IMMIGRANTS AS VOTERS IN ELECTORAL AUTOCRACIES: THE CASE OF MAINLAND CHINESE IMMIGRANTS IN HONG KONG

Abstract

Migration to electoral autocracies has become increasingly common. Extant studies, however, accord little attention to the immigrants' influences on the domestic politics of these regimes. We argue that immigrants have attributes (status quo bias and lack of prior exposure to local politics) that make them an attractive co-optation target of the authoritarian regime. We provide a case study of mainland Chinese immigrants in Hong Kong to illustrate our argument. Since the sovereignty transfer, the Hong Kong government has devised various schemes to attract these immigrants, while pro-establishment political parties and groups have actively sought to co-opt them. Using two distinct public opinion surveys, we also find that immigrants are more likely to approve of the political and economic status quo, and less likely to vote for pro-democracy opposition parties than the natives. In addition, we find no evidence that exposure to political information can change the immigrants' vote choice.

INTRODUCTION

History is replete with examples of people fleeing autocracies in search of a better life. ¹ In recent decades, as a fair number of autocracies have achieved significant economic development, more people voluntarily migrate into authoritarian regimes to improve their lot. Little is known about the political attitudes of these immigrants and their political impacts on the receiving autocracies. Are they pro-authoritarianism? Do they support democracy? Are they less politically informed and less likely to participate in politics? These are important questions that demand more scholarly attention.

This article is intended to fill this void. Our central argument is that immigrants have characteristics that make them an attractive co-optation target compared with the natives from the perspective of an authoritarian government. First, they are more likely to be satisfied with the political status quo because the decision to migrate is inherently a self-selection; had they strongly disapproved of the political status quo in the receiving country, they would not have chosen to migrate in the first place. Another characteristic is that prior to their arrival, immigrants share no common experience with the natives, including the experience of participating in any pro-democracy movement. As a result, they are less sympathetic toward the cause of the opposition.

These attributes are particularly relevant to authoritarian regimes that hold somewhat competitive elections. A key challenge for these “electoral autocracies” is to garner enough support in elections to stay in power with some degree of legitimacy. If an authoritarian government needs to invest resources in consolidating political support in society, it should devise tactics to bring immigrants on its side, as it is more cost-effective to co-opt immigrants than natives. The political attitudes of immigrants, however, are not immutable. Greater exposure to local political information may alter their political identification, although the size of the exposure effect is ex ante unclear.

We provide a case study of post-1997 Hong Kong to illustrate our argument. While Hong Kong is not an independent state after the sovereignty transfer in 1997, it should be classified as an electoral autocracy, not only because it enjoys a high degree of autonomy, including the preservation of its own currency, judicial system, legislature, and constitution, but also because it allows for multi-party elections that take place on a regular basis. Some may say that these elections lack a level playing field, which is a common characteristic of electoral autocracies. Yet, blatant electoral fraud such as vote-buying has been largely absent, and opposition parties have managed to win a sizable number of seats in every election. In this regard, Hong Kong is no different from a typical electoral authoritarian regime that allows for the existence of multiparty competition, while struggling to keep the opposition in check (Fong 2017).

Our analysis is also relevant because immigrants make up a significant portion of the city's population. In particular, about one million (or one seventh of the population) immigrants from mainland China settled in Hong Kong after 1997. The large migrant stock provides rich empirical data for studying immigrants' political attitudes. Finally, Hong Kong citizens are still able to gain access to alternative political information. Hong Kong does not have China's massive Internet censorship or the great firewall. It therefore provides an important opportunity to analyze the extent to which exposure to political information under a relatively free media environment can alter immigrants' political identification.

Unraveling the intention of the ruling elite of an authoritarian regime is a challenging task because its decision-making process is generally opaque. Yet, it is still possible to infer its intention from observable behaviors. In the present context, if our argument is correct, we should observe a relatively lax immigration policy in Hong Kong and the ruling elite's active pursuit of the immigrants' political support. Indeed, our case study provides qualitative evidence in support of these behavioral implications. We also examine the “supply side” of the story; namely, the immigrants' status quo/pro-government bias. Drawing on the data from two distinct public opinion surveys, the Asian Barometer Survey (ABS) Wave 3 and Wave 4 and the Hong Kong Election Study (HKES), we find

1. (1) mainland Chinese immigrants in Hong Kong are more satisfied with the political and economic status quo than natives;

2. (2) they are less likely to support pro-democracy opposition parties than natives;

3. (3) they consume political news more often than the natives; and

4. (4) exposure to political information is not correlated with immigrants' electoral support for the opposition parties, but is correlated with their political identification with them.

Our findings have profound implications for Hong Kong's political struggle for full democracy. While the opposition continues to obtain support from native Hong Kong people, it is fighting against a more consolidated pro-government coalition who see their support base increased daily as a result of a continuous influx of immigrants from mainland China. Perhaps the only hopeful sign for members of the opposition elite is that information exposure after arrival can somewhat increase immigrants' political identification with the opposition.

The rest of our article is divided into five sections. We begin with a discussion of the theoretical motivation. We next present our argument and then provide the historical background of our case. This is followed by empirical results and finally conclusions.

THEORETICAL MOTIVATION

There is a large literature on the political participation of immigrants in host countries. A common finding is that immigrants tend to have lower political participation than natives (Ramakrishnan and Espenshade 2001; Cho 1999; Junn 1999). The degree of political participation can vary with the personality characteristic and external environment of immigrants. For example, Leal (2002) show that the political participation of Latino immigrants in the United States increased when they become more informed about politics. Black, Niemi, and Powell (1987) showed that the older Canadian immigrants tend to be more politically active than the younger ones. Finifter and Finifter (1989) studied American migrants in Australia, and found that immigrants who are less ideologically committed to a US party tend to relinquish former party identification more easily.

