English tutors’ emotions, agency, and identity development in the aftermath of China’s Double Reduction policy

2.1 Shadow education

Shadow education, also known as private tutoring, refers to compensated after-class learning in academic school subjects (Bray and Lykins 2012). Shadow education generally mimics the curriculum in mainstream schools but occurs in a for-profit context (Bray 1999). The practice has become increasingly popular worldwide because of a confluence of growing emphasis on students’ academic performance and rising competition (Dang and Rogers 2008; Zhang 2013). Shadow education involves both individual tutoring sessions and organized preparatory courses provided by educational institutions (Bray and Silova 2006), and it encompasses various tutoring modes, including one-on-one sessions, small group instruction, and large lectures (Bray 2013). Tutors in shadow education provide instructional sessions to students either individually or under the employment of tutoring institutes (Bray 2010). Depending on the circumstances, these instructional activities can occur face-to-face in different settings like homes or classrooms, or increasingly, be conducted online (Curran and Jenks 2023).

Shadow education has received increasing public attention because of widespread concern over its potentially negative effects. Extensive after-class tutoring reinforces the academic burden of students and the anxiety of parents (Wang 2022). Scholars also note that shadow education’s focus on exam-oriented skills fosters a learning culture that prioritizes exam achievements rather than lifelong learning and holistic development (Yung and Yuan 2018). Moreover, shadow education is identified as potentially contributing to social inequality since high-income families have the capacity to access extensive and superior tutoring services, while middle and low-income families are forced to direct their limited resources towards shadow education in order to avoid their children falling behind (Bray 2010; Byun and Baker 2015; Wang et al. 2024). Therefore, shadow education can function to categorize students into different social classes, further affecting their learning opportunities and schooling experiences (Zhang and Bray 2018).

2.2 Language teacher identity and agency

We treat identity as dynamic, multiple, and negotiated. From this perspective, identity is not solely a byproduct of individual characteristics and values but is also shaped by social contexts (Pennington and Richards 2016). Therefore, identity is constantly negotiated and reshaped in diverse contexts as we interact with new people or adapt to new roles (Varghese et al. 2005; Wenger 1998). A wide variety of research emphasizes the role of identity in teachers’ personal and professional development (Freese 2006; Olsen 2008; Varghese et al. 2005). Researchers highlight that teachers’ identities are multiple and encompass both personal and professional dimensions (Rodgers and Scott 2008; Varghese et al. 2005). Teacher identity is shaped by a continuous process of interpreting and reinterpreting both teaching and sociocultural experiences (Xu and Connelly 2009). A wide range of different social factors influence language teachers’ identities (Kayi-Aydar 2019), producing dynamic identities affected by both internal and external factors (Beauchamp and Thomas 2009; Flores and Day 2006; Rodgers and Scott 2008). Crucially, new contexts are found to lead to “identity stress” that prompts teachers to reevaluate their values and behaviors (Pennington and Richards 2016). Scholars have observed that teachers’ emotions are crucial to their identity and practice (Benesch 2012; Lemarchand-Chauvin and Tardieu 2018; Song 2016). For example, Song (2016: 633) highlights how teachers’ emotions integrally contour their identity, noting that how “teachers understand, experience, perform, and talk about emotions is closely relevant to their sense of self.” Song (2016: 633) concludes that “understanding emotions is crucial in order to elucidate the personal, cultural, and political components of teacher identity.” Therefore, the current investigation is attuned in particular to teachers’ emotional responses to the DRP and its aftereffects, as well as how these emotional responses contour teachers’ subsequent exercise of agency, identity development, and career choices.

