is transformed into the control capacity of an authoritarian government. Even though scholars cannot always achieve access, company websites, product manuals, and other commercial information are usually more accessible compared to government documents.

The second type of site refers to industrial events such as industrial conferences, tech exhibitions, and workshops. Similar to the role of annual academic conferences, these events are important opportunities for companies to promote their businesses, display what they are working on and network with manufacturers, suppliers, clients and customers. For outsiders, such as researchers, attending these events could help get a quick but clear picture of who are the most influential companies in this field and what are their most-updated products. For example, the first time I observed the operation process of a police-oriented surveillance platform was at the 2017 Smart City Expo in Beijing. In the Chinese context, most surveillance-service providers are technology companies that focus on data mining and data processing. Their booths are frequently found in tech expos featuring keywords such as Big Data, Smart Cities, Cloud Computing, Internet Security, or Artificial Intelligence. In addition, the attendants of these events include not only industrial actors but also students who would like to enter the industry, and attending them could provide the opportunity for researchers to build their network for interviews. Considering that these events are accessible for the general public and their information can be easily found online, ⁵ I view them as a valuable information source to study tech-based control in non-state fields.

The third type of site I suggest here is not a pure non-state site. Rather, it is the online platform (ccgp.gov.cn) used by the Chinese state to publicize information for public procurement. On this platform, we can not only find recent public procurement policies but also historical records of governments’ procurement announcements, bidding invitations, and transition briefings. In my case, this platform has helped me learn what administrative agencies are the main clients of surveillance services and how Chinese governments frame commercialized Internet surveillance under the category of public procurement. Although regime-produced data may have the issue of low validity and reliability, they may still yield the least interesting observations (Ahram and Goode, 2016; Yusupova, 2019). At least researchers can also deploy various triangulation techniques to avoid the bias caused by unreliable data (Denzin, 2016; Gold, 1997; Rambukkana, 2019).

Gaining access to the surveillance industry

After choosing the appropriate field sites, researchers need to tackle another two problems associated with access: securing entry into the sampled organizations/communities and recruiting people to contribute data (Shenton and Hayter, 2004). As mentioned in the previous literature, gaining entry to the field and negotiating access can pose a challenge and in some situations be problematic, particularly in groups and communities that are excluded or difficult to reach (Thummapol et al., 2019). Upon reflecting on my fieldwork experience, I realized that the social reality resulting from the rapid evolution of tech-driven authoritarianism has not only created the phenomenon being studied, as it also opened up new horizons for research practice in the field. Qualitative researchers should utilize the change to develop creative methods of information collection.

In the context of authoritarian regimes, it is evident that various types of collaboration with local universities or local professors can help foreign researchers build connections to the field (Li 2021; Thunø, 2006). I followed this strategy when I attempted to enter the field in 2016. Since I obtained my master’s degree from a Chinese university, I asked my MA advisor to help build my initial connection to the Internet surveillance industry. He told me that his friend was a mid-level manager in a media organization that provides Internet surveillance service to government clients. Unfortunately, although my previous advisor attempted to persuade the friend to offer me a fieldwork opportunity in his organization, this strategy did not work well. The gatekeeper thought my research topic is too sensitive, and he did not think his own supervisor would agree with my entry. It is worth noting that the media organization was previously a subordinate agency of the party state. Although the reform of the media industry has changed this organization’s operating structure (Stockmann, 2013), this organization is still under the supervision of the related party agency. According to ex post facto analysis of my previous advisor, this place is exclusive to outsiders due to its government background and bureaucratic office culture.

After being rejected by my advisor’s friend, I then turned to a communications professor who had also worked in my previous university. He joined this university after my graduation, so I did not know him before. I am interested in him because this professor is not only an expert in big data but is also a co-founder of a technology company that provides Internet surveillance services in the market. I planned to rely on his industrial connections to gain access. While the professor agreed to meet with me, during our talk he directly told me that my research was too sensitive for a qualitative research study, and he suggested that I change the research focus and use social media data for analysis. Although I failed to gain access via the professor, I learned a key piece of information from my talk with him: The use of big data technologies to monitor and analyze Internet information is a popular topic in Chinese academia, and technology companies recruit many graduates who are majoring in social sciences (such as communication studies, political science, and even sociology) for data analysis in the surveillance industry.

This information implies an important change brought about by the industrialization of Internet surveillance: the engagement of social sciences in the surveillance industry. Although computational technologies provide the foundation for data-based surveillance software, technology companies still need to interpret the data they capture from the Internet. Social science knowledge/skills can help technology companies transfer complex computational language into policy language that government clients can understand. Students and faculty with an educational background in social sciences constitute the labor pool from which technology companies can hire analysists or consultants to analyze monitored Internet comments and provide policy suggestions for Internet management.

In China, the industrialization of Internet surveillance therefore not only provides students in social sciences with a new choice for a non-academic career but also creates a channel for researchers to gain access. I contacted my schoolmates, most of whom have a major in sociology or communication studies, to ask them whether they or their friends have connections to the industry. One friend, who was working in a consulting company, told me that her company has a partnership with a technology company that provides surveillance software for government clients. With her help, I established my initial connection with the industry and was then given the opportunity to visit the introduced company and recruit participants there.

The second strategy I used to gain access is via social networking software. The application of digital networking software is an important feature of tech-driven authoritarianism. A growing body of methodological literature has highlighted the use of digital communication tools to create new opportunities for researchers to approach hard-to-reach communities (Kaufmann and Tzanetakis, 2020; Ward, 2011). In China, QQ is widely used instant messaging software developed by the Chinese tech giant Tencent. This software has a group talk function called QQ group that allows users to engage in group discussions with people with similar interests and backgrounds. Similar to a Facebook group, users can search these groups by keywords and request access to the group manager. I searched the keywords “Internet-information surveillance” (yuqing jiankong) and found many QQ groups that had been created by industrial actors for networking and information sharing. I sampled three groups with the largest number of members ⁶ and sent out joining requests with my real name and a statement of research interest. Fortunately, all of my requests were approved. Joining these groups gave me the opportunity to observe what are the popular topics featured in group discussions. I noticed that the active users are mostly salesmen and data analysists. They shared information about seeking new clients and discussed their experiences of solving technological problems. Sometimes the active group members, when they were physically located in the same area, organized off-line events for networking. In addition, considering that these groups usually require people to join with their real names, positions and company information, it is highly convenient for me to seek access to certain actors as potential informants. After 2 weeks of observation, I introduced myself in these groups and asked whether anyone is interested in conducting a face-to-face interview with me. I also posted a survey questionnaire in these groups to introduce my research interest and identify the respondents who were willing to accept my interview request. All these strategies worked well. Finally, through engaging in group discussions on QQ, I recruited 15 interviewees, received three invitations for company visits, and participated in one off-line gathering that included five group members located in Beijing. ⁷

Recruiting new industrial actors as key informants

In addition to new field sites and access channels, the increasing state-market collaboration is also creating new types of actors from whom we can collect information about the operation of tech-driven authoritarianism. In the tech industry, although certain occupations (such as data scientist and programmer) are receiving increasing scholarly attention from the social sciences (Avnoon, 2021; Franklin, 2021; McFarland et al., 2016), how these occupations are connected to the operation of authoritarian power is rarely discussed. In my case, the industrial actors, especially those working in technology companies, are the specific types of non-state informants deserving more methodological attention. My fieldwork shows that at least two types of actors should be considered as key informants for researchers on China’s Internet surveillance and even the more general issue of tech-driven authoritarianism. Their occupational characteristics in the Internet surveillance industry indicate their potential to contribute fruitful information about authoritarian control in the digital era.

The first type of actor is product managers. In tech companies, a product manager is a key employee that is responsible for the success of the product. Despite the differences among business environments, product managers usually refer to the people who collect requirements from customers and manage the delivery of ordered products (Springer and Miler, 2018). In tech-driven authoritarianism, a product manager is an employee that mediates the transformation between companies’ technological capacities and the state demand for Internet control. They not only need to analyze the requirements of government clients but also need to propose tech solutions based on the companies’ budget limit, financial goals, and current resources. In my case, I interviewed six product managers working for technology companies that provide surveillance products to Chinese governments. Their rich technical knowledge and experience in interacting with clients helped me learn a lot about the production process through which technological development is used to address China’s Internet control demand. This information also helps me think about the market-state conflict in China’s tech-driven authoritarianism, which is not covered in my initial proposal. In addition, product managers often coordinate teamwork and supervise production, and their position requires that their network includes various actors within their company or even across the whole industry. In this sense, product managers can be seen as ideal gatekeepers when researchers would like to understand the ecology of the studied industry.

Knowing that I am from a western academic environment in which scholars investigate digital surveillance phenomena around the world, they show curiosity about Western business models in both commercial surveillance and state surveillance. It was discussed how when facing professional elites, self-presenting as a perceptive and well-informed sociologist is a suggested positioning strategy (Li, 2021). This strategy worked well in my interaction with the product managers. One time, when I talked about the case of Cambridge Analytica and the related dispute around it, ⁸ one product manager showed great interest, and he then brought up more political topics such as the difference between China’s Internet surveillance and western ones and how both affect people’s privacy. Another time, when I introduced how data scientists can mine sentiment data and use geographic information to visualize the result as an emotion map, the interviewed product manager was excited, and he immediately called his team member, a data scientist, to discuss whether the geographic information could be incorporated into their new product. This experience proves that an appropriate positioning strategy can enhance the informants’ recognition of the researcher and encourage them to share more insider knowledge. It also shows that, compared with government officials, product managers are relatively open-minded. I believe that this open-mindedness towards an outsider researcher is partly driven by their professional training and the culture of tech companies, which benefit the establishment of links between my academic world and their tech-industry world.

The second type of actor I suggest researchers to focus on is salespersons. Salesperson is an occupation that is uniquely positioned to manage the relations between clients and the organization. The duty of a salesperson includes interacting with customers, answering questions about products, assisting with finding the best product to match customers’ needs, and performing the final sale (Geiger and Finch, 2009). Based on my experience, I suggest salespeople as key informants for the following three reasons. First, salespeople, similar to product managers, are important mediators in bridging the interaction between surveillance companies and government agencies, but their sales experience can provide more knowledge about the bureaucratic administration of the state. In my case, to secure an order, salespeople from tech companies usually need to interact with officials in different departments and sometimes address the changes brought about by (horizontal or vertical) government divides. They have insider knowledge about which government agencies are engaged with the Internet-governance mission and what type of government actor is the decision maker. In this sense, with limited access to officials, salespeople can be considered as an alternative source for learning about the administrative process of authoritarian governments’ Internet control. Second, salespeople usually have an expanded social network that can help researchers access more industrial actors or even government actors. In my case, one of my interviewed salespersons invited me to follow him to take a business trip to Hebei to see how they open the market in rural China. He also told me he knew the local officials very well so he could help me build connections. According to my observation in those QQ groups, active sales members are usually very familiar with each other, and they even recommend clients to familiar salespeople working in other companies. Third, the sales group can also provide information about how the surveillance market is embedded in China’s political context. As was complained about by several interviewed salespeople, the sale process is largely shaped by China’s political culture. For example, local governments usually purchase surveillance software from media-owned technological companies because these companies have media resources and a political background. In my research, many details about the government-market interaction in the bidding process were shared by interviewees with a sales background.

Despite the key role of salespeople in my fieldwork, I also want to highlight that, compared with interacting with product managers, recruiting salespeople requires the use of different fieldwork approaches. In my case, as most salespeople do not have a technology-related background, they usually have high pressures in terms of sales performance (Cron, 1984), and they are not that interested about my academic knowledge in digital surveillance. Researchers need to apply more flexible strategies to establish trust with participants. For example, when I interacted with a young salesman who had just entered the industry, I could tell he had anxiety about the future of the industry and his career development. When he knew I had already interviewed 15 industrial actors, he pressed me to summarize my findings from my previous fieldwork. In that interview, I spent almost 40 min describing the initial impression I got from the industry. Positioning myself as a trustworthy and neutral industry researcher undoubtedly relieved the career anxiety of this young salesperson and helped me gain his trust.

However, simply positing myself as an expert may not always work. In some other cases, knowing I am specialist in studying Chinese governments and that I have some fieldwork experience in government agencies, the interviewed salespeople showed more interest in what surveillance software was used in my previous fieldwork sites and whether I could help them build connections. When I told them that those agencies were already their companies’ clients, I could feel that they were not as interested in me or my interview. On most occasions, the approach of “borrowing power from the powerful” (Ostrander, 1993; Undheim, 2015) was still the most efficient strategy. In all my interviewees, I found that the two most informative salespeople were introduced by their managers or sales supervisors. All these experiences show that the access to informants is a precarious, ongoing, and implicit bargaining process and can be impacted by various factors such as trust, respect, reciprocity, professional prestige or even self-esteem (Undheim, 2015).

Despite these industrial actors provide alternative choices for qualitative researchers with limited access to government actors, we should not assume that non-state actors are all open-minded and more willing to share information on politically sensitive topics. In authoritarian contexts, criticism on policies and engaging in ideological debates are still considered politically sensitive by many people and may bring interviewees with negative implications (Janenova 2019). Therefore, careful research design and flexible interview strategies should be applied to secure informants’ cooperation and minimize potential risk.

In my own research, I applied the strategy of ‘depoliticizing’ to reduce the risk. Depoliticizing refers to the practice that researchers ‘frame the research topic in a way that has the best chance of reducing any sensitivity around it’ (Art, 2016, cited in Glasius et al., 2017: 41). ⁹ To learn what frame is accepted in Chinese contexts, besides discussing with my local friends, I searched Internet surveillance (wangluo jiankong) in China National Knowledge Infrastructure (CNKI), an academic dataset that covers all scholarly literature published by Chinese journals. By reviewing the title and abstracts of retrieved Chinese literature, I find that, even though the surveillance study is largely absent in Chinese academy due to the sensitiveness of this topic, social scientists in China are still allowed to: (1) introduce social theories on surveillance (Li et al., 2008); (2) analyze Western surveillance projects such as PRISM (Xu 2013), and (3) discuss technological and practical issues in the practice of Internet surveillance (Chu and Zhu, 2017). To depoliticize my research, rather than directly asking interviewees on the political implication of China’s surveillance, my semi-structured interviews were usually started with questions on the practical details of interviewees’ daily work and their opinions on Western surveillance practices (etc., Snowden disclosures and Cambridge Analytica scandal). ¹⁰ As the interview progressed, I also flexibly adjusted my interview focus based on whether I built more trustful relation with interviewees and obtained their permission to discuss more sensitive topics.