An obstacle to political participation is that migrants invariably live in their own ethnic communities, and have less integration with the host society (Bauer, Epstein, and Gang 2002). Some political environments may be more receptive than others and allow better engagement for the immigrants. Just and Anderson (2013) showed, by data from the European Social Survey, that a friendly opinion climate toward immigrants will promote institutionalized political action. Using the same survey, they show that citizenship is a powerful determinant of immigrants' political participation (Just and Anderson 2012). Other scholars find that in some cases a threatening environment in the host country can stimulate immigrant's political participation, including voter turnout (Ramakrishnan 2005; Pantoja, Ramirez, and Segura 2001) and voter registration (Cho, Gimpel, and Wu 2006).

Recent studies of immigrants' attitudes focus on the political effects of migration on the sending countries. By analyzing election data from Mexico, Pfutze (2012) shows that migration promotes democratization by increasing the electoral support for opposition parties. Careja and Emmenegger (2012) showed that in Central and Eastern Europe, people who had migration experience to the West tend to have more positive attitude towards democracy.

In the age of globalization, looking for better job opportunities is the most common motive for migration (Borjas 1999). The improvement of material well-being can make migrants more resourceful and politically active. The improvement of status can also bring about expression of more liberal views and other political aspirations (Mishler and Rose 2001; Careja and Emmenegger 2012). The urge to participate, in theory, should root in the basic political attitudes of the immigrants. The political ideology of the migrants in turn should be closely connected with their political exposure prior to migration. Those who came from an autocratic regime have less exposure to democratic norms. As a result, some argue that immigrants from non-democracies are less capable of participating in politics even if they move to democratic countries (White et al. 2008; Black, Niemi, and Powell 1987). Others contend that migrants socialized in a politically repressive environment tend to distrust politicians and government officials (Portes and Rumbaut 2006; Bueker 2005; Ramakrishnan 2005; Fennema and Tillie 1999), which may hinder their political integration in the host society. In contrast, DeSipio (1996) and De la Garza, Falcon, and Garcia (1996) counter that these migrants would more likely participate in politics precisely because they lacked such opportunities prior to migration.

Most of the studies of immigrants' political attitudes and participation, however, are based on data in democratic countries. One reason for this is that authoritarian regimes offer little opportunity for political participation by their citizens, let alone immigrants. Recent decades have seen a change, however. First, more and more authoritarian regimes allow for multiparty and somewhat competitive elections (Levitsky and Way 2010; Geddes 2005). Second, as a fair number of autocracies have achieved significant economic development in recent decades, more people voluntarily migrate into authoritarian regimes in search of a better life. Given that many electoral authoritarianism or hybrid regimes today allow citizen to participate in regular elections, the electoral impact of immigrants should not be overlooked.

Malaysia provides one example of how an authoritarian regime is able to make use of immigrants to boost its own electoral support. In 2013, Malaysia held its general elections, in which the ruling coalition, Barisan Nasional (BN), obtained its worst electoral performance in history, although it was still able to secure a majority in the parliament. There were reports of numerous electoral malpractices committed by the ruling parties. One of them was to mobilize tens of thousands of foreign workers, who came from less developed countries such as Bangladesh, Indonesia, and Myanmar, to vote with faked identity cards (Cochran 2013). Sadiq (2005) also argues that the influx of immigrants from the Philippines has strengthened BN's electoral support. While BN's practice might be seen as electoral fraud, there are governments that adopt an “express citizenship” policy to immigrants. For example, it can take as few as two years for foreign workers to get citizenship and voting rights in Singapore. ²

Figure 1 shows that many authoritarian regimes, defined as countries with a score of less than zero on a 21-point scale developed by the Polity IV project, actually have a fairly sizable population of foreign-born individuals. Although Hong Kong is not shown, its figure is comparable to Singapore, which is one the highest. As of 2011, 39.5 percent of the citizens of Hong Kong were not born in the city (HKSAR Census and Statistics Department 2012).

Figure 1 International Migrant Stock as a Share of Total Population by Country, 2010

OUR ARGUMENT

Why would many non-democratic regimes be willing to incorporate the immigrants into the voter population? We argue that incorporating outsiders into the electorate is consistent with the political interest of electoral autocracies. Only those who find the political status quo of the host country acceptable would self-select to migrate (the selection effect). It is therefore not surprising that the immigrants would favor the ruling elite over the pro-democracy opposition that seeks to change the fundamental political order.

This is not to say that political considerations are the primary motivation behind one's immigration decision or the key reason for a host country's acceptance of immigrants. Most people migrate for economic reasons. Precisely because of their individual economic concerns, they would accord relatively less attention to other broader social issues in the host country such as its political development. Similarly, the prime objective behind an authoritarian government's immigration policy may not be political, but this does not imply that the authoritarian government would refrain from taking advantage of the immigrants for other political goals.

Immigrants should also exhibit greater tolerance toward the political status quo for two possible reasons. First, immigrants tend to come from less affluent regions or countries. They are often able to gain a better living standard after migration, which may further increase their support for the authoritarian incumbent.

Second, compared with natives, immigrants often lack political knowledge about the past performance of the ruling elite. As a result, their dissatisfaction with the regime should be lower, which also implies that they are less likely to support the opposition. In addition, prior to their arrival, they had no exposure to any local pro-democracy movement. Consequently, they are unlikely to have familiarity with the cause of the opposition. Note, however, their exposure to local political information after migration may improve their local political knowledge, and possibly the attitude toward the opposition. Whether this exposure effect can narrow the gap between natives and immigrants in political preferences is ex ante unclear.

In summary, ideological predisposition and the lack of prior exposure to local political information are important attributes that make immigrants likely supporters for the authoritarian incumbent. While it is true that the selection effect associated with movement to a non-democratic country might be over-ridden by socialization into democratic norms, the extent to which this actually occurs is an empirical issue.

MAINLAND CHINESE IMMIGRANTS IN HONG KONG: BACKGROUND

Postwar Hong Kong has been seen as a “refugee society” (Hughes 1968; Lau 1984). Since 1949, hundreds of thousands fled the economic hardship and political turmoil in mainland China to come to Hong Kong. The coping strategy of Hong Kong government before 1980 was labeled as “touch base policy” Chinese and British soldiers on both sides of the border would try to stop the illegal immigrants, sending them back to China after arrest. If the illegal immigrants managed to get past the border and reach the city center, they would be given identity cards and allowed to stay, work and live in Hong Kong as legal residents.

Before the 1980s Hong Kong had no democracy, but personal freedom was largely respected, and in general had much better living standards than mainland China. For decades, the dominant paradigm in Hong Kong politics saw the immigrant nature of the society as the key explanation for political stability amidst rapid industrialization and urbanization. These immigrants from China were seen as materialistic, caring mostly about their families and making a living than about political ideologies, which brought a low level of political participation despite lack of democracy and welfare benefits before the 1980s (Lau 1984).

The “touch-base” policy was scrapped in 1980. From then on even if illegal immigrants managed to get to the city center, they would not be allowed to stay but would be sent back to China. This largely stopped the regular influx of immigrants from the north after 1980. The Hong-Kong-born took up a larger and larger portion of the Hong Kong population, giving rise to a Hong Kong identity quite distinct from that of the mainland.

Since decolonization and democratization started in the 1980s, Hong Kong has seen the rise of a domestic democracy movement. Since limited popular elections were introduced in 1991, ³ the pro-democracy parties steadily obtained 55 to 64 percent of the popular vote share. The 1989 Tiananmen crackdown, the brutality of which shocked many Hong Kong residents, was a major event that drove many Hong Kong people to become adamant supporters of democracy, and to have a more negative view of the Beijing government. Seeing the pro-democracy parties as anti-China, the Chinese government was reluctant to grant full democracy to Hong Kong, lest these anti-China elements should take control of the Special Administrative Region after 1997.

To keep its free-wheeling capitalism attractive to international investors, Beijing instituted a hybrid regime in Hong Kong after 1997. Constitutionally the Basic Law, or mini-constitution of Hong Kong after 1997, promised full election of the Chief Executive and the whole legislature, but did not specify when this would be delivered. By 2012, 15 years on from 1997, Beijing had allowed no more than half of the legislature to be elected by universal suffrage. The other half were elected from functional constituencies with a very narrow franchise, largely controlled by conservative and pro-Beijing business and professional groups. The pro-democracy parties were hence constrained as a permanent minority opposition in the legislature after 1997. The Chief Executive has never been popularly elected, but was elected from an Election Committee representing largely the same groups in the functional constituencies. By controlling the business and professional elites in the Election Committee, Beijing was largely able to dictate the election outcome (Wong 2010).

The hybrid nature of the regime means that pro-democracy groups in Hong Kong were involved in a long drawn-out fight for full democracy. Since the 1980s, there was common fear among Hong Kong people that their long-cherished freedom would be lost under Chinese rule, because of undue intervention from the north. A democratic government, many in Hong Kong believed, will be better able to fend off intervention from China and defend their freedom and way of life. It would also mean that the Hong Kong government's policies would be more geared toward the interests of Hong Kong people, instead of being tightly controlled by Beijing.

Beijing faced some constraints in dealing with such a restive population. It could not transplant its heavy-handed authoritarian rule to Hong Kong, because basic freedoms were promised in the Joint Declaration and the Basic Law. Reneging on these promises would damage Hong Kong's image as a free capitalist haven, which ran against Beijing's own economic interests. There were limits to the co-optation strategies because the cost of co-opting all the pro-democracy citizens was simply too large (Wong 2012; Lam and Lam 2013). As in most hybrid regimes, constitutional elections must be conducted in Hong Kong, and Beijing wants to guarantee that the democrats cannot extend their power and popular support through the limited elections. One possible way to weaken the democrats' influences and to entrench Beijing's political control was to outnumber them with a more docile population—one that was unfamiliar with Hong Kong's history of opposition, unaffected by Hong Kong people's collective memory of the Tiananmen crackdown, and most of all, less concerned with whether or not the city is fully democratic.

Immigrants from mainland China fulfilled all these criteria. As these immigrants had been previously socialized in a fairly different political environment, their political knowledge and ideology would be dissimilar to that of the Hong Kongers. For example, owing to the tight media censorship in China, the Tiananmen crackdown remains a mystery to most mainland Chinese. In addition, the repressive political environment and autocratic political education in the mainland also limits the people's exposure to democratic values. Immigrants from China thus will be less able to understand the pro-democracy opposition's obsession with democracy.

As mentioned, the flow of illegal immigrants to Hong Kong from the mainland largely stopped after 1980. Under the Basic Law, although Hong Kong was repatriated into Chinese sovereignty after 1997, migration from mainland into Hong Kong was not without constraints. Mainland Chinese who wants to migrate to Hong Kong must apply through the mainland authorities. If the application is approved, they can legally migrate to Hong Kong. Usually these immigrants will obtain permanent residence status after seven years in Hong Kong, which would make them entitled to all rights and welfare, including the right to vote, like Hong-Kong-born residents.

MAINLAND CHINESE IMMIGRANTS' POLITICAL PREFERENCES

Our central argument is that immigrants tend to have a status quo or pro-government bias, which makes them an attractive co-optation target of the authoritarian government. In this section, we devise various tests to check if such a bias exists among mainland Chinese immigrants in Hong Kong. It is important to emphasize that the point of comparison here is with the Hong Kong natives, not with the residents in these immigrants' home regions. From an authoritarian incumbent's perspective, whether immigrants are more anti- or pro-authoritarian than their fellow nationals is unimportant. Even if they are the most anti-authoritarian type in their own country, the authoritarian incumbent would still welcome them as long as they are more supportive of the political status quo than the natives in the host state.

A cursory examination of the Asian Barometer Survey Wave 3 data reveals that immigrants and natives differ significantly with respect to their vote choice. As may be seen from Table 1, two-thirds of the self-identified pan-democratic supporters are native Hong Kong people. On the other hand, only about 41 percent of immigrants indicate that they voted for a pan-democratic party. The division is quite clear: natives tend to support the pro-democracy opposition, while immigrants tend not to. It is noteworthy, however, that only 433 out of 1,207 respondents in the ABS data report their vote choice.

Table 1 Vote Choice: Natives vs. Immigrants

Data from: Authors' calculation based on Asian Barometer Survey Wave III.

Note: “Others” include both pro-establishment parties and parties/candidates with unidentified political affiliations. The data comes from Question 33 of the Asian Barometer Survey Wave III. The question asks respondents to indicate whom they voted for in the 2012 Legislative Council election. Respondents are allowed to name two parties at most because an ordinary voter would have two votes: one vote for the geographical constituency and another for the District Council (Second) functional constituency. Column proportions are reported. Standard errors of column proportions are in parentheses.

Table 1 shows only descriptive statistics. In addition, it examines vote choice, which is only one indication of political attitudes. In what follows, we will conduct a systematic empirical analysis of the differences between Chinese immigrants and native Hong Kong people. In particular, we are interested in four questions:

1. (1) Are Chinese immigrants different from the natives in regime evaluations and regime preferences?

2. (2) Are they less likely to vote for the opposition than natives?

3. (3) Are they less exposed to political information?

4. (4) Is exposure to political information correlated with political support for the opposition among immigrants and natives, respectively?

Depending on the questions, we use either ordered logit or logit regression to estimate the differences between the immigrants and natives. The logit specification is presented below:

$${\rm ln} \left( {\displaystyle{{P_i} \over {1 - P_i}}} \right) = \delta \times {\rm Immigrant}_i + x_i^{\prime} \,\,\beta + {\rm \epsilon}_i,$$

where P _i is the probability that an outcome variable y _i (e.g. vote for opposition parties) is equal to one, δ is the coefficient on the variable of interest Immigrant _i , which takes the value of “1” if respondent i is a Chinese immigrant and “0” otherwise, β is a vector of coefficients on a set of control variables x _i , and ϵ _i is an error term assumed to be independent and identically distributed. Our control variables include gender, education, income, marital status, age, age squared, district and religion fixed effects, and a dummy variable for Wave 4 respondents. Age squared is intended to capture the potential nonlinear effect of age on the outcome variables.

As already noted, mainland Chinese immigrants arrived in Hong Kong at different points in time, and they may have differing political attitudes. For instance, those who arrived in Hong Kong early may be more sympathetic toward the opposition because they might be refugees escaping from the Communist persecution. In the following analysis, we compare natives separately with two groups of immigrants: (1) all immigrants and (2) immigrants who arrived in Hong Kong after 1997.

Are Chinese Immigrants Different from Natives in Regime Evaluations and Regime Preferences?

The first set of hypotheses is concerned with immigrant satisfaction with the political and economic order of Hong Kong. We expect to see that the immigrants are more satisfied with status quo than the Hong Kong natives.

• Hypothesis 1a. Chinese immigrants have a more positive evaluation of Hong Kong's political institutions than the natives.

• Hypothesis 1b. Chinese immigrants would view the economic situation of Hong Kong more favorably than the natives.

• Hypothesis 1c. Chinese immigrants are more satisfied with Hong Kong's democratic development than the natives.

The next three hypotheses are related to the difference in regime preferences between Chinese immigrants and the Hong Kong natives. A potential reason for their difference is that these immigrants had been socialized in an authoritarian environment prior to their arrival in Hong Kong. Being exposed to political propaganda that emphasizes the advantage of one-party rule and patriotism, they may have formed political values vastly different from native Hong Kong people. These political values shape their worldview as well as how they evaluate government performance. To find out whether Chinese immigrants have a different value system from Hong Kong people, we derive a set of hypotheses related to regime preferences.

• Hypothesis 2a. Chinese immigrants have a weaker support for democracy than the natives.

• Hypothesis 2b. Chinese immigrants are more likely to approve of strongman rule than the natives.

• Hypothesis 2c. Chinese immigrants are more likely to approve of a paternalistic government than the natives.

We test the above hypotheses using data from the Asian Barometer Survey Wave 3 and Wave 4. The data were collected from face-to-face interviews, during which respondents had to answer an extensive set of questions concerning their political values, party identification, and attitude toward democracy and political institutions. The respondents were selected using a stratified random sampling design. The survey was conducted between September and November of 2012 (Wave 3) and between February 2016 and May 2016 (Wave 4). There were 1,207 and 1,217 individuals successfully interviewed in Wave 3 and 4, respectively. The question corresponding to each hypothesis is shown in the Appendix.

As may be seen from Table 2, there are significant differences in regime evaluations between Chinese immigrants and native Hong Kong people. The coefficient on the variable of interest is positive and statistically significant in the specifications of “Institutional Pride” and “Economic Evaluation,” regardless of whether we examine all immigrants or those who arrived after the retrocession. The results suggest that Chinese immigrants are more satisfied with the existing political and economic status quo than native Hong Kong people. The results support Hypotheses 1a and 1b. The effect of immigrant status is also substantively important. Being a Chinese immigrant is associated with a 0.354 and 0.317 increase in the log odds of being in a higher level of institutional pride and economic evaluation, respectively, although there is no significant difference between immigrants and natives in “Satisfaction with Democratic Development.” Hypothesis 1c is not supported by the data.

Table 2 Regime Evaluations and Regime Preferences: Immigrants vs. Natives

Data from: Asian Barometer Survey (ABS) Wave 3 and Wave 4.

Note: The estimation strategy for all specifications above is ordered logit. District and religion fixed effects are controlled, but not reported. Standard errors are in parentheses. *p < .05, ** p < .01.

When it comes to regime preferences, the difference between mainland Chinese immigrants and native Hong Kong people is less remarkable. The coefficient on the variable immigrant is statistically different from zero only in one specification; mainland Chinese immigrants, including those who came before and after 1997, tend to favor a paternalistic government more than the natives. ⁴

It would be more informative to express the effect in probability. Figure 2 shows the average marginal effects of the immigrant status on the probabilities of giving different answers to the two questions related to institutional pride and economic evaluation based on the first two columns of Table 2. It is clear that Chinese immigrants are less likely to give the answers “Bad” or “Very Bad” when asked to evaluate the Hong Kong current economy, and more likely to answer “Average” or above. Similarly, they are more likely to agree with the statement “I am proud of our [Hong Kong] political system,” and less likely to disagree with it.

Figure 2 Average Marginal Effects of Immigrant Status on Institutional Pride and Economic Evaluation

The results in this section indicate that mainland Chinese immigrants in Hong Kong appreciate the political and economic status quo more than native Hong Kong people. Although there is no evidence showing that the immigrants are less supportive of democratic values, they are nevertheless less resistant to paternalistic rule.

Are Immigrants Less Likely to Support the Opposition?

This set of hypotheses is related to the difference in political identification and voting behavior between Chinese immigrants and the Hong Kong natives. To the authoritarian incumbent, how the immigrants act is arguably more important than what they think. For this reason, in addition to political identification, ⁵ we also examine turnout and vote choice.

Thanks to the electoral formula of the Legislative Council, which is list proportional representation using Hare Quota and Largest Remainder, the party system in Hong Kong is extremely fragmented (Carey 2017; M. Wong 2015). Despite the dominant political cleavage between the pro-establishment camp and the opposition, there also exist a plethora of independent candidates and political groups. ⁶ Because those who do not support the pro-establishment parties may not necessarily support the opposition, it is important to examine political support for each of the two rivalling camps. Hence, we have the following hypotheses:

• Hypothesis 3a. Chinese immigrants are more likely to identify themselves with pro-establishment parties than the natives.

• Hypothesis 3b. Chinese immigrants are more likely than natives to vote for pro-establishment parties.

• Hypothesis 3c. Chinese immigrants are less likely than natives to identify themselves with opposition parties.

• Hypothesis 3d. Chinese immigrants are less likely than natives to vote for opposition parties.

• Hypothesis 3e. Voter turnout of Chinese immigrants is lower than that of the natives.

From the establishment's perspective, an unwanted scenario is that the immigrants are more likely to support the opposition. Apart from that, it makes little difference between whether immigrants are less likely to support the opposition or the immigrants are more likely to support the pro-establishment camp, as both outcomes would weaken the opposition.

To test the above hypotheses (3a–3e), we use two distinct datasets. The first is the ABS Wave 3 data, which cover the Legislative Council election in 2012. ⁷ The second dataset comes from a new multi-wave election study known as the Hong Kong Election Study (HKES). ⁸ We use the HKES's post-election public opinion survey of the District Council election in 2015. The post-election survey was implemented by YouGov on a sample of its online panel members. Conceivably, the HKES and the ABS are not exactly identical. For instance, the former questionnaire does not contain a question asking immigrants' length of residence in Hong Kong, which prevents us from identifying post-1997 immigrants. That said, the two surveys share many common variables, including political identification, turnout, and vote choice.

The results, which are presented in Table 3, seem to favor the establishment more than the opposition. First, consider the all-immigrants sample. Mainland Chinese immigrants are more likely to vote for the pro-establishment parties (the second column), and less likely to identify themselves with the opposition or to vote for them (the third and fourth columns). The coefficients on the variable of interest immigrant are statistically significant. The effect is also of substantive significance. For example, converting the coefficient into probability, Chinese immigrants are 6 percent less likely to identify themselves as pan-democratic supporters (the third column) and 13.6 percent less likely to vote for pan-democratic opposition parties (the fourth column). The ABS data also show that Chinese immigrants are less likely to show up in the voting booth than native Hong Kong people. The results support Hypotheses 3b, 3c, and 3d.

Table 3 Political Identification and Vote Choice with ABS Data: Immigrants vs. Natives

Data from: Asian Barometer Survey (ABS) Wave 3 and Wave 4.

Note: The estimation strategy for all specifications above is logit. District and religion fixed effects are controlled, but not reported. Standard errors are in parentheses. * p < .05, ** p < .01.

Another striking difference between the two groups is their voting behavior. Chinese immigrants are less likely to identify themselves with or vote for pro-democracy opposition parties. The coefficient on the variable of interest is negative and statistically significant. Converting the coefficient into probability, Chinese immigrants are 6 percent less likely to identify themselves as pan-democratic supporters and 13.6 percent less likely to vote for pan-democratic opposition parties. The results support Hypotheses 2a and 2b. When it comes to turnout, the coefficient of interest has a negative sign, although it is not significantly different from zero. In other words, we find no evidence suggesting that Chinese immigrants are less likely to show up in the voting booth than native Hong Kong people.

For immigrants who arrived in Hong Kong after 1997, we still find a significantly negative coefficient on immigrant when it comes to the identification with the opposition. Although the post-1997 immigrants are less likely to vote for the opposition than are natives (the second last column), the coefficient is no significantly different from zero. It is important to note that the number of observations in the specifications related to vote choice shrinks to about 200, which is likely to increase the variance of the estimate. Finally, the post-1997 immigrants are significantly less likely to vote than natives. Converting the coefficient on immigrant in the last column to probability, the post-1997 immigrants are 23 percent less likely to vote than are native Hong Kong people.

The HKES data also show a similar correlation between immigrant status and political support for the pro-establishment camp. As may be seen from Table 4, the coefficients on immigrant are positive and significant at 1 percent in the two specifications related to identification with and voting for the pro-establishment camp. Converting the coefficients into probability, mainland Chinese immigrants are 13.1 percent more likely than native Hong Kong people to identify themselves with the pro-establishment camp and 21.4 percent more likely to vote its parties. The coefficients on immigrant are negative in specifications related to the political support for the opposition, but they are not significantly different from zero. In addition, there is no detectable difference in voter turnout between immigrants and natives (the last column).

Table 4 Political Identification and Vote Choice with HKES Data: Immigrants vs. Natives

Data from: Hong Kong Election Study (HKES) 2015 District Council Election Post-Election Survey.

Note: The estimation strategy for all specifications above is logit. District and religion fixed effects are controlled, but not reported. Standard errors are in parentheses. * p < .05, ** p < .01.

Combining the data from the two surveys, which are based on elections at different levels, we find strong evidence that immigrants are unlikely supporters for the opposition. On the contrary, if they decide to vote, they are more likely to support the pro-establishment camp or vote someone other than the opposition.

Political Information, Vote Choice, and Chinese Immigrants

As discussed, while the lack of prior exposure to local political information makes immigrants less likely to understand and support the cause of the opposition, subsequent exposure to political information after arrival may alter their political preferences. The extent to which this can change their political identification or vote choice is ex ante unclear. In this section, we examine the correlation between exposure to political information and political attitudes. In particular, we derive several testable hypotheses:

• Hypothesis 4a. Natives have greater exposure to political information than Chinese immigrants.

• Hypothesis 4b. Having greater exposure to political information would change Chinese immigrants' political identification.

• Hypothesis 4c. Having greater exposure to political information would change Chinese immigrants' vote choice.

• Hypothesis 4d. Having greater exposure to political information would change Chinese immigrants' turnout rate.

To test the above hypotheses, we operationalize “exposure” using an ABS question that asks respondents to report how often they gain access to political news. The value of this variable runs from 1 to 5, the higher the greater exposure to political information. Unfortunately, the HKES survey does not contain a question on respondents' exposure to political news. Instead, it has a question that asks respondents the frequency of sharing news with acquaintances through social media or mobile communication platforms (e.g. Whatsapp, WeChat). Assuming that exposure to political information is positively correlated with one's frequency of news sharing, we use this question to proxy the respondents' exposure to political information.

First, consider the difference between immigrants and natives with respect to their exposure to political information. As may be seen from Table 5, perhaps surprisingly, mainland Chinese immigrants are more frequently exposed to political information than natives, regardless of whether all immigrants or a subset of them who arrived in Hong Kong after 1997. In the HKES data, however, we find no detectable difference in the frequency of news sharing between natives and immigrants. Taken together, we find no evidence for Hypothesis 4a.

Table 5 Exposure to Political Information and News Sharing: Immigrants vs. Natives

Data from: Hong Kong Election Study (HKES) 2015 District Council Election Post-Election Survey and Asian Barometer Survey (ABS) Wave 3 and Wave 4.

Note: The estimation strategy for all specifications above is ordered logit. District and religion fixed effects are controlled, but not reported. Standard errors are in parentheses. * p < .05, ** p < .01.

The ABS data do not permit us to investigate why the immigrants' exposure to political information is greater than the natives. One possibility is that the immigrants are anxious to assimilate into Hong Kong society. This provides them a strong incentive to gain more exposure to local news media.

With the exposure variable, we re-run the specifications related to “Pro-Establishment,” “Vote for Establishment,” “Pro-Opposition,” and “Vote for Opposition” as in Table 3. For each outcome variable, we run the specification on three samples: all immigrants, post-1997 immigrants, and natives. Owing to insufficient observations, we are unable to run specifications related to vote choice in the sample of the post-1997 immigrants. The results are presented in Table 6.

Table 6 Exposure to Political Information, Political Identification, and Vote Choice with ABS Data: Immigrants vs. Natives

Data from: Asian Barometer Survey (ABS) Wave 3 and Wave 4.

Note: The estimation strategy for all specifications above is logit. District and religion fixed effects are controlled, but not reported. Standard errors are in parentheses. * p < .05, ** p < .01.

As may be seen from the table, the coefficient on immigrant is statistically significant only in the specifications related to political identification, which indicates that being exposed to more political information would make one more likely to identify with the opposition camp. The finding is consistent with Hypothesis 4b. The effect applies to both the samples of all immigrants and natives. Interestingly, for natives, exposure to political information is also positively correlated with political identification with the pro-establishment camp. Note that although greater exposure to political information may alter one's political identification, it may not be sufficient to change one's vote choice, as the immigrant variable is not significantly different from zero in any of the vote-choice specifications. We, therefore, find no evidence in support of Hypotheses 4c.

We also try to test Hypotheses 4b, 4c, and 4d using the HKES data. We re-run the specifications in Table 6, the results of which are displayed in Table 7. As shown in the Table, although news sharing is correlated with neither one's political identification nor one's vote choice, it is predictive of one's probability to vote. In the specifications related to turnout, the coefficient on the variable of interest immigrant is positive and significant in both the native and immigrant samples. Only Hypothesis 4d is supported by the HKES data.

Table 7 Sharing Political Information, Political Identification, and Vote Choice with HKES Data: Immigrants vs. Natives

Data from: Hong Kong Election Study (HKES) 2015 District Council Election Post-Election Survey.

Note: The estimation strategy for all specifications above is logit. District and religion fixed effects are controlled, but not reported. Standard errors are in parentheses. * p < .05, ** p < .01.

IMMIGRATION POLICIES IN THE HONG KONG SPECIAL ADMINISTRATIVE REGION (HKSAR)

Our findings in the previous section demonstrate that mainland Chinese immigrants in Hong Kong have fairly strong predispositions toward the status quo and the government, which should make them an attractive target of co-optation by the pro-Beijing ruling coalition. Although we are unable to pry into the inner workings of the ruling coalition, we can nevertheless examine observable practices of the HKSAR government and pro-Beijing parties and groups, from which we can draw an inference of the ruling coalition's attitudes toward the immigrants. We begin with the government's immigration policies.

Immigration Policies of the HKSAR

There are multiple schemes by which mainlanders could apply to migrate to Hong Kong. The most significant one is the “One-way entry permit scheme.” After 1997, every day a maximum of 150 mainland Chinese could come to settle in Hong Kong, which means about 55,000 every year. A majority of these daily quota were devoted to family reunion, mostly wives, children, and relatives of Hong Kong residents. Not surprisingly, their labor force participation rate is significantly lower than natives (Chiu and Lui 2004). Those who do join the labor market, tend to concentrate in low-skilled sectors due to their lack of local credentials or language barriers. The detailed criteria of selection by the mainland authorities were never disclosed, and the Hong Kong government had little say over the decisions about who can come to Hong Kong.

Other than the “One-way entry permit scheme,” the Hong Kong government introduced different immigration schemes after 1997, with the manifest policy goal of attracting talents and investments. Most of these schemes, however, were targeted at Chinese citizens or mainland Chinese, and the schemes have been gradually relaxed in recent years. These schemes could largely be divided into three types: (a) those that attract professionals or “talents” or investors to come to Hong Kong; (b) those that facilitate mainland students in Hong Kong to stay and work in Hong Kong after graduation; (c) those that encourage for spouses and dependents of migrants from various schemes to come to Hong Kong.

In 2003, the Hong Kong government initiated the Admission Scheme for Mainland Talents and Professionals. Enterprises in Hong Kong can apply to employ mainland professionals for teaching, research, high-skilled and other professional posts. In 2006, the Quality Migrant Admission Scheme allowed a quota of 1,000 migrants every year. Although the scheme included non-Chinese applicants, it was seen as a relaxation of control for mainlanders since the “quality migrants” do not require a pre-contract by a Hong Kong employer, as the 2003 scheme stipulates. The rules were further relaxed in 2008 and 2015, including relaxation of rules on age limits and extension of residing in Hong Kong, making it easier for the migrants to get to the seven-year residence limit.

Since the 1990s, Hong Kong has been a popular place for mainland students to seek university and postgraduate education. Since 1999, the Hong Kong government allowed mainland students to apply for Hong Kong universities as undergraduate students. In 15 years, the number of mainland students increased from less than 3,000 in 2002 to about 12,000 in 2014. In 2007, in a bid to make Hong Kong a “regional educational hub,” the Hong Kong government pushed the new Immigration Arrangements for Non-local Graduates to facilitate non-local students to stay beyond graduation to find jobs in Hong Kong. Mainland students can stay in Hong Kong unconditionally for one year after graduation (previously three months), even if they cannot find an employer immediately after graduation.

The various migrant schemes were further relaxed in recent years, mostly making it easier for spouses and dependents to come to Hong Kong following the imported talents. Starting from 2007, the spouses and dependents (including children under 18 and parents over 60) of investors and talents approved under the various migrant schemes and the non-local graduates can apply to come to Hong Kong. The scheme also applied to the mainland spouses and dependents of non-Chinese or overseas Chinese.

Table 8 showed the number of mainlanders who were approved to reside in Hong Kong under various migrant schemes from 2011 to 2016. It shows that while some schemes are not restricted to Chinese applicants, mainland Chinese made up a large chunk of the successful applicants. Other than the Quality Migrant Admission Scheme, the success rates were also very high, showing that the Hong Kong government was very forthcoming in inviting or allowing mainland immigrants to Hong Kong.

Table 8 Application and Approval Figures of Immigration Schemes, 2011–16

Data from: Authors' own compilation based on government reports and statistics.

Thanks to the various immigration schemes, Hong Kong has continued to attract a large number of immigrants, especially those coming from mainland China. A useful indicator of the scale of the influx is to compare the number of immigrants to the number of live births. Table 9 shows that the ratio between the two is almost 1:1. ⁹ In other words, immigrants have become a main source of population growth in Hong Kong.

Table 9 Number of Immigrants vs. Number of Live Births in Hong Kong

Data from: Various Reports from the HKSAR Census and Statistics Department.

Note: The annual number of immigrants consists of the following categories: One-Way Permit Holders, Admission Scheme for Mainland Talents and Professionals, Capital Investment Entrant Scheme, Quality Migrant Admission Scheme, and Immigration Arrangements for Nonlocal Graduates. The last three categories comprise non-mainland immigrants. According to HKSAR Government Press Release (2013), mainland immigrants made up for 95 percent of all immigrants under these three immigration schemes.

Electoral Mobilization of Immigrants

As discussed, the ideological predisposition and the lack of prior exposure to local political information make immigrants an attractive co-optation target for electoral authoritarian regimes. In the case of Hong Kong, the co-optation task is largely done by pro-Beijing parties and a number of intermediate organizations. In particular, through the liaison and coordination of the Central Government Liaison Office (CGLO), ¹⁰ they connected with the immigrants from China and turned them into a solid support base. The CGLO had full information of the mainlanders who were about to come to Hong Kong. The pro-Beijing Hong Kong Federation of Trade Unions had numerous offices in mainland cities, which would contact these would-be immigrants and offered logistic or material help before they came (Chang and Tseng 2011). When the immigrants reached Hong Kong, they would be contacted by their respective hometown associations, offering material assistance and logistic help. These hometown associations were coordinated under the Hong Kong Federation of Fujian Associations and Hong Kong Federation of Guangdong Associations, both umbrella organizations that worked closely with the CGLO. ¹¹

In 2010, the New Home Association was set up in Hong Kong, which quickly expanded to an enormous organization with five district offices providing services to new immigrants. By 2015, it claimed to have more than 100,000 members. ¹² It recruited members largely from its three offices in the mainland cities, referrals from other pro-Beijing associations, and their own activities. Membership is free, but members would receive numerous free gifts and benefits including food, concert tickets, discounted medical services, scholarships, and discounted tours and meals. ¹³ In collaboration with the hometown associations and other pro-Beijing community organizations in Hong Kong, it offers many social, recreational and cultural activities for the new immigrants. This allows these associations to maintain good connections with the new immigrants from China. In election times, all these associations became effective mobilization machines which would draft their members as campaigners and mobilize them to vote.

Professionals and students who entered and stayed in Hong Kong after graduation would be wooed by different kinds of “activities” and organizations. “Youth associations” were organized on the basis of their respective hometowns or home provinces (e.g. the Hangzhou Hong Kong Youth Association). They would arrange activities related to career development such as visits to enterprises in China, and talks on starting up a business or financial investment, which attract the migrant professionals. There were also other “soft” activities such as boat cruises, sports activities or wine-tasting. ¹⁴ The membership records of these associations would become valuable databases useful for electoral mobilization.

Note that the membership of the above organizations is not confined to immigrants who arrived in Hong Kong after 1997. For example, long-time immigrants can also join activities organized by their hometown associations any time. From the establishment's perspective, the goal of setting up these umbrella organizations is to expand its local political support by reaching out to the potential supporters, and an immigrant's “potential” lies in his or her ideological predisposition toward the establishment, not in his or her length of residence in Hong Kong.

How do the opposition parties react to pro-Beijing parties' co-optation of immigrants? They tend to have limited options. As pointed out by S. Wong (2015), the opposition parties can often count on ideological appeals to reach out to their potential supporters, as they are working under severe resource constraints. In the case of mainland Chinese immigrants, the opposition parties usually have little detailed information about the immigrants. For one thing, they are unable to contact the immigrants prior to their arrival, which is something pro-Beijing groups have been trying to do. The opposition parties also lack resources for the distribution of material benefits, when compared to the pro-government groups. Their constituency services appeal to all, and seldom aim specifically to immigrants. Their ideological positions on values such as democracy, human rights or rule of law, in general have less appeal to the new immigrants. In addition, the rise of radical opposition parties, which advocate localism or even separatism from mainland China, has further constrained some traditional opposition parties, who would be criticized for selling out local Hong Kong people's interests if they try to fight for the immigrants' rights.

CONCLUSION

Extant studies on electoral autocracies focus predominantly on how authoritarian governments manipulate elections to secure victory (Hyde 2011; Levitsky and Way 2010; Schedler 2002). Important as it is, electoral manipulation alone cannot guarantee regime survival; an autocratic government cannot survive without at least a certain degree of political support from society. In this article, we examine one potential source of support that has been under-studied in existing research: immigrants. We argue that electoral authoritarian states can often benefit politically from the influx of immigrants, as immigrants, who self-select to migrate into these states, tend to support the politically conservative incumbents. Indeed, in our case study of Hong Kong, we find that mainland Chinese immigrants are more satisfied with the political status quo of the city than native Hong Kong people, and more supportive of pro-Beijing parties.

Perhaps immigrants, with increased knowledge and resources, may eventually become more sympathetic toward the opposition, together with its pro-democracy cause. Yet, in the case of Hong Kong, even though mainland Chinese immigrants consume political news more frequently than the natives, greater exposure to political information does not make them more likely to vote for opposition parties. In other words, the selection effect seems to be far more important than the effect of political knowledge or learning.

Our findings have important implications for Hong Kong politics. First and foremost, Hong Kong society has experienced an intense struggle for democratization, which manifested itself in an unprecedented occupy movement in 2014. An increasingly vocal, if not radical, civil society has posed serious challenges to the legitimacy of the establishment. Under these circumstances, mainland Chinese immigrants form an important and solid base of support for the pro-Beijing conservative coalition in the city's limited elections to counter the challenges of the pro-democracy opposition.

In addition to their electoral effects, mainland Chinese immigrants may also have influences on non-institutional political participation. Facing the challenges of the civil society, the ruling coalition has increasingly taken a hard line against the popular demand for political liberalization, as shown by a rising number of counter-mobilization in recent years (Cheng 2016). The pro-Beijing conservative coalition may take advantage of mainland Chinese immigrants' status quo bias to mobilize them to participate in counter-protests. If this is the case, political polarization is likely to escalate in years to come.