Language Standardization under Authoritarianism

Previous research on standard language promotion highlights two major roles of language: defining identity and facilitating communication. Regarding their identity-defining roles, studies have shown that a standard language fosters a shared identity among fellow speakers of an ‘imagined community’ in nation building (Anderson Reference Anderson2006). A classic example is the promotion of the French language among rural populations, which led to the creation of a French identity in the late nineteenth century (Weber Reference Weber1976). Similarly, in the post-colonial context, the promotion of Tanzania’s Swahili (Blommaert Reference Blommaert1996; Miguel Reference Miguel2004) and Indonesia’s Bahasa (Anderson Reference Anderson2006; Bazzi et al. Reference Bazzi, Gaduh, Rothenberg and Wong2019) promoted a strong national identity while diluting subnational identities prone to conflicts. Following the collapse of the Soviet Union, the replacement of Russian with indigenous languages in newly independent states also helped create a shared national identity (Fierman Reference Fierman2009; Laitin Reference Laitin1998).

The identity-defining role of standard languages becomes particularly significant in authoritarian regimes, as it can be instrumental in strengthening their rule. One common source of authoritarian legitimacy is nationalistic appeals (Cannady and Kubicek Reference Cannady and Kubicek2014; Dukalskis and Lee Reference Dukalskis and Lee2020; Nathan Reference Nathan2020). Existing studies show that language standardization enables autocrats to generate such appeals, either by uniting diverse groups or by reinforcing the status of the dominant group. Some leaders have adopted an ethnically neutral language to unite groups.Footnote ² As mentioned, Tanzania’s Nyerere and Indonesia’s Sukarno and Suharto promoted Swahili and Bahasa, respectively; both languages were lingua francas rather than the mother tongue of any major group, helping to establish a strong national identity across populations and thus bolstering the cause of the rulers (Blommaert Reference Blommaert1996, 247). Others have pursued a more contentious strategy, imposing the ethnic majority’s mother tongue as the national language. Thailand’s seemingly homogeneous society today is the result of a strong monolingual policy initiated in the early twentieth century under King Chulalongkorn, who sought to assimilate Chinese communities perceived as a threat (Liu and Ricks Reference Liu and Ricks2012). Without necessarily forcing assimilation, Malaysia’s authoritarian-leaning government under the National Front shifted from multilingualism to monolingualism in 1971, thereby enhancing the regime’s legitimacy through appeals to Malay nationalism (Liu and Ricks Reference Liu and Ricks2012; Ostwald and Subhan Reference Ostwald and Subhan2021).

In contrast to the prevalent focus on identity, a second strand of literature examines the communicative function of language. While language is often treated as a proxy for ethnic identity, individuals commonly speak a repertoire of languages to interact with non-coethnics and to navigate different levels of government, which may operate in local, regional, or official state languages (Laitin Reference Laitin1992). Liu and Pizzi (Reference Liu and Pizzi2018) show that the purported economic advantages of ethnolinguistic homogeneity extend to countries with substantial non-native speakers conversant in official languages. Marquardt (Reference Marquardt2018,Reference Marquardt2022) further demonstrates that fluency in a metropolitan language grants access to state services and high-status group membership, affecting policy preferences independently of ethnic identity.

From this perspective, the promotion of a standard language is crucial for facilitating communication between the state and society, thereby supporting state building. Drawing on Max Weber’s concept of ‘rationalization’, which includes standardizing calendars, measures, and currency, Laitin (Reference Laitin1992) argues that language rationalization, defined as ‘the territorial specification of a common language for purposes of efficient administration and rule’ (9), helps the state collect taxes, monitor citizen activities, and enforce regulations through a shared language. Along similar lines, scholarship on ‘legibility’, first introduced by Scott (Reference Scott1998), suggests that the rise of mass literacy during state and nation building reduced communication costs between citizens and officials, enhancing the state’s information-gathering capacity essential for tax collection and regulation of private affairs (Lee and Zhang Reference Lee and Zhang2017; Zhang and Lee Reference Zhang and Lee2020).