Teacher emotion and identity negotiation are intrinsically linked to teacher agency (Beauchamp and Thomas 2009; Buchanan 2015; Tao and Gao 2017; Tao et al. 2024). Scholars demonstrate that teacher agency is a critical factor in teachers’ professional development, as it enables teachers to exercise their autonomy in achieving their goals (Tao and Gao 2017; Tao and Gao 2021). Teacher agency is dynamic, relational and mediated by sociocultural contexts (Tao and Gao 2021). Teachers’ interpretations of their emotional experiences can reflect their perceptions of their current situations, which further shape their agentic enactment (Miller and Gkonou 2018; Tao et al. 2024). In turn, the enactment of teacher agency can influence teachers’ emotional experiences as well (Tao et al. 2024). For example, when teachers’ agentic actions contribute to the achievement of teaching goals, negative emotions may be replaced by positive ones. Teacher agency is affected by educational policies at both institutional and national levels (Tao and Gao 2017). Because they are frequently excluded from the formulation of educational policies, teachers’ responses to policy implementation are primarily influenced by how these policies affect their teacher identities (Van Veen and Sleegers 2006). Therefore, teachers exhibit agency by engaging in actions that align with and strengthen their self-perceptions (Buchanan 2015).

3.1 Participants and procedure

Participants were recruited via purposeful sampling (Patton 2014). The criteria for participation in the study were: (1) being a Chinese national; and (2) having worked at a private tutoring institution in Mainland China when the Double Reduction Policy was announced. A total of 12 participants were recruited, including 10 female tutors and 2 male tutors. Among these participants, three were classmates of the first author and enrolled in the same Master’s program in China’s Greater Bay Area. The remaining nine participants were recruited through referrals provided by these initial contacts. The purposeful sampling used to recruit teachers yielded a geographically diverse participant pool (the 12 teachers were working in nine different cities). However, participants possessed socioeconomic privileges compared to typical tutors in Mainland China, as reflected in their comparatively high levels of education. The tutors’ experience in the private tutoring industry ranged from two to eight years. Of the twelve participants, eight were still working in the industry at the time of the interview, while four had already left the industry. Among the eight who remained in shadow education following the DRP, four were still affiliated with private tutoring institutions, while four had become independent English teachers operating outside the auspices of any formal institution. Further relevant information about the participants is included in Table 1. The authors stopped participant recruitment after conducting interviews with 12 teachers because significant data saturation had already been reached, and new data seemed unlikely to provide new codes or data relevant to the research question (Fusch and Ness 2015). It is not entirely surprising that data saturation would be reached so quickly, given that research on data saturation suggests it can occur within sample sizes of twelve and below (Guest et al. 2006).

Semi-structured interviews were conducted in Mandarin with each participant. Initial questions addressed participants’ experiences in shadow education, their feelings towards the Double Reduction policy, and their career trajectories. Follow-up questions were tailored to participants’ initial responses. For example, participants were prompted to elaborate on their motivations for working in private tutoring institutions, their career changes and adjustments necessitated by the policy, and their short-term and long-term career plans after the policy. The interviews were conducted online (i.e., Tencent Meeting), and the recorded portion of the interview lasted between 35 and 50 minutes. The recordings were automatically transcribed by a subscription-based AI service, after which the first author carefully reviewed and edited the transcripts. Next, the transcripts were translated into English by the first author.

The protocol for the interviews necessitated some caution due to the sensitive nature of the topic. The topic was sensitive because (1) it involved participants recalling events that often had resulted in a negative impact on both their standard of living and their professional identity; and (2) it involved participants making potentially negative comments about a national-level government policy. Indeed, some of the participants were concerned that their comments might reflect poorly on them, their former companies, or the DRP itself. In assuaging participants’ concerns and eliciting productive answers, the study was aided by the fact that the first author shares the same linguistic, national, and cultural background as the interviewees. The first author was able to build rapport by sharing her experience of having taught English in China. Nonetheless, the first author had never worked as a full-time English teacher and was not directly affected by the DRP – and thus could not fully relate to the participants’ experiences.

3.2 Data analysis

The data were analyzed via thematic analysis (Clarke and Braun 2017). Thematic analysis consists of systematic procedures for generating codes and identifying recurring patterns within qualitative data (Clarke and Braun 2017). By breaking down the discourse in interview transcriptions into smaller components and categorizing them into broader themes, thematic analysis allows researchers to organize qualitative data and assist in interpreting within and across participants’ experiences. In coding the data, we adopted a discursive-analytic perspective that seeks to analyze speech alongside the social and cultural contexts of the speaker as well as the broader sociohistorical context (Jenks 2020). The first author and second author reviewed the data independently to generate initial codes related to the research question. Next, they reviewed their initial codes together and collaboratively developed themes. Finally, they iteratively coded the transcripts again to further refine these themes.

4.1 Initial reactions: anxiety and concern

In the immediate aftermath of the policy’s announcement, teachers’ initial reactions were concern, anxiety, and even disbelief:

I was very anxious at first. At that time, anxiety could be described as occupying more than 90% of my emotions (Brandon).

I just felt like, “Is this my bad luck?” […] everything seemed to collapse suddenly (Edward).

At first, I thought it was impossible and that the policy would be revised in a month or two, maybe three at most (Sophia).

Teachers’ shock was accompanied by concerns about not only the policy itself but also the way that their employers responded. The teachers explained that in response to the policy, the companies they worked for immediately cut their number of lessons with students. The reduction in hours had a tremendous impact on teachers’ wages because the majority of the teachers were employed on a salary system whereby each earned a low base salary that was supplemented by additional income earned from each lesson. Crucially, teachers reported that the companies they worked for had purposefully avoided firing them directly, as doing so would have necessitated the companies paying them the “N+1” compensation. In China, “N+1” refers to a legal obligation that requires companies to compensate employees with severance pay based on the number of years they have worked for the company (denoted as “N”) plus an additional month’s salary (Lu 2023). In this context, being fired was less worrisome than having their hours cut:

[I and other teachers] were mainly concerned about whether the institution could fully provide the “N+1” compensation, as it is a significant amount of money (Grace).

By cutting teachers’ hours without firing them, companies sought to indirectly force teachers to resign. The companies’ logic was sound because teachers’ base salary was so low that many felt compelled to quit:

The base salary was only around 6,000 yuan, which isn’t enough to live on in Beijing […] At that time, I didn’t think I would be laid off. I just didn’t have any classes scheduled. Although I still received a base salary, I eventually chose to resign on my own (Sophia).

The institution implemented some unfavorable policies to push us out because if they fired us, they would have to provide N+1 compensation. So, they made it difficult for us to stay, and many of us left voluntarily. Most teachers didn’t receive the N+1 compensation because we couldn’t tolerate the institution’s measures or because there weren’t enough classes to teach (Zoey).

The teachers’ reference to the “N+1” policy illustrates how teachers’ responses and reactions to education policies are affected by broader labor and employment policies. Because some teachers in this study were not directly fired, they experienced a prolonged period of employment uncertainty, which has been shown to negatively affect even those individuals who still have a job (Mantler et al. 2005). The experiences of the teachers in this study can be contrasted with those of foreign teachers abroad working for online platforms. Those teachers could be immediately fired since they were not protected by any type of labor laws (Curran 2024).

4.2 Negative evaluations of the policy

The second theme pertains to teachers’ negative feelings about the policy itself. When looking back on the policy and evaluating the policy in terms of its stated goals of reducing competition and easing families’ stress (Ren and Zhao 2024), the teachers’ remarks were highly negative, describing the policy as “ineffective” (Grace, Jade), “meaningles s” (Edward, Sophia) and “failing to achieve its intended goals” (Brandon, Daisy, Luna). Participants grounded these negative evaluations in what they believed was the policy’s failure to achieve its intended goals:

Interviewer: Having discussed a lot about your personal experiences, to what extent do you think this policy has achieved its original goals?

Luna: I don’t feel like the Double Reduction policy has actually reduced [families’] academic burden.

Moreover, participants claimed that the issue of educational inequality, which has traditionally been seen as being exacerbated by the prevalence of shadow education (Bray and Kwo 2014), had not been mitigated. In fact, some teachers felt inequality had even worsened as a result of the policy:

I believe the policy aims to sift out children from families who do not prioritize education and redirect them towards vocational paths, while families valuing education might end up spending more on education (Edward).

Whether the burden is reduced or not is debatable, but the stratification and solidification of social classes have definitely succeeded (Sophia).

The teachers’ observations are in line with extant research suggesting that families with more substantial financial resources are more inclined to continue investing in shadow education for their children after the DRP (Liu et al. 2024). The teachers’ negative evaluation of the policy’s effects highlights a disconnect between the Chinese government’s stated aims and the reactions of a key group of stakeholders affected by the policy: English teachers working in shadow education.

Interestingly, one-third of the participants who moved into public education after the DRP said that they continued to be affected by the DRP. For example, participant Ava said she felt that she could not assign too much homework to students without risking violating the DRP. This concern is based on the fact that assigning take-home assignments to students in the lower grades is considered unacceptable under the regulations (Wang 2021; Zhao 2024). The fact that even mainstream school teachers are concerned about violating the DRP highlights that the DRP’s effects extend beyond shadow education and may affect the identity development of mainstream school teachers as well.

4.3 Evolving identity: selling knowledge versus cultivating students

Only one of the teachers remained working for the same private education institute at the time of the interview. The other teachers often directly compared their previous working experience at institutes with their current positions. In doing so, they often inadvertently constructed teacher identities along a continuum that was anchored by negative identities associated with teaching for institutes, and positive identities associated with their current jobs as either mainstream school teachers or independent private tutors. When describing their jobs at large institutes (e.g., TAL and New Oriental), they used words such as “salesperson,” “short-term,” “service industry,” and “performance-oriented” to describe their negative perceptions of their teacher identity. They contrasted these negative traits with what they perceived as ideal teacher traits, including “traditional,” “cultivating students,” and “nurturing and guiding.” The teachers’ highly negative comments about their previous institutes they worked for reflected both their disgruntlement at how they had been treated by the institutions in the aftermath of the DRP, as well as the success of the Chinese state policy’s consistent negative descriptions of shadow education institutes. For example, news articles often suggest that excessive capital investment in the after-school tutoring industry exacerbates social inequality (e.g., Zhou 2021).

In retrospectively discussing the effects of the DRP, many teachers referenced the DRP as marking an important turning point in their lives, one that moved them from a more negative position along the spectrum above to a more positive position. Comparing their current jobs with their jobs before the DRP, participants portrayed their previous working experiences and teacher identities in a predominantly negative light. They cited diminished social status, intensive workload, and a lack of personal fulfillment. Externally, they emphasized the low social status associated with being an English tutor in a private institution, regardless of the content they taught, the grade level they worked with, or their financial conditions. However, in contrast to previous studies (Xiong et al. 2022), the teachers did not attribute their low social status to a lack of academic achievements or educational qualifications. Instead, they attributed their perceived low social status as being due primarily to their jobs lacking professionalism or violating what they viewed as ideal teacher norms. For example:

[before the DRP] it felt like we must treat parents and children as if we have to hold them in the palms of our hands. It creates a very strong sense of being in the service industry for me (Brandon).

Here, Brandon used the Chinese phrase “hold in the palms of hands”（捧在手心, peng zai shou xin), which carries with it the connotation of treating something or someone very carefully. In this case, Brandon used the idiom to refer to the need to prioritize continued enrollment, highlighting that he felt that he needed to coddle parents. The comparison to being in the “service industry” contrasts the Confucian ideal of a teacher as a trusted and respected cultivator of students’ holistic development (Cortazzi et al. 2009) with the negative image of education as commodified and financialized – the exact issue that the DRP was intended to address. Crucially, Brandon’s statement reflects tutors' negative emotions about their previous position, as reflected in their comparison of a traditionally high-status job (teacher) to a traditionally low-status job (service industry). Similarly, several teachers expressed disappointment and shame that working for institutions had forced them to engage in non-teaching-related tasks, such as persuading parents to renew contracts (Brandon, Edward, Grace). Teachers argued that this behavior undermined the professionalism of their work. As with the private tutors in Xiong et al. (2022), participants felt that such behavior aligned them more closely with the socially undesirable identity of a salesperson rather than the traditionally venerated identity of an educator, which participant Luna described as “dignified and honorable.” Looking back on their experiences working for private institutes, participants argued that because the main focus of private institutes is on grades and exam scores, private tutors are seen as merely responsible for “imparting knowledge” (教书, jiao shu) rather than “nurturing students”(育人, yu ren):

Interviewer: Could you elaborate further on the differences in experience between the two roles (i.e., private tutor and public-school teacher)?

Fiona: The goal of tutoring institutions is very singular and focused, emphasizing a high-intensity, fast-track educational approach. I hesitate to call it “education” […] students addressing me as “teacher” made me feel unworthy of such a title.

Participants' negative emotions towards their previous roles were related to their personal beliefs and values regarding the role of educators. This is in keeping with Canrinus et al.’s (2012) suggestion that teachers’ ratings of their self-efficacy, job satisfaction, motivation, and occupational commitment are dynamic over time and that these factors are closely linked to their sense of professional identity. The incongruency between their actual duties as a tutor and their personal understandings of a teacher as a holistic mentor resulted in negative emotions that prompted critical self-evaluation.

Although leaving shadow education institutes was seen as a positive move in terms of teachers’ self-perceived social status, it also brought with it considerable uncertainty. One participant described the status of private English tutors in the aftermath of the DRP as follows:

Working as an independent teacher is a bit like being in a field army (野战军, ye zhan jun) (Ava).

The term “field army” refers to the People’s Liberation Army (PLA) and its operations during the Chinese civil war, marked by mobility, guerrilla tactics, and constant evasion of the enemy (RUSI 2004). Ava's use of this metaphor suggests that following the DRP, the role of private tutors has become increasingly marginalized, lacking stability or clear career progression. This situation affected teachers emotionally and prompted reevaluations of their identities, ultimately leading them to reexamine the role and meaning of being a teacher vis-à-vis emerging discursive differences between “private” and “public” teaching. The divergence in policy practices between public and private education systems contributed to significant challenges for tutors attempting to work in public school positions or to remain in the shadow education sector (Yang et al. 2024). Overall, the DRP forced teachers to re-evaluate their identities in light of China’s changing social, cultural, and economic environment.

4.4 Reframing the DRP as a positive turning point

The third and final theme relates to teachers’ somewhat contradictory belief in the positive role of the DRP in their lives. Although the teachers spoke about the anxiety they experienced at the policy’s announcement as well as their belief that the policy has been ineffective, they spoke positively of the DRP’s effects on their own lives. Many teachers claimed that the DRP forced them to improve themselves. Indeed, more than half of the participants came to Hong Kong to pursue higher education in the years immediately following the DRP:

Interviewer: Since you mentioned earlier that working as an English teacher in an institution might not be a long-term plan, I was wondering if this policy led to significant changes in your career plan?

Edward: My goal was to take some time to pursue a master’s degree to enhance my professional skills. I wanted to see if I could develop, improve, and expand my expertise and teaching scope. At that time, I was uncertain about the stability of being an English tutor, so I decided to enrich myself.

As the quote above illustrates, many teachers responded to the uncertainty they faced after the DRP by trying to improve their competitiveness in the job market. Crucially, all the teachers who pursued an MA degree studied English Language Teaching, which they chose to either enhance their teaching skills or increase their employment opportunities:

Interviewer: Did the policy affect your career planning, both in the short and long term? Have you made any adjustments?

Daisy: When this job became less stable, I thought to myself, “Why not take this opportunity to improve myself further?”

A similar reply to the same question was offered by Ava, who observed: “If you truly master the art of teaching, it’s a valuable skill that ensures you can make a living anywhere you go.”

Evident in the teachers’ narratives is their collective embrace of a neoliberal mindset wherein individuals embrace continuous self-improvement (Shin and Park 2016). Tao and Gao (2017) suggest that teachers exercise agency by actively engaging in continuous learning to improve their teaching competencies. In this study, many English tutors exercised agency by pursuing higher education and then transitioning to new career roles, such as mainstream school teachers. While demonstrating teachers’ exercise of agency, these actions also reflect the teachers’ internalization of neoliberal subjectivities that prioritize continuous self-improvement in a highly competitive teaching environment (Curran 2024).

The DRP was intended to counter the endless competition associated with neoliberalism (Read 2009). However, by vastly reducing the number of jobs available to English teachers, the policy can be understood as introducing increased precarity for individual teachers. The result was internalization of neoliberal subjectivity and the embrace of self-improvement as an ongoing and ceaseless goal (Park 2021). Participants’ pursuit of continuous self-improvement is unsurprising given that a previous study about the DRP similarly found that the policy spurred some teachers to pursue higher education (Lyu and Lam 2025). Indeed, teachers’ agentic actions are affected by both personal and contextual factors (e.g., personal beliefs as well as policy changes) (Tao et al. 2024). We argue that participants’ pursuit of higher education can be interpreted as an agentic choice to mitigate the downsides of the DRP while also simultaneously signaling a commitment to developing their professional identities as language educators. Crucially, the teachers identified the DRP itself as a crucial catalyst in their decision to invest in their human capital. Asked to reflect on their current “feelings about this policy,” participants gave surprisingly similar answers:

If this policy didn’t exist, I might still be at [the company] today, or even still living in confusion. In this case, I was essentially lost for four years. From this perspective, it has actually benefitted me, the benefits outweigh the drawbacks (Brandon).

Without this policy, I might not have had the determination to make changes (Edward).

If the Double Reduction policy hadn’t been introduced, I might have naturally given up on retaking the postgraduate exam and continued working at the institution because I felt I had a talent for teaching (Sophia).

If the policy hadn’t been implemented, I would have stayed at the institution, and my time would have been wasted (Zoey).

The teachers' positive evaluation of the DRP’s impact on their own lives starkly contrastes their negative assessment of the ban’s overall effects on curbing social inequality. Crucially, even those teachers who eventually chose to continue working in shadow education as private tutors expressed their appreciation that the DRP had forced them to “grow”:

[Without the DRP] I might still be at New Oriental, stuck in a rut, with no time to grow, like “boiling a frog in lukewarm water” (温水煮青蛙, wen shui zhu qing wa) (Zoey).

The teacher’s use of the “boiling a frog” metaphor implies that without the DRP, her career would have gradually but surely deteriorated. In this context, the policy served as a catalyst, prompting the teacher to focus on aspirations she may have previously hesitated to pursue. Buchanan (2015) suggests that teachers’ emotions are linked to negative feelings when they are unable to achieve their objectives, which prompts them to exercise agency to effect change. In the aftermath of the DRP, some English teachers found that their identities no longer aligned with their professional identities and the evolving social context or cultural expectations. This prompted them to critically reflect on both their identities and the nature of shadow education. For example, one participant currently working in a public school reported that a student-centered teaching philosophy was more consistent with her personal values than the exam-oriented approach prevalent in shadow education. Consequently, teachers viewed the DRP as an important opportunity to redefine and improve themselves to better align with their desired identities, with a particular emphasis on enhancing their professional competence. In other words, while teachers were pessimistic about the societal impacts of the policy, most were highly positive about the effects that the policy had on their personal career trajectories. However, it is also important to note that the teachers who expressed positive emotions about the DRP’s impact on their identity and career development were the ones who had pursued higher education. Yet, to pursue an MA degree, these teachers necessarily possessed higher levels of economic capital than the average tutor in China. Relatively less privileged tutors, for whom job loss was not followed by educational advancement, might have expressed less positive attitudes in their reflections on the impact of the ban than the participants in this study.

According to the participants’ accounts, the implementation of the DRP forced them to reflect on and reassess their career trajectories. Interestingly, although the teachers framed their former jobs at tutoring institutes negatively due to the “salesperson” nature of the work, they also embraced an entrepreneurial mindset whereby they started to view themselves as salable products, or “companies of one” (Read 2009). This sentiment was reflected in a participant’s (Ava) assertion: “If you truly master the art of teaching, it’s a valuable skill that ensures you can make a living anywhere you go.” These reactions are surprisingly similar to those of Western online English teachers discussed in Curran (2024), who lost their jobs teaching Chinese children as a result of the DRP. Like the teachers in Curran (2024), the tutors in the present study responded to the DRP with anxiety and fear, yet ultimately concluded that the DRP was beneficial in that it forced them to move out of their comfort zones and improve themselves. The similarity in these findings underscores that neoliberal discourses of self-entrepreneurship are common among shadow education tutors from different national contexts, reflecting the dominance of these discourses globally (Holborow 2015; Shin and Park 2016).