Protecting Farmers and Workers in Socialist Market Transitions: Mass Attitudes Toward Imports in Asia

After decades of market liberalization in Asian countries, how do people see trade, mainly imports, as potentially hurting domestic production? Do they consider them a threat to their societies’ less privileged groups? Using survey data of China, Vietnam, and Cambodia, and citing Thailand as a comparative case representing a liberal economy in the Southeast Asian region, we examine the effect of economic interests, collectivist ideology, and political trust on trade preferences. Empirical results indicate that economic optimism is only weakly related to attitudinal openness to imports, while income has a mixed effect. Collectivism and political trust in the central government and the elite stimulate antipathy against imports. The findings shed light on the sources of protectionist attitudes in Asian authoritarian regimes undergoing socialist market transition among the general public.


Introduction
China, Vietnam, and Cambodia are three prominent cases of socialist market transition in communist countries in Asia. They have endeavored to craft a market economy after decades of a socialist economy. Subnational governments are encouraged to experiment with innovative growth models. Many grass-roots enterprises, whether owned by towns and villages, para-state entities, or private households, have seized the opportunity to benefit from deeper involvement in global commodity chains (Beresford, 2008;Hill & Menon, 2014;Malesky & London, 2014). The three socialist regimes seem to have shifted their development onto a different, marketoriented track despite their governments remaining quintessentially authoritarian, a lamentable condition as castigated even by the most recent research on their state-society relations. (Blake, 2019;Chang, 2022;Hirsch, 2020;Howell & Pringle, 2019). Evolving along with their remarkable economic growth is a new trend of value changes that celebrate neoliberal ethics of personal responsibility, private interests, self-cultivation, expressive emotion, and desires. As Springer (2010) documented in the case of Cambodia in the post-conflict era, the socialist market transition has represented a society-wide justification of these neo-liberalized values, which involves no less than the restructuring of the rational relationships between individuals, the state, and local and global markets (Springer, 2009;Tran, 2015;Yan & Sautman, 2010).
However, do the mass attitudes tilting toward market neoliberalism supplant the idea of reduced state intervention in these authoritarian regimes with a façade market economy? Undoubtedly, the institutional changes in the economy, education, and social mobility have given birth to a calculating, proactive, and self-disciplined selfhood, particularly among the young professional class (King et al., 2008). However, cherished individualism and neoliberal ethics do not necessarily incur the idea of a lesser state for the public. Socialist market transitions have aroused acute tensions, inciting concerns over how authoritarian regimes should distribute economic opportunities, benefits, and social costs. Empirical evidence from large-scale social surveys shows mixed results. More robust state intervention seems acceptable for better social protection (Wong & Lee, 2000). Vietnamese, however, display fervent attitudinal support for the private economy, market competition, tolerance of greater income inequality, and more individual responsibility for their life conditions (Pham & Pham, 2007). To our knowledge, there is no empirical evidence on the public 1 Academia Sinica, Taipei 2 Soochow University, Taipei belief about trade openness in Cambodia. We need to examine how people perceive trade openness in post-communist countries systematically.
Our study is the first on the mass public's belief in policy protections of labor and peasants with a cross-national comparative design. Focusing on policy-specific issues, we examine how the general public in the three societies sees openness for imports, and assess their potential impact on the labor and rural peasants and their preferences for import limitation. The findings advance our understanding of a sympathetic feeling toward the precariat class in the three socialist countries and provide a unique view of how people respond to the changing relationships between the state, individual, and society during a socialist market transition. The research issue is fundamental because imports play a substantial role in these economies, especially in Vietnam and Cambodia. Recent economic growth has relied heavily upon materials, equipment, and technical know-how via trade, not to mention increased consumption goods pouring into the domestic markets. The two countries currently show a higher percentage of imports over GDP than Thailand, an ideal open market economy this study uses as a benchmark for comparison. China has continued to depend heavily on foreign inputs to realize its role as a world factory. Nevertheless, its gigantic economic base has resulted in a smaller fraction of imports (see Figure 1).
In this paper, we ask what, after a long journey of market liberalization, are the levels and sources of protectionism for labor and peasants in these three authoritarian regimes, which have undergone socialist market transitions and are highly engaged in the global market? Previous research highlights a prima facie hypothesis that it is personal as well as sociotropic interests that may stimulate openness. The interestsbased argument might not fully account for our three cases. People's perception of international economic competition and the impact on labor and farmers can be conditioned substantially by political ideologies idiosyncratically evolved from a legacy of socialist collectivism. We advance two competing hypotheses in this study. For the general public, holding collectivist values and having high political trust in government and the ruling elite account for much variance in the antipathy toward openness and favoring more protection for farmers and workers in the market. Evidence will be evaluated across China, Vietnam, and Cambodia and contrasted with Thailand. Thailand has a slightly higher GDP per capita (US$5,076 for 2010) than China ($4,550), while those of Vietnam ($1,318) and Cambodia ($786) lag far behind. The data are the third and fourth waves of the Asian Barometer Survey, collected from 2010 to 2011 and 2014 to 2015. Building on this survey evidence, we expect that the findings will contribute to settling the debate about whether ideologies such as social values or political trust outweigh material interests to determine people's preferences.
