Public Opinion in Authoritarian States

The latent causes of faction are thus sown in the nature of man; and we see them everywhere brought into different degrees of activity, according to the different circumstances of civil society.

Dean Karalekas writes the following in his PhD thesis, Identity and Transformation: Perceptions of Civil-Military Relations in the Republic of China (Taiwan):

In late May, 2016, the newspapers in Taiwan reported on a poll conducted by the Taiwanese Public Opinion Foundation in which 80 percent of respondents reported that they identified as Taiwanese, with a mere 8.1 percent identifying as Chinese (7.6 percent identified as both). This is in stark contrast to similar polls conducted in the late 1980s and early 1990s in which the vast majority reported identifying as Chinese or both. What happened in the short span of a single generation that could change an entire nation’s self-perception so radically?

In fact, no such sea change ever took place. Indeed, neither historical roots nor system-level changes are sufficiently able to influence group identity on the scale of what has been measured (Chu 2004: 498). Rather, according to researcher Frank Muyard (2012), what happened in the 1990s was not a rising Taiwanese identity, but rather a newfound freedom to express that sentiment, including in polls. As with many oppressed and/or colonized peoples throughout the world and throughout history, the Taiwanese identity can be conceptualized in opposition to an external “other.” Before the era of democratization, Taiwaneseness stood in opposition to the KMT and that party’s authoritarianism, White Terror, and Leninist efforts to control all forms of social expression by subsuming Taiwaneseness into a subset of Chineseness. For evidence of the expression of this phenomenon one need look no further than the political accretion of opposition forces who chose the name Tangwai (黨外), or literally ‘outside the party’.

In the 1990s democratization opened up new avenues for exploring the idea of Taiwaneseness and what it means. Several things happened in the 1990s. First, the DPP established itself as a legitimate and legal alternative to the KMT and standard bearer of Taiwan-centered politics. The rise of democratic politics meant that the Taiwan identity could no longer be defined as resistance to the System: the Tangwai were now part of the system in the form of the DPP and its allies. Further, the KMT under Lee Tung-hui, who was president throughout the entire decade, co-opted many DPP programs and positions, and thus appeared to be Taiwanizing. That made it difficult to oppose the KMT as an anti-Taiwanese party (Wang & Liu 2004: 577). Survey results from the late 1980s and early 1990s show a high number of respondents identifying as Chinese because, according to Muyard (2012), the Taiwanese had long learned that it was safer to lie to the State. As society opened up through democratization and the end of the authoritarian era, people felt more comfortable publicly admitting to a sense of Taiwaneseness, and telling the truth in such surveys. Therefore there was no real “emergence” of a Taiwanese identity in the 1990s, but rather the uncloaking of a theretofore hidden one. [1]

The historical context for this passage was a weak, and in some cases non-existent, sense of nationalist identity among Han Chinese in the Qing Dynasty. In those days, Han Chinese tended to identify with the village or (if in the south) clan first, Han-ness (made more relevant in the Qing by the rule of an alien Manchurian power and the privileges granted Westerners in the concessions) second, and regional identities (which mapped onto linguistic divides inside China itself) third. Those who identified with the Chinese state were almost exclusively intelligentsia—first Confucian literati, shuffled across the Qing Empire as agents of the state, and then later the thinkers, activists, and statesmen enraptured with the ideals of the May Fourth movement, the sort of people who read national political journals and Herbert Spencer. These intellectuals believed that the weak sense of national identity then prevalent among the bulk of the Chinese people was one of the central reasons China had suffered through so many bitter misfortunes in modern times. When these same folks seized power in the 1920s, 30s, and 40s, they put building a coherent national identity at the front of their political programs.

The Taiwanese missed out on all that. They went directly from control by the minimalist (one might even say “night watchman”) Qing empire, in which they were an imperial backwater, to direct control by Japan in 1895, well before popular nationalism began to blossom on the mainland. Japanese colonialism was benign by contemporary imperialist standards; economic investment in the island and the early, conscious co-option of local elites into the Japanese colonial system gives the era a sort of rosy glow. Many Taiwanese are still nostalgic for this era today. Very few native Taiwanese are nostalgic for the era that followed.