Conclusion

In this article, I have discussed how ethnographers and qualitative researchers can explore tech-driven authoritarianism when access to governments and officials is restricted. Despite various structural constraints, the authoritarian nature of a regime does not necessarily hinder opportunities for scholars who are interested in examining the connection between digital technologies and authoritarian domination. Knowledge is produced in collaboration with all the elements that surround the studied phenomenon (Kaufmann and Tzanetakis 2020). Currently, the development of technologies and the engagement of digital technologies is changing the social reality of authoritarian domination, which might open up new opportunities for qualitative and ethnographic investigation. Taking China’s Internet surveillance as an example, my own fieldwork shows that although the state remains the main actor in promoting the political application of digital technologies in an authoritarian context, the increasing impact of the private sector and publicly accessible digital tools is providing new research space for scholars to explore state control in non-state institutions.

My article contributes to the discussion on developing a better methodology for exploring sensitive topics in authoritarian contexts. Theoretically, I argue that a wider and more nuanced understanding of the studied phenomenon will help researchers develop creative ways to collect information on it. The relationship between technological development and authoritarian domination is usually narrowly framed as a Big-Brother move solely operated by the state. The drawback of this narrow assumption includes the simplification of the studied phenomenon and the restriction of the research focus, both of which reinforce each other in the knowledge production process. Instead, I argue that tech-driven authoritarianism is a state-market complex in which various interactions between governments and non-state actors mediate the political application of digital technologies. Using this understanding as a theoretical compass could help scholars identify new fieldwork opportunities not mentioned in the previous literature.

Empirically, my article sheds light on the knowledge accumulated on doing fieldwork within high political pressure contexts. I shared my own study of China’s Internet surveillance to discuss three elements of fieldwork: where to go for fieldwork, how to secure access, and what type of informant to recruit. When the access to state institutions and state actors is limited, I suggest that scholars develop various flexible fieldwork approaches for information collection. In my case, the engagement of the private sector not only provides new non-state fields, such as technology companies and tech expos, but also new choices for informant recruitment, such as product managers and salespeople. The popularity of digital communication tools also empowers researchers to gain access to this hard-to-reach community through online discussion groups and information-sharing platforms.

My work also aims to contribute to the literature on critical qualitative research in which scholars question the dominance of the Anglo-American core and the current divide between the core and periphery in the production of qualitative knowledge (Alasuutari, 2004; Hsiung, 2012; Hsiung, 2015; Li, 2021). As pointed out by previous studies, many theoretical issues are available to frame various surveillance practices and tech-driven political phenomena in Western contexts, while an antagonistic state-society relationship is the dominant narrative directing the discussion in non-western countries such China (Hou, 2017). I believe this divide is partly rendered by the core-periphery dichotomy in the knowledge production in tech-driven politics. When the experience of non-western contexts is only portrayed as the realization of the Orwellian prophecy, scholars of non-western cases are theoretically and methodologically restricted within a narrow dialogue space in which governments and officials are perceived as the only actors that matter, even though they are usually inaccessible in fieldwork. This approach further hinders the possibility that scholars will develop various theories and methods for exploring and generalizing the complexity they encounter in the social impacts of digital technologies in the periphery.

I expect that my methodological reflection will call for more attention to the methodological innovations developed with non-western experience. The fact that my research is mainly concentrated on fieldwork in China does not mean that I regard these types of problems as being unique to this specific regime. On the contrary, many of these issues are prevalent across various countries, and given the fact that limited access is a problem in many countries, a China-based experience with various coping strategies should be able to contribute to ongoing discussions on general fieldwork methods (Thøgersen and Heimer, 2006). On the one hand, it is evident that the diffusion of China’s digital control experience is proceeding with China’s One-Belt-One-Road project and has already impacted other regimes in the Global South (Woodhams, 2020). This implies the possibility that the industrialization of Internet surveillance and expanded state-market collaboration may become a more widely applied authoritarian move in the future. On the other hand, the methodological challenges faced in research on authoritarian regimes pertain, in varying degrees and extents, to democratic politics as well (Ahram and Goode, 2016). As state-market collaboration has been widely considered as the fundamental mechanism underpinning Western democracies (Bennett, 2015; Howard et al., 2005), exploring the political engagement of non-state institutions should also be a critical methodological strategy to use to address the access problem hindering most scholars in the fieldwork research of digital politics.