The structure of this paper is as follows: we first review the literature on trade and import preferences, focusing on two particular vulnerable and significant groups, workers and farmers in post-communist countries. Next, we argue how value positions such as collectivism and political trust shape people's attitudes toward imports during a socialist market transition and propose the hypotheses for testing. Then, we explain our measurements and methods, and report the findings from analyzing the cross-national survey datasets. Lastly, we conclude by summarizing the findings and noting the limitations of our study.

Protecting the Workers and Farmers: Imports, Interests, and Politics
Scholars claim that personal interests shape policy choices. The perceived advantages from economic growth or distributive policies are considered primary factors affecting attitudes toward openness after socialist market transition in the East European societies, post-communist democracies (Csaba, 2008). Additionally, the wage level is negatively associated with protectionist preferences (Scheve & Slaughter, 2001). Job insecurity leads to less preferences for trade liberalization (Johnston, 2013;Naoi & Kume, 2011) but more support for state ownership (Runst, 2014). However, some studies provide a rebuttal that personal economic status has little explanatory power for the choice of openness policy (Mansfield & Mutz, 2009;Wolfe & Mendelsohn, 2005). People's ignorance about the impacts of market openness is arguably one of the most critical reasons some citizens endorse specific trade policies hurting their economic interests (Rho & Tomz, 2017). Indeed, trade openness creates winners and losers, yet most of the general public has difficulty knowing whether they are winners or losers (Rogowski, 1989). Researchers have also looked into how skill levels mold trade preferences. The more highly skilled workers prefer free trade in wealthy countries, while they are less likely to do so in developing countries (O'Rourke et al., 2001). While empirical studies inspired by the economic interest explanation have presented no more than mixed Source. https://data.worldbank.org results concerning the economic interest argument, an individual's gain, rather than their material (or objective), economic conditions during the economic reform may account for substantial variance in attitudes toward openness.
There are two plausible pathways toward protectionism in this regard. First, people who perceive a growing national economy can probably support more protection for the less privileged groups because they have more confidence in the future returns and therefore can afford to care more for those who become increasingly vulnerable to the market. This perception has been termed "socio-tropic evaluation" and is closely associated with confidence in governments in postcommunist societies in Eastern Europe with the socialist market transition (Kluegel & Mason, 2004). Second, prospective economic mobility, expecting a better economic condition for individuals ("egocentric evaluation"), can also lead to more substantial support for protectionism rather than more openness. Thus, we propose a revised hypothesis concerning perceived economic interest as follows: Hypothesis 1: The more an individual expects the national economy or personal economic conditions to improve, the stronger their support is for protectionism for farmers and workers.
In contrast to this revised economic interest viewpoint, this study advances a powerful argument that policy choices often are affected by a belief in specific values or a commitment to societal ideals (Funk, 2000). The mass public has specific value positions based on which they favor policies that they deem as desirable to achieve the common goods. These ideal interests or perceived best interests for the larger social groups (in our case, labor and farmers) can constitute a basis for policy preferences distinct from the calculated personal gains or losses. In what follows, we highlight such a political value explanation to account for attitudinal support for protecting labor and peasants in the socialist market transition.
Collectivism has long been a moral value in Asian societies. It refers to the belief that individuals give their unquestioned loyalty and sacrifice their personal goals to be accepted by the larger group and recognized as in-group members (Voronov & Singer, 2002). To see oneself to the whole (nation) constitutes a basic morality. An individual's identity is meaningful only when embedded within the society or state. In China and Vietnam, collectivism originated from conventional Confucian teaching, but a legitimate demand to sacrifice personal interest for the nation-state evolved into being only in the early 20th century.
Chinese nationalists equate "the nation's interest" to the people's livelihood or well-being. Sacrificing one's interest in the nation is justifiable and highly patriotic. However, excruciating it is, an individual should bear the pain for the sake of the grand pursuit of a nation's strength and prosperity. Edgerton-Tarpley (2014) traced the evolution of the sacred idea of sacrificing one's life for the country in modern China and pinpointed a striking similarity between the nationalist and Maoist states in enforcing this collective discourse for mobilizing people to sacrifice individual interests during lethal famines. In the Vietnamese context, virtues in the Confucian tradition such as righteousness, collectivism, courage, and devotion to the nation have been essential in crafting a personhood model and various reform movements (Bradley, 2004). Shohet (2013) further traced the anthropological origin of the ethics of sacrificing in the Confucian family ideology. In the Vietnamese context of ancestor worship and generational relationships, sacrifice represents more than a gift and a religious ritual of respect for gods and ancestors. It mediates the relationship between the sacred and profane modes of existence. Most relevant to this study is that it is reified into a nationalist act, a moral code for daily life justifying the hardship endured by a person for the family, community, and country. Shohet (2013) forcefully argues that a person's sacrifice shows an understanding that personal desires and needs, which are inseparable from those of others to whom he is, she is normatively related, in what is called "asymmetrical reciprocation." In this hierarchical and patriarchal relationship, people are obliged to pay back their debts for what the larger community has done for them, like country and family. In the post-conflict period, Cambodia's government has also endeavored to craft a new national identity through public education. The formal educational system aims at socializing students to possess "a sense of national and civic pride" and develop "a strong belief in being responsible for the country and its citizens" (Ministry of Education, Youth, and Sport, 2005). Whether this state-sanctioned ideology of collectivism and nationalism has been prevalent among the general public has not been documented. Tan (2008) suggests that this line of moral socialization echoes the traditional cultural preferences for social harmony, conformity, and passive acceptance of authority. Nevertheless, admittedly, little empirical evidence about collectivist values has been presented from Cambodia.
In advancing a general proposition, we reason that people concerned with the interests of groups or nations should be more attentive to the harmful impacts of imported goods on the local economy and community. Thus, the second hypothesis is derived: Hypothesis 2: Those who hold a collectivist value are more likely to favor protectionism for farmers and workers.
According to the initial hypothesis, political trust in government denotes people's confidence in looking after their best interests (Hetherington, 1998). High political trust implies confidence not merely in the integrity of public officers but also in the ability and efficiency of the government institutions. Political trust is one basic foundation of legitimacy and policy support (Levi & Stoker, 2000). We contend that when individuals believe the government to be trustworthy, they are likely to expect the state to adopt policies for the sake of the weak class. People expect the powerful state to look after their interests in the paternalist culture. In contrast, people place political trust in a government that helps facilitate individual liberty and autonomy in a market economy, with minimum state intervention in the Western rational model of the statesociety relationship (Sigley, 2006).
In the context of China, Vietnam, and Cambodia, the paternal and governmental relationship has embraced an ethos of official duty to care for the public well-being, in the sense that a benign ruler and the officials should be highly concerned with what the people need, even if authoritarian states restrict fundamental rights and exploit societal resources at hand (Cherry, 2016;Sigley, 2006). Socialist market transition and trade openness policies in socialist China have generated a state's concomitant reengineering rather than retreat. Economic liberalization has not been an end in itself. It has been an instrument auxiliary to the ultimate socialist order (Yan, 2001). As Crane (1998) argues forcefully for the case of China, this instrumentalist view means that the development of the national economy serves the political interests rather than the other way around. Indeed, the states have increased their political grip on production, exchange, and consumption activities even when the socialist countries have become more open to the global economy. While both countries have made great efforts to respond to people's demands effectively, a paternalist ideology soothes out conflicting social interests induced by economic growth. The state's paternalist ideology prescribes power relations in an exchange of absolute authority with fatherly benevolence. There exist some cross-country variations, however. As a study of Chinese labor unions reveals (Zhang, 2009), under this paternalist spell, the union leaders have minimal effectiveness, while the government plays the role of a protector for the workers' legal rights and welfare during the socialist market transition (Lan et al., 2015). In comparison, however, the state in Vietnam has been described as less a protector than a regime using deterrence and fear more regularly (Cherry, 2016;Fforde & Homutova, 2017). There is a need for research to check whether this difference impacts the evolution of protectionism.
Based on the paternalist ideology argument, we hypothesize that the belief in government constitutes one critical factor supporting a socialist order protective of the people's interests for the three authoritarian regimes. Thus, political trust is positively related to the protectionist mindset.
Hypothesis 3: Those with higher political trust in government are more likely to favor protectionism for farmers and workers.

Data
We use the dataset of the Waves 3 (ABS W3) and 4 (ABS W4) of the Asian Barometer Survey (ABS; Hu Fu Center for East Asia Democratic Studies at National Taiwan University, 2014), which collected unique information about attitudes toward imports for China, Vietnam, Cambodia, and Thailand. By comparing these three authoritarian regimes with socialist market transitions to Thailand's mature market economy, we adopt a research strategy of a worst-case scenario. We note that Thailand's current dynamics of growth have been driven by openness in trade and foreign investment (Chachavalpongpun & Samphantharak, 2020;Yusoff & Nuh, 2015). Economic openness seems to be well-received among Thai people nowadays, although this country was hit hard by the Asian financial crisis more than two decades ago. We test the robustness of our theory for political values and attitudes toward imports and social protection in a socialist market transition.
The Thailand, respectively. Based on a face-to-face interview method, it applies the probability sampling techniques by a multistage area approach for the randomized sampling and the adequate coverage of rural areas and minority populations. Because this study is interested in people's attitudes whose socio-economic situations have potential influences, respondents' ages range from 18 to 65 (see Table 1 for summary statistics).

Variables
Dependent variables. This study utilized two measures to capture the public response to openness and protection of the less privileged classes as the dependent variables. The first question is: "Foreign goods are hurting the local community" (on a 4-point scale from "strongly agree" to "strongly disagree"). Note that this question does not specify which groups hurt in the local community are hurt. It could refer to either the ways of life or economic activities in a locale. To strengthen measurement, we adopted another measure, "We should protect our farmers and workers by limiting the importation of foreign goods." Ideally, the interests and well-being of farmers and workers are at the core of a socialist social order. Respondents who agree with this statement reveal an inward-looking disposition to protect both sub-populations from the potential harms of imports. These two measures tap an attitude of economic nationalism because they highlight the impacts of competition with foreign goods. Wave 4 of ABS does not include the first question, so we used only data from ABS 3. We pooled the responses to the second question in analysis, but the different levels observed are controlled using a dummy of waves. Figure 2a examines respondents' evaluation of whether foreign goods are hurting the local community. In China, the percentage of people who agree (47.2%) is slightly less than those who disagree (52.8%). In contrast, Vietnamese and Cambodians share the Thai's protectionist sentiments.
Vast majorities reported that imports harmed local society in Vietnam (72.2%), Cambodia (86.0%), and Thailand (89.5%). Moreover, most people "strongly disagree" that foreign goods hurt the local community. Compared with Vietnam, the protectionist sentiments are intense in Cambodia and Thailand. The Thais especially are alert to threats from imports. Thailand is the only market economy where the communist party never rules, and its level of imports is not exceptionally high. It seems that the sheer volume of imports does not include covariate highly with this attitude. Nevertheless, almost 90% of Thai support protectionism, a grave concern not observed among the three post-communist countries implementing the socialist market transition.
To further probe into the protectionist sentiments, we display the level of the perception of the potentially detrimental effect of imports on farmers and workers across studied countries. Figure 2b shows that the Chinese are distinctively less worried about the harm of imports caused to the farmers and workers than people in Vietnam, Cambodia, and Thailand. In China, slightly more than half the respondents disagree with the statement (8% strongly disagree and 43.9% somewhat disagree). On the other hand, in Vietnam, Cambodia, and Thailand, a great majority of respondents support protectionism, as we observe remarkably high figures (80.3%, 89.6%, and 88.8%, respectively), showing great sympathy for farmers and workers. Here, Vietnam, having a highest dependence on imports, does not express a largest concern with the economic conditions of the farmers and workers.
A substantial proportion of people in socialist China seem to subscribe to the idea of openness to global imports at the expense of the local less-privileged classes. However, this is not the pattern for Vietnam and Cambodia. For the latecomers to socialist market transition, a drastic restructuring of the relationship between the market, state, and the individual toward less regulation remains a concern. Nor is it entirely welcome in Thailand, which has long been experiencing an open market economy. The obtained finding also casts doubt on a previous interpretation of the socialist market transition for China and Vietnam. Scholars have suggested that because Vietnam's economic growth is more dependent on regional and global markets than China's, it should be more open to the world market. Turley and Womack (1998), one of the very early studies in comparing ideologies underlying the socialist market transition in socialist Asia, highlights such a contrast: "China opened its door and expected the world to walk in. Vietnam opened its door and expected to go out and find the world, adjusting to global taste" (p. 118). Vietnam is presumed to be more agile and open than China in pursuing global opportunities. This evaluation, however, does not hold well against our findings. Vietnam is loather to be open to imports. Higher dependency on the global market does not necessarily lead to a stronger desire for openness. Cambodia is late to adopt a socialist market transition and less open to the global market among the three post-communist countries. Compared with the other three countries, Thailand stands in contrast with a thriving market economy.