The Kuomingtang (“Nationalist Party”), leadership felt far less secure than the Japanese colonial administrators did. They came from a realm of rebellion, subversion, and Communist plots; they saw these things where ever they went. They initially treated the Taiwanese people as a treasure trove to be plundered for the sake of winning the war across the strait; that, combined with the distinctive corruption of the 1940s KMT, led to genuine popular resistance to KMT rule. This resistance was extinguished through massacre. Taiwan became a Leninist police state. The Nationalists embedded themselves into the military, industry, media outlets, and the educational system. Through violence, surveillance, corruption, and terror they got the native Taiwanese to fall in line.

The Nationalists also began imposing their Chinese nationalist program on the people in Taiwan. Had this new nationalist identity not been associated with the terrors of martial law and the stifling control of a Leninist state it would likely have had a warmer reception. Even today, for many Taiwanese appeals to Chinese identity carry with them the flavor of martial law.

That is the context. Now for my question: did Taiwanese identity spring up out of nowhere in the 1990s because it was always there, just hiding? Was it a strong current rushing just beneath the surface, only concealed because “the Taiwanese had long learned that it was safer to lie to the State”?

There are many authoritarian regimes in the world today about which we might ask similar questions. To take an easy example, consider Tang Wenfang’s 2018 essay for American Affairs, “The ‘Surprise’ of Authoritarian Resilience in China.” Tang uses reams and reams of Chinese polling data in his arguments about the nature of the PRC regime.[2] But this data comes from country where having the wrong opinion can get you in trouble. In countries like China, what can polling really tell us?

The purpose of censorship and thought management differs from regime to regime. The contention of this post is that for many of the most effective authoritarian systems, controlling the thoughts of the ruled is secondary to shaping social cleavages in the population.

We began by looking at the rapid transformation of Taiwanese public opinion once the KMT’s old instruments of censorship and coercion had been dismantled. Consider another example of opinion change, one far faster than the dramatic growth of Taiwanese identity in post-martial law era:

Image source: Mark Fahey, “Free Trade: Americans Inconsistent But Like it More Than Their Leaders ,” CNBC (12 August 2016).

Witness an incredible transformation! In less than two years an entire wing of the Republican Party metamorphosed from a collection of free trade zealot into a cartel of pro-tariff troglodytes! Why did this happen? Did the Trump administration dismantle a national system of spies, censors, and political prisoners? Were the Republican masses really anti-trade all along, but afraid to proclaim their opinion until their man was in power?

Political convictions do not work the way most people think they do.

“It is difficult to get a man to understand something,” quipped Upton Sinclair, “when his salary depends upon his not understanding it!” [3] It is easy to see the truth behind this statement. But also easy to see are its limitations. Very few Americans have a direct financial stake in same-sex marriage, nuclear power, the size of the American military, or a balanced budget. The number of Americans whose salary is dependent on a particular understanding of affirmative action or global warming is very small. Even when an issue of public importance has a direct and undeniable impact on household finances (say tax rates or health care), members of that household will often vote against the party whose policies would save them the most money. If asked to guess an American’s opinion on tax cuts, you are would be better off knowing whether that person is pro-life or pro-choice than you would knowing their income.

This is all very odd. A man’s opinion about tax rates should not have anything to do with his opinion about abortion, but in the American context they often do. A woman’s conclusions on the life history of Brett Kavanaugh should hold no relation to her opinions about conservationism. One’s convictions about transgender bathrooms should tell you very little about one’s attitude towards Russia or Iran. Yet they do!