In authoritarian contexts, the communication-facilitating dimension of standard language promotion remains underexplored, particularly in comparison to its more extensively studied role in generating nationalistic appeals. This gap is striking given the pervasive information control under authoritarianism, including censorship, surveillance, and propaganda. A notable exception is Yue Hu’s (Reference Hu2024) study on China, which – building on his earlier work (Hu Reference Hu2018) – finds that citizens with higher listening proficiency in the official language are more likely to trust and align with the government (Hu Reference Hu2024, chapter 3). This operates through two channels. At the cognitive level, the official language acts as the voice of political authority (Hu Reference Hu2024, chapter 4; see also Hu Reference Hu2020). At the informational level, listening skills help citizens better receive and understand state messaging. Although the latter channel is not fully developed in the analysis, it points to an important implication: standard language promotion can strengthen authoritarian rule by enhancing the reach of propaganda. This deserves special attention, as examining propaganda highlights the state’s capacity not only to collect but also to disseminate information, an equally vital but understudied aspect of state–society communication.

Extending Hu’s work, this study seeks to identify the causal effect of the state’s deliberate promotion of standard language and further explore the propaganda mechanism. Using the case of China, I hypothesize that standard language promotion increases citizens’ satisfaction with the government, and that exposure to putonghua increases television news consumption – a key channel for state propaganda delivered exclusively in putonghua. The scope condition is that promotion efforts are aimed at facilitating communication rather than defining identity. China presents a suitable case, given its emphasis on promoting a ‘common speech’ for communication rather than imposing national symbolism (Chen Reference Chen1999; Simpson Reference Simpson, Chan and Minett2016),Footnote ³ at least until recent shifts yet to be discussed. Furthermore, a large part of standard language promotion is directed towards speakers of regional Chinese varieties within the Han majority – China’s dominant ethnic group, comprising over 90 per cent of the population. That said, the hypotheses may be less applicable in countries where language policy is more closely linked to identity, especially in highly ethnically diverse societies.

Putonghua Promotion in China

China’s official language is ‘putonghua’, which translates to common speech and is also known as standard Beijing Mandarin. By 2020, 80.72 per cent of the population was reported to be proficient in putonghua, establishing it as the lingua franca of the country (China News Service 2021). While this figure may be an upper-bound estimate – presented by the Ministry of Education to underscore the achievements of its putonghua promotion initiative – the contrast with the same agency’s 2000 report of 53.06 per cent suggests that proficiency has risen considerably. However, putonghua is not the first language for most people in China. A nationally representative survey conducted in 2010 indicates that only 14.4 per cent of the population speak putonghua as their primary home language. These individuals are largely concentrated in northeastern regions: Beijing and the capital, as well as Liaoning, Jilin, and Heilongjiang provinces (Figure 1).

Figure 1.

Percentage of individuals primarily speaking putonghua at home by province.

Note: created by author based on the 2010–12 China Family Panel Studies. Provinces shaded in gray indicate areas not covered by the survey. Home language information is available for 94.4 per cent of the 2010–12 survey participants who were born on or before 2002 (aged 10 and above), totaling 44,375 out of 47,020.

The vast majority of the population speaks various other languages, including not only ethnic minority languages such as Mongolian, Tibetan, and Uyghur but also regional varieties of Chinese such as Cantonese, Shanghainese, and Sichuanese. Although these regional varieties share written characters with putonghua, they differ significantly in pronunciation, vocabulary, and at times grammar, making their spoken forms mutually unintelligible. Many are predominantly oral; when written, speakers may resort to rare or non-standard Chinese characters to approximate local sounds. For this reason, speakers of regional varieties, especially those more distant from putonghua, describe learning putonghua as comparable to learning a foreign language (Interviews R1-2021, R8-2023, R10-2023, R33-2024). Even speakers of Mandarin dialects, though linguistically closer to putonghua, report mutual unintelligibility, particularly in rural areas (Interviews R13-2023, R14-2023, and R35-2024 on Liaoning, Shandong, and Sichuan dialects).

Given this complex linguistic landscape, Chinese leaders have long sought to promote putonghua. Although their commitment fluctuated over time,Footnote ⁴ a significant milestone was reached in 2001 with the Law on Standard Spoken and Written Chinese Language.Footnote ⁵ As the first national law on language, the law actively mandated the use of putonghua in state organs, education, publication, broadcasting, and public services (Articles 9–13).Footnote ⁶ The law’s legislative goal was three-fold: ‘economic and cultural exchange’ (Article 1), ‘national unity and ethnic solidarity’ (Article 5), and ‘the construction of socialist material and spiritual civilization’ (Article 5). The first two were often intertwined, as the government aimed to encourage economic transactions across language groups while integrating minorities into this process (Lin and Jackson Reference Lin and Jackson2021). Importantly, the third objective was fundamentally about propaganda. According to a legislator involved in drafting the law, it entailed aligning the form and content of putonghua with socialist ideology, while countering the spread of ‘corrupt ideological content like colonialism, feudalism, pornography, and vulgarity, which obstructs the progress of socialist civilization’ (Wang Reference Wang2000).

Following the national enactment, provincial-level governments adopted local regulations to implement the law, with the timing varying from as early as 2002 to as late as 2014. This uneven pace may reflect a mix of factors. Administrative capacity was crucial, since compliance required various resource-intensive measures such as training putonghua-speaking teachers and administering proficiency exams. In provinces with substantial ethnic minority populations, implementation was likely further complicated by the question of how strongly to prioritize putonghua over minority languages (Zhang and Cai Reference Zhang and Cai2021) – a key concern in the drafting process that generated tensions between education departments and ethnic affairs commissions at the national level (Rohsenow Reference Rohsenow, Zhou and Sun2004, 35–6; Zhou Reference Zhou, Zhou and Sun2004, 91). In addition, provinces with strong local cultural pride, like Guangdong, appeared slower to comply, partly due to popular resistance against putonghua promotion (Gao Reference Gao2015) and implicit local government endorsement evident in the continued support for Cantonese in public by Guangdong’s senior officials even in 2013 (Guangdong Department of Education 2013). Finally, variation may also have reflected shifting central expectations, with greater attention devoted to global cities and to politically sensitive peripheral regions.

Despite variation in timing, the law was ultimately implemented nationwide and brought real changes to people’s daily lives, particularly in schools. Previously, many schools utilized regional languages or, less frequently, ethnic minority languages as their instructional language; putonghua was taught as a class subject. However, the 2001 law strongly mandated that putonghua should become the primary medium of instruction. The implementation particularly focused on elementary and middle schools, which constitute the compulsory education period in China. As depicted in Figure 2, the transformation was remarkable, especially at elementary schools. The percentage of elementary school fourth- to sixth-grade students primarily speaking putonghua at school was only 36.21 per cent in 2010, but soared to 67.22 per cent in 2020. Aligned with this trend, several interviewees from different provinces shared similar observations: in their extended families, which primarily speak dialects, only the younger members – roughly those born in the 2000s – speak putonghua even when others speak to them in dialects, likely because they use putonghua at school (Interviews R14-2023, R19-2023, R21-2023).

Figure 2.

Percentage of students primarily speaking putonghua at school by survey round.

Note: created by author based on the 2010–20 China Family Panel Studies surveys. School language information was collected for students aged 10 and above (elementary fourth grade and above). On average across surveys, the information is available for 94.26 per cent of these individuals. Solid lines indicate the period of compulsory education, the primary realm of putonghua promotion efforts.

The political consequences of this dramatic shift in the language practice, however, remain to be examined. Through the lens of standard language promotion as a tool for autocrats, the subsequent sections investigate the impact of putonghua promotion on citizens’ government satisfaction in China.

Findings

Before presenting the main findings, I first show descriptive evidence regarding citizens’ government satisfaction based on the putonghua cohort and local compliance with the putonghua promotion mandate. To reduce noise for visualization, I reorganize the cohort into three-year segments, determined by the age of respondents at the time of adopting the provincial regulation. Specifically, I put respondents aged 1 to 3 at the time of adoption into one group, those aged 4 to 6 into another, and so on. I also divide local compliance levels into two groups: ‘high’ compliance for counties above the median of the sample, and ‘low’ compliance for those below the median. Figure 5 depicts the mean government satisfaction for each three-year grouping, for high- and low-compliance counties, respectively.

Figure 5.

Government satisfaction by putonghua cohort and compliance.

Note: groups 1–3 (aged 16 and above at the time of provincial legislation), situated to the left of the left vertical line, were untreated by the reform. Groups 6–8 (aged 9 and below), situated to the right of the right vertical line, were fully treated. Groups 4–5 (aged 10–15), situated in between, were partially treated.

Among cohorts unaffected by the reform, government satisfaction is consistently lower in high-compliance counties compared to low-compliance counties, indicating a parallel trend (see the Supplementary Material for further details). However, in high-compliance counties, the satisfaction level markedly increases for cohorts partially affected by the reform and surpass that of low-compliance counties for cohorts fully affected by the reform. In contrast, satisfaction levels in low-compliance counties exhibit relatively mild changes across cohorts.

The main results regarding the effect of putonghua promotion on citizens’ government satisfaction are presented in Table 1. In Columns 1–3, I use a bandwidth of 0, covering individuals born between 1986 and 2005. For Columns 4–6, I use a bandwidth of 3, covering those born between 1983 and 2008. Columns 1 and 4 are reduced models without controls. Columns 2 and 5, my preferred specification, incorporate controls. Columns 3 and 6 add county-level controls, combined with province fixed effects. Across the models, the estimated effect of putonghua promotion ranges from roughly 0.50 to 0.60. Given the sample mean of roughly 3.5, the coefficients indicate that exposure to putonghua at school increases government satisfaction by about 15 per cent. Substantively, an effect of 0.5–0.6 means that, in counties with full compliance, individuals in the treated cohort report government satisfaction about 0.5–0.6 points higher than comparable individuals in the untreated cohort. To more directly link this interpretation back to the underlying survey items, which asked respondents to rate the severity of various problems in the country on a 0–10 scale, this 0.5–0.6 increase in government satisfaction corresponds to individuals viewing these governance problems – including government corruption, environmental protection, economic disparity, employment, education, medical service, housing, and social security – as 0.5–0.6 points less severe.

Table 1.

Effect of putonghua promotion on government satisfaction

Note: coefficients are estimated using ordinary least squares. For columns 2 and 5, control variables include gender, ethnic minority or not, urban or rural residency, parents’ highest education, and the family’s relative income level. For columns 3 and 6, they additionally include county GDP per capita (logged) and county urban population proportion. For columns 1, 2, 4, and 5, fixed effects are for county and birth year; standard errors are clustered by county. For columns 3 and 6, fixed effects are for province and birth year; standard errors are clustered by province. **p < 0.05; ***p < 0.01.

To better illustrate how the effect of putonghua cohort varies by compliance level, I further estimate the following modified regression equation:

(2) $y_{itpcg} = \sum _{q=1}^{5} \beta _q(putonghuacohort_{pt} \cdot compliance_{cq}) + {\bf X}^\prime _{itpc}\Gamma + \zeta _{c} + \mu _{t} + \epsilon _{itpc},$

where q denotes quintiles of compliance levels. Figure 6 presents the results. Across the five compliance levels, the effects clearly exhibit an upward linear trend. This suggests that the putonghua promotion has negligible impact on government satisfaction in counties with lower compliance but effectively increases satisfaction in those with higher compliance. The effect becomes statistically significant at the highest compliance quintile.

Figure 6.

Effect of putonghua promotion on government satisfaction by compliance levels.

Note: thick black lines represent 90 per cent confidence intervals, and thin gray lines represent 95 per cent confidence intervals. The cohort bandwidth is 0.

The findings remain robust across various alternative specifications, as presented in Table 2. In column 1, I exclude individuals categorized as partial treats (age 10–15). In columns 2–5, I employ alternative coding schemes for putonghuacohort _pt . First, I lump three-year cohorts within the partial treatment category, allocating the value of 1/3 to ages 13–15 and 2/3 to ages 10–12. Secondly, I lump all six-year partial treats and assign 1/2. Thirdly, I assign non-linear values that decline quadratically, assuming that earlier exposure may have stronger effects ( $1-({{\rm age}-9}{7})^2$ ). Finally, I shift the time frame by one year, categorizing individuals aged 17 and above (instead of 16) as untreated, and those aged 10 and below (instead of 9) as fully treated. This adjustment accommodates variations in the school entry age, typically 6 or 7. In column 6, I consider those attending high school at the time of adoption as partially treated rather than untreated and use the scale spanning from 1/10, 2/10, to 9/10. As further shown in the Supplementary Material, the findings remain robust to using (1) alternative compliance measures, (2) disaggregated government satisfaction indicators, (3) the stacked difference-in-differences approach, (4) various bandwidths of cohorts (0 to 10) included in the analysis, and (5) previous survey rounds.

Table 2.

Robustness checks results (bandwidth = 0)

Note: coefficients are estimated using ordinary least squares. Control variables include gender, ethnic minority or not, urban or rural residency, parents’ highest education, and the family’s relative income level. Fixed effects are for county and birth year. Coefficients for the (county-level) compliance variable are not shown as they are absorbed by county fixed effects. Standard errors are clustered by county. *p < 0.1; **p < 0.05; ***p < 0.01.

In the Supplementary Material, I show the heterogeneous effects between Han and ethnic minority respondents. The results suggest that the positive effect of putonghua promotion on government satisfaction is present for both groups.Footnote ²² However, since most ethnic minority respondents in the sample reside outside autonomous regions (see footnote 7), these results should not be generalized to all ethnic minority populations in China. I leave a more comprehensive examination of ethnic minorities’ responses to putonghua promotion for future research.

Mechanisms

Why might putonghua promotion at school increase government satisfaction? To explore the mechanisms, I now turn to individual-level evidence by utilizing school language records from all rounds of CFPS survey between 2010 and 2020.Footnote ²³ Consistent with the main analysis, I focus on individuals who reside outside of putonghua strongholds and whose mother tongue is not putonghua. To preview, I find that early exposure to a putonghua-speaking environment at elementary and middle school is associated with increased consumption of political news via television, a key channel for state propaganda delivered exclusively in putonghua.Footnote ²⁴

Some caveats are in order. First, compared to the main analysis, the individual-level analysis involves a slightly younger population, as it covers those who were attending elementary or middle school (fourth to ninth grades) during the period of the 2010–20 surveys. Language records from elementary school are largely available for those born between 1997 and 2010, and for middle school, between 1994 and 2007. Secondly, these findings do not establish causality. While the evidence is only suggestive, it provides useful insights into how the shift in the medium of instruction affects individuals’ behavioral patterns, which may influence their attitudes towards the government. In the Supplementary Material, I apply the main causal inference design to television political news consumption, which shows a statistically significant increase at the 0.10 level.

Television news remains one of the most powerful propaganda tools in China (Di Reference Di and Shirk2011; Pan et al. Reference Pan, Shao and Xu2022). Its content is regulated by the National Radio and Television Administration, which reports directly to the Propaganda Department of the Chinese Communist Party (Pan et al. Reference Pan, Shao and Xu2022). Sensitive topics never appear on television – more strictly than in books and newspapers – and even local news channels are closely supervised by propaganda departments at each administrative level (Di Reference Di and Shirk2011). Importantly, coverage of national news – more relevant to this study’s focus on citizens’ satisfaction with the (central) government – is effectively monopolized by China Central Television (CCTV). Each night at 7 p.m., CCTV’s flagship newscast, Xinwen Lianbo, is broadcast simultaneously across dozens of central and provincial channels. As of 2019, it drew about 55–60 million viewers – roughly 4.6 per cent of China’s 1.2 billion population (Pan et al. Reference Pan, Shao and Xu2022). The content, language, and format have remained largely unchanged over time (Huang Reference Huang2015), typically featuring government leaders’ activities and economic, social, and scientific achievements shown in a positive light.

In this context, television political news consumption may persuade viewers to hold positive evaluations of government performance. The influential studies by Huang (Reference Huang2015,Reference Huang2018) suggest that news programs in China, especially Xinwen Lianbo, are overtly propagandistic, serving less as instruments of persuasion than as signals of the government’s strength through the very act of propaganda. However, there is evidence that such programs can still be persuasive. For instance, Pan et al.’s (Reference Pan, Shao and Xu2022) experimental study shows that messages in the typical Xinwen Lianbo format can effectively sway people’s policy positions closer to the government through framing. Furthermore, Hu and Shao (Reference Hu and Shao2022) find that there is a non-trivial number of intentional viewers who, by watching the news, seek confirmation of social stability or exposure to positive stories such as the government’s anti-criminal or anti-corruption efforts and national achievements. Taken together, television news programs can – at least for some – change their views of the government or reinforce already-positive orientations towards it. Notably, so-called ‘soft propaganda’, such as emotional films and television dramas, has been found to foster more nationalistic attitudes but not necessarily more positive evaluations of government performance (Mattingly and Yao Reference Mattingly and Yao2022). By contrast, this study’s focus on television news makes it plausible that it influences citizens’ attitudes towards government performance. Consistent with this reasoning, the CFPS 2020 survey shows that political news consumption on television is positively associated with government satisfaction (p-value < 0.01).

The television news is delivered exclusively in putonghua. Interviews reveal two major observations in this respect. First, putonghua-exclusive television is less appealing to viewers not proficient in the language. An ethnic minority interviewee from south-central China noted that many elderly people in their neighborhood rarely watch television news because they neither understand the spoken language nor the subtitles in Chinese characters (Interview R21-2023). Another interviewee from a Cantonese-speaking region mentioned that they used to watch Hong Kong TV shows instead of Chinese state media, as these shows were in Cantonese and accessible in their area due to its proximity (Interview R18-2023).Footnote ²⁵ Secondly, people also learn putonghua through watching television news. One interviewee mentioned that some parents intentionally turn on the news to expose their children to the most standard pronunciation of putonghua (Interview R17-2023). Additionally, several interviewees noted that although their parents’ generation may not be fluent speakers of putonghua, they have developed listening skills from watching television shows (Interviews R8-2023, S2-2023, S3-2024). This aligns with online discussions recommending television news as an effective method for improving putonghua skills.Footnote ²⁶

Combined, I hypothesize that individuals exposed to a putonghua-dominant environment at school are more likely to consume television news either because of their heightened proficiency in putonghua or to enhance their putonghua skills. I test this hypothesis in two ways: one using cumulative records of elementary and middle school language from all 2010–20 surveys, and the other using current records from the 2020 survey. In the cumulative-record approach, I first identify individuals who attended elementary (middle) school during the survey period. I then code whether or not they reported putonghua as their primary language at least once during elementary (middle) school. Finally, I take unique individuals from all 2010–20 surveys, prioritizing their most recent responses. In the current-record approach, I simply focus on elementary and middle school students in the 2020 survey, coding whether or not they reported putonghua as their primary school language. For the outcome variable, I use a question on television political news consumption in the past week, with responses ranging from 0 to 7 days.Footnote ²⁷ Among the individuals included in the analyses, the average consumption is approximately 1.5 to 1.7 days in a week.

Figure 7.

Effect of putonghua usage at school on TV political news consumption.

Note: for panel (a), I regress whether respondents reported putonghua as their primary language at least once during elementary and middle school, separately, on television news consumption. The reduced models omit controls. The full models incorporate controls: gender, ethnic minority or not, urban or rural residency, parents’ highest education, and the family’s relative income level. The reduced and full models include county, birth year, and survey fixed effects; standard errors are clustered by county. The models with county-level controls add counties’ GDP per capita (logged) and urban population proportion, with province, birth year, and survey fixed effects; standard errors are clustered by province. For panel (b), I regress whether respondents reported putonghua as their primary language at school, interacted by the school level (elementary or middle), on television news consumption. All other specifications are the same as panel (a), but with an additional control for the school level and without survey fixed effects.

Figure 7 visualizes the results. The estimates indicate that exposure to a putonghua-speaking environment during middle school – compared to during elementary school – has clear and statistically significant positive effects on television political news consumption. The effect is particularly pronounced when using the current records, as shown in panel (b): middle school students who currently primarily use putonghua at school consume political news via television approximately 0.45 days more in a given week than their peers who do not. Exposure during elementary school produces effects that are insignificant or marginally significant, yet the directions are consistently positive and aligned with the middle school estimates.

There exist at least two alternative mechanisms linking exposure to putonghua with government satisfaction: first, the reinforcement of national identity or nationalistic attitudes due to greater reliance on the country’s official language; and secondly, improved access to resources such as education that are available only to individuals proficient in putonghua. While I do not draw definitive conclusions regarding these mechanisms, I offer preliminary evidence that may guide future research (see the Supplementary Material for details). First, individuals’ exposure to putonghua in elementary and middle school does not appear to be associated with a stronger nationalistic attitude, as proxied by a question on anti-West sentiment, suggesting that this mechanism is less plausible.Footnote ²⁸ Secondly, such exposure does appear to be positively associated with enrollment in post-secondary education – opening the door to high-income jobs and elite networks, lending more plausibility to this mechanism. Still, a thorough examination is needed to ascertain whether there is a causal relationship and whether this translates into greater satisfaction with the government, since education is also associated with more critical political attitudes in China.

Beyond education, improved access to other resources like government services – particularly those requiring interactions with street-level bureaucrats such as at community service centers, complaint centers, and police stations – represents another related mechanism, though this study does not directly explore it. In many government offices, signs read ‘Please speak putonghua’, and one interviewee noted that because her father is not proficient in putonghua, he lets her mother handle all public communication (Interview R36-2024). Such dependence likely reduces one’s ability to engage with government services and diminishes the chance of satisfactory first-hand experiences. Overall, the proposed mechanism of TV political news consumption does not exclude these alternative pathways; further research is needed to examine these dynamics more comprehensively.