Independent variables. We use three key indicators of an individual's economic conditions to understand their distinct influences. The first one is the household's income level, as was reported by the respondent. All respondents were asked to place themselves on a quintile scale, which was designed based on official statistics about household income differences. In addition to this objective income indicator, subjective economic condition is also measured by two proxies. The first is the expected economic condition of the family, by assessing its economic condition a few years later (from "much better" to "much worse" on a 5-point scale). The second one refers to the country's expected economic condition, measured similarly (on a 5-point scale; see Appendix Table A1 for summary statistics).
We used two indicators to capture the orientation of collectivism. As we have argued, the idea of sacrificing is one primary element for conceptualizing collectivism in the cultural contexts of China, Vietnam, and Cambodia. Therefore, the two questions are selected primarily for their measurement: "In a group, we should sacrifice our interest for the sake of the group's collective interest" and "for the sake of national interest, individual interest should be sacrificed" (from "strongly agree," coded 4 to "strongly disagree," coded 1 on a 4-point scale). A positive response shows higher collectivism because the nation and group precede personal interest in the questions. We calculate the average of the two items for a collectivism indicator.
Political trust indicates confidence in the government and public authorities. We use two measures, the reported level of trust in both the national government (from "a great deal" to "none at all" on a 4-point scale) and response to "You can generally trust the people who run our government to do what is right" (from "strongly agree" to "strongly disagree" on a 4-point scale). The first refers to confidence in the institutions of the central government, which conceptually differs from the other indicator, focusing on high-ranking government officials in charge of operating the government. Because the responses designated for the two questions are categorically different, we treat each as a distinct predictor in the analysis.

Controls.
We consider several control variables in the estimation. Scholars have demonstrated that support for openness policy varies based on gender, age, endowment factors, and skill level (Mayda & Rodrik, 2005;Scheve & Slaughter, 2001). The more highly skilled workers tend to be cosmopolitan in values, and they are more secure in the market and more likely to enjoy foreign goods for consumption (Tsai & Iwai, 2013). We use education years as an indicator of skill. Education is negatively associated with protectionism.
Additionally, we regroup employment status.
(1) upper while collar, as the reference group, (2) lower white-collar, (3) manual worker, (4) peasants, and (5) all others (unemployed, retired, students, and house workers). Those who reported having no job were too small percentage in Vietnam, Cambodia, and Thailand that they were not separated as a dummy. In China, the occupational information provided in the AB surveys lack information about the skill levels for people working in the labor market. Their employment positions are categorized as: (1) employed, as the reference group, (2) unemployed, (3) peasants, (4) all others, in addition to a group whose employment status is (5) missing. The employed are less likely to favor protectionism because their job security is high (Johnston, 2013;Naoi & Kume, 2011).
Urbanites are also crucial for respondents to disfavor trade restrictions. City residents tend to have more opportunities to benefit from contacts with the outside world, thus being less supportive of restrictive trade policy. Finally, we explore the effect of global exposure by measuring a respondent's frequency of "following major events in foreign countries and the world" (5-point scale, from "very closely" =5 to "not at all" =1). Kaltenthaler et al. (2004) find that citizens who are less "cosmopolitan" are more prone to endorse regulation on international trade. We assume that global exposure is positively associated with trade openness (Tsai & Iwai, 2013).

Model Specification
The technique of ordered logistic regression is used for estimation. This model assumes that there is an unobserved continuous dependent variable Y*, which is a linear function of observed independent variables (Treiman, 2009, pp. 342, 343): The residuals are distributed in the logistic function. In fact, what is observed is a set of ordered categories, Y = 1. . .K; in the current case, Y = 1, 2, 3, and 4 (our dependent variables range from 1 to 4). And where the k i is cut-off points on the latent variable. The probability that an outcome y = k is The expected probability is that an observation (in our case, preferences for limiting imports) will have a value of k. As the above function shows, it is the difference between the probability of reaching the upper bound cut-off point and getting the lower-bound cut-off point (Treiman 2009, p. 343). A positive sign indicates an increase in the contribution to the odds ratios of a specific category over the lower categories. We report the odds ratios by exp (−b k ) for a list of independent variables x j , which include collectivism and political trusts, as well as other demographic controls.

Results
We now turn to the estimation results or two responses to imports. Table 2 demonstrates the outcomes from the ordered logistic models for evaluating imports' impact on the local community. The analysis used approximately 11% fewer respondents for various reasons, including "do not understand the question," "decline to answer," or "don't know." It is not proper to assume that missing information is missing at random. Caution should be exercised in interpreting the results. The effects of family income and two perceived economic conditions are insignificant, except in Vietnam, where the middle-income group reports being 1.87 (i.e., 2.865 − 1) times more alert than those with low family income to the potential impact of imports, which is significant at the level of .05. The empirical evidence generally does not support H1. Greater expectation by an individual that the national economy or personal economic conditions will improve does not augment or abate their concerns for imports' harm to the local community.
As is expected, collectivism shows a positive influence in China, Cambodia, and Thailand. The odds ratios indicate that in these three countries, people are more likely to be alert to the potential harm caused by imports by roughly 40%, 27%, and 34%, respectively (at the significance level of .01), given a one-unit increase in collectivism. Trust in government fails to achieve a statistically significant result, except for Vietnam, where it is just what we expected-high trust in government leads to more concern over harm from imports. Restated, the Vietnamese are about 89%more likely to see imports negatively, given a one-unit increase in trust in government. China is the only country where a one-unit increase in trust in the elite results in greater likelihood of distrusting imports, which is a significant 41.4% (p < .001). However, in Thailand, a negative correlation of trust in the government with limiting imports (its odd ratios is smaller than 1) might reflect the tendency of the supporters (surveyed during the second half of 2010) for the government then led by the military and its alliances, instead of those favoring the populist pro-Takshin parties which promised redistributive social policies (Hewison, 2017). Overall, evidence for H3 is less substantial.
Regarding control factors, we obtained several interesting findings worth further discussion. Males in China, as in Thailand, are less inclined to support import limitations. Age effect is assertive in three out of the four studied populations: older people are more alert to whether increased imports hurt the community. Education appears to have mixed influences because it increases attention in Cambodia and Thailand but has a detrimental effect in China and Vietnam. That is, in the latter, respondents who are more highly educated endorse openness (see also Blonigen, 2011;Rho & Tomz, 2017). Employment status has little effect on the equation. The peasants are a focal group in the model, but its positive influence is shown only in Vietnam. Urban residents in Vietnam are more attentive to imports' impacts. More global exposure leads to fewer worries about foreign goods hurting the local community in China. However, in Vietnam, it boosts protectionism (Viet Nam News, 2011). The context of global exposure might differ across the two countries. Particularly in Vietnam, knowledge about the origins of imports can be decisive for whether these goods are welcome or not. There has been resistance among Vietnamese to the importation of Chinese goods. This is perhaps the reason why global exposure does not lead more Vietnamese to welcome foreign goods with open arms.
The estimation results for protecting farmers and workers are displayed in Table 3. The dummy variable of the Wave 4 in the equations shows changes over two periods in the Note. Standard errors in parentheses. *p < .05. **p < .01. ***p < .001.
pooled cross-sectional data. We find that only in China does it indicates an increase (at 33%) of "limiting imports to protect farmers and workers" compared with the dataset in Wave 3. In Thailand, the trend shows a decrease over time.
The Chinese respondents in the top socio-economic stratum do not express more concern about the impact of imports than those in lower social strata. Instead, Vietnamese in the middle-income category seem to be more supportive of farmers and workers by the margins of 95% (significant at the .001 level), compared with the low-income group. There is a similar pattern observed among Thai, although the gap is smaller (37%). Higher hopes held by an individual for improvement in either the national economy or personal economic conditions are not related to more support for protectionism to help farmers and workers in a consistent manner. Thus, evidence supporting H1 is mixed.
In contrast, collectivism remains a powerful predictor in three socialist countries, China, Vietnam, and Cambodia, as well as in one with an open market economy, Thailand, where a one-unit increase in collectivism increases a greater likelihood (ranging from 21% to 67%) of wanting to protect farmers and workers. These significant results support H2.
Trust in government in Vietnam leads to strong attitudinal support for farmers and workers by a margin of 31% on the condition of a one-unit increase of such trust. On the other hand, a one-unit change in trust in the elite systematically generates significant effects among Chinese, Vietnamese, and Thai by 35%, 35%, and 28%, respectively. Note that in Cambodia, a one-unit increase leads to a drop of approximately 12% in the attitude favoring protecting farmers and workers. Cambodians who support the national leaders show a neoliberal tendency to somewhat discount the interests of farmers and workers. Thus, we have mixed evidence for H3.
The control variables generate similar effects on the second dependent variable when they reach significance. Age produces a positive and significant impact on two populations. In both China and Vietnam, males and the highly educated feel less need to protect farmers and workers at the expense of imports Occupational status does not covariate highly with protectionism. Note that for China, the category of the peasants, a term this study uses in an indiscriminate manner, is available for only the wave 3. It carries a significant positive regression sign when this wave is estimated solely, suggesting that these peasants, who probably are migrants working outside their hometowns, tend to have a greater concern with those at the lower stratum on the social ladder. In Thailand, the urbanites take a position in favor of fewer restrictions on imports. Global exposure in China predicts lower support for protecting farmers and workers, while in Cambodia, global exposure in fact brings higher support. This difference alerts us to the diverse contexts in which people with a cosmopolitan orientation react to imports and consequently adopt specific policy positions.
In sum, as an open market economy, Thailand serves as a reference to the other three, socialist countries. Presumably, the Thai would show more support for imports. Nevertheless, the results indicate the opposite: as far as the level of attitude toward imports is concerned, Thai people are more restrictive than those in the other three countries. In hypothesis testing, collectivism (H2) and trust in the elite (H3), outweigh material interests (H1) in restrictivism both in the three socialist countries and in Thailand, an open market economy. In other words, the shared social values and political trust prevail in the mass publics' trade policy preference. The political values and detachment from what is called marketfriendly policies in socialist authoritarian regimes in Asia has important implications which we will further discuss in the concluding section.

Robustness Check
We perform alternative measures and estimations to ensure that the findings are reliable. First, among Vietnamese, family income distribution is skewed, so that it might generate biased estimates. We replace it by combining the possession of two household facilities-cable television and a refrigerator-and use it as a proxy of family income. We re-estimated the equations and obtained the same outcomes, except collectivism was significant.
Second, a respondent's affiliation with political organizations might influence attitudinal response to imports. Unfortunately, the ABS does not provide political party membership with public sectors variables. The Vietnam data offer a unique factor of feeling close to the Communist Party. However, we found a weak correlation with import restrictivism. It was dropped for further analysis.
Third, the ordered logistic estimation in Tables 2 and 3 assumes that the coefficient of a specific predictor is equivalent across all logistic regressions of various categories. This parallel-lines assumption does not hold firmly in our case, in that we conducted estimation by the partial proportional odds models (Fullerton & Xu, 2012). Alternative models in the family of generalized ordered logit regression researchers give up parsimony for detailing the various magnitudes of the effects across categories, which can be increasingly complicated for our case. It becomes hard to explain theoretically how these variations occur. For instance, according to the Gamma index of deviation from proportionality (Williams, 2006), in China, the coefficient of collectivism for liming imports the difference between "agree" and "strongly disagree" is about 1.5 times stronger (p < .001). Deviations as such might be interesting when we seek apparent differences across equations in the models. Nevertheless, for an overview of attitudes toward imports, our results suffice.
Finally, missing values occurred, especially in the collectivism and political trust variables, ranging from 5.4% in Cambodia to 13.2% in Vietnam. We generated an imputed dataset and re-estimated the models to determine whether substantial differences exist compared to the results reported in Tables 2 and 3. We chose the multiple imputations by chained equations (MICE) method. MICE techniques can handle different variable types (in our case, ordered categorical dependent variables) because each variable is imputed using its imputation model (White et al., 2011). Re-estimation based on ten imputation cycles shows similar outcomes for collectivism and political trust for three socialist populations. Among Thai, the influence of trust in government on "hurting local community" is insignificance. All in all, our main findings for testing the three proposed hypotheses remain stable despite missing information in the studied survey data.

Discussion and Conclusion
Scholars of socialist market economies have been understudying how the general public perceives a need for restricting imports (Baughn & Yaprak, 1996). Based on recent survey data from three socialist countries-China, Vietnam, and Cambodia-and liberal capitalist Thailand, we have tested revised economic interest hypotheses and two new advanced political ideology and political trust hypotheses in cross-national comparative settings. The results show that collectivism and political trust, rather than economic interests, are crucial determinants preference for imports. Both factors induce opposition against them, except that trust in elite operates differently in Cambodia.
Our findings advance the knowledge about protectionist preferences in East Asia's authoritarian regimes which have been undergoing socialist market transition in the following regards. First of all, an individual's economic resources measured by income, compared to their subjective economic conditions, are more strongly correlated with protecting the workers and farmers. However, its influence is less observable in the attitude about limiting the imports. The special interests argument might expect that people with higher income benefit more from open policies and therefore welcome imports and give less attention to potential impacts on the less privileged classes relying on local economies. This seems to be the case of China, but is not applicable to the other studies cases. The perceived economic condition, as is indicated by one's family or the national economy, does not prove to be a solid explanatory factor, contrary to our expectations. An optimistic economic outlook for the family and the nation does not instigate acceptance of imports consistently across the studied countries. A possible explanation of this insigficance (H1) is that the even for a specific income or status group, the members range widely regarding their perception of potential benefits from either openness or strict limitation of imports.
This article contributes mainly to the political psychology literature on ideology, regime trust, and antipathy toward imports by highlighting the importance of a collectivist ideology and trust in government by communist parties. These political factors help explain why in Asia's socialist societies many people do not favor imports even when they have vehemently pursued socialist market transition and openness policies. We do not intend to invalidate the economic interest argument entirely. We propose a comprehensive framework beyond income-related considerations to fully comprehend public beliefs toward imports and their perceived impacts on local societies.
We argue that the collectivist value observed in socialist China, Vietnam, and Cambodia has a distinct implication (referring to H2), which is also applied to Thailand, a capitalist market society. This might reveal a common belief in the Asian region that the lower class should be well protected by the paternalist government. What is more important in such a state model is to strengthen the protective role of the state, rather than to advocate the space of civil activism or bargaining power of the farmers or workers (Arnold, 2017;Hirsch, 2020). Political trust, however, is somewhat complicated in its pattern of influence. The evolution of trust in the studied countries has been facilitated by doling out public spending to compensate the less privileged people whose living conditions had deteriorated owing to trade liberalization and high exposure to external uncertainties and shocks. Numerous studies have confirmed this "compensation hypothesis" (Adserà & Boix, 2002;Hays et al., 2005;Rodrik, 1998). This perspective identifies the potential losers from globalization. For China, Vietnam, and even Thailand, increased trust in the elite (rather than in government) leads to more preferences for barriers to imports and protection for the less privileged classes (in favor of H3). Again, there is a paternalist implication herein. The elite in these Asian societies assumes, implicitly or explicitly, a role of paternalistic-technocratic bureaucrats that can understand the misfortunes of the ordinary people and take care of their well-being. The ruling class crafts extensive bottom-up planning and implements the rigid top-down policy with ultimate power and wisdom for the common good. This growth model is well-aligned in socialist market transition, political trust, and individualization.
In contrast, for the case of Cambodia, trust in the elite seems to tell a different story which needs further explanation, because it lowers the awareness of protecting workers and farmers, a significant deviation from a socialist legitimating ideology but conforming more squarely with the logic of the neoliberal zed market (a finding partially disapproving H3). The ruling class, composed of a military-commercial alliance, has been using a patronage system which redistributes societal resources by way of asset-stripping of natural resources, as the industrial base has been much weaker than in China or Vietnam (Hughes, 2020). Cambodia is also well known for its extremely repressive measures against the opposition from civil society (Arnold, 2017). It is in this context that those supporting the elite are less concerned with following a socialist legacy and with prioritizing the interests of workers and farmers. Thus, what has been responsible for a relationship between the elite and people is intriguing and highly variant within the socialist countries in Asia. Unfortunately, we do not have enough space to explore it further (Benedikter & Nguyen, 2018). In sum, we have indicated that a political ideology model operates influentially in formulating the antipathy toward imports among Chinese, Vietnamese, and Cambodians. The findings should bring our attention to the intricate, nuanced relationships between the state, market regulation, and reactionary responses to imports.
There are some limitations to this study. First, the response to imports reflects market regulation only partially. We do not capture other essential elements in the measures. Discrimination in favor of national products and voluntary constraints in buying foreign goods often incur a wide variety of public practices in a nationalist campaign. The phenomenon of "buying Chinese" or "buying Vietnamese," as has often been seen, is one indicator worth consideration in expanding measurement with regards to patriotic emotions in goods consumption (Hooper, 2000). To our knowledge, a campaign for commercial nationalism in Cambodia has not yet emerged.
Second, protectionism does not necessarily preclude trade in particular goods or with specific countries. These trade regulations reflect macro-level policies by the state. They do not necessarily affect micro-level preferences for specific foreign products among the general public. Moreover, macro-policies about tariffs, quotas, and duties can be difficult questions for the general public. Not all respondents can answer whether the less privileged would benefit from the restriction of imports-indeed, we observed quite a proportion of the sample who did not provide answers on two dependent measures. However, it might still be helpful to have a design that explores the potential impacts of industryspecific imports, such as agricultural produce or home electronic appliances, for further fine-grained evaluation.
Third, our conceptualization has relied heavily on the experience of China, Vietnam, and Cambodia. It can be generalized to other post-communist countries, like Laos and East European countries, which allows more in-depth theorizing of how political ideology influences preferences for trade policies and relevant market regulations.
Lastly, the survey data unfortunately are not fully up to date. The latest waves of ABS do not contain the questions regarding the respondents' occupational backgrounds or attitudes toward industry-specific imports. Workers in the importing sectors may see foreign goods and trade policy differently than those in exporting sectors. Future research toward this direction would make an excellent addition to the literature.