Economists Eric Gould and Esteban Klor published an interesting paper a few weeks ago that analyzed American voters who believe that abortion is the single most important issue in American politics. In the 1980s abortion was a relatively non-partisan issue. People on both sides of the issue could be found on both sides of the isle. This has changed. The issue is now highly partisan. If your priorities are pro-life, you join the GOP. If pro-choice politics drives your value system, you sign up with the Democrats. In the late ’80s and ’90s many Americans switched parties or became affiliated with a party for the first time precisely for this reason.

That is not too surprising. “One-issue voter chooses the party that agrees with them on their one issue!” Nothing shocking there. But Gould and Klor were not interested in the decision to join one of the two parties. They were looking at something different: how a one-issue voter’s view of other issues changed after they joined their new party. Did Republicans who joined the party solely because of its position on abortion end up changing their views on taxes and regulation? Was a person who signed up with the Democrats purely because of its pro-choice platform influenced by that platform’s positions on oil, Iraq, or education?

A few excerpts from their study:

The results show that individuals not only changed their party as of 1997 in accordance to their abortion views in 1982, but this party realignment led individuals to shift their preferences over other policy dimensions accordingly. That is, as people switched parties due to their initially held abortion views, their other political views tended to follow suit. These findings hold for a variety of political, economic, and social issues that are not directly related to abortion, but are stronger for issues that are more partisan nature…. Individuals with pro-life attitudes in 1982 are 14.5 percentage points more likely to hold a conservative view in 1997 than pro-choice individuals even after controlling for this attitude in 1982… The size of the coefficient implies that a standard deviation increase in the likelihood of supporting the Republican Party increases the chances of holding a conservative view by 0.44 percentage points…

We see that a pro-life view in 1982 causes an increase of almost 11 percentage points in the likelihood that individuals oppose the legalization of marijuana and that women should have an equal role in society as men, both measured in 1997. A pro-life view in 1982 increases the chance that a person supports prayer in public schools in 1997 by 14 percentage points…

A priori, there is no reason to believe that views on abortion are correlated with views on unions, big business, or the size of the government. If anything, individuals supporting the right of the government to interfere with a person’s reproductive decisions by restricting or banning access to abortion should also be a likely candidate to favor bigger and more intrusive government interventions in other areas. This, of course, is contrary to reality, where individuals that have pro-life attitudes also tend to favor a smaller government. We argue that this correlation is caused by these views being bundled by the Republican Party. Of course, the same argument applies to having a pro-choice attitude and preferring a bigger government, which are bundled into the platform of the Democratic Party. The reduced form estimates in Table 7 show that there is a positive correlation between pro-life views in 1982 and holding a Republican point of view on economic attitudes (see columns 2, 6,and 10)…. Finally, the results of the second stage confirm that identifying with a given party increases the likelihood that the individual adopts the party’s view on unions, big business, and the size of the government. [4]

Political convictions are not a careful calculation of interests. They are rarely grounded in an ideology cannily crafted from first principles. Like most of our beliefs, the reasons for political convictions are constructed after we have settled on them.[5] Attempts to make the tangle of positions endorsed by Republicans or Democrats congeal into one coherent ideology are post-hoc justifications. Ideological coherence ultimately matters less than commitment to a coalition.

At the broadest level, the level which speaks equally to hunter gatherer bands and mass democracies, political ideology is a product of what psychologists have dubbed “coalition psychology.”[6] Humans form alliances. We are primed to identify with groups. It is second nature to divide our social world into various groups of ‘us’ and ‘them.’ These groups are distinguished by shared symbols, signals, and norms. These can be obvious: slang, dress codes, extreme behavior. They can also be subtle. Our sense of what is a ‘proper’ or ‘odd’ opinions or behavior fit in this last category.

Most of the time, humans act normally. We believe normal things.

“Normal” is another way of saying, “What is considered acceptable within my in-group.” Changes in political convictions usually have more to do with a shifting sense for what is normal (or in the parlance of opinion, “reasonable”) than with some killer argument delivered at a debate podium. If you think about your own experience the truth of this will be rather obvious. Almost no one changes their mind because of one day’s argument. Instead, we slowly start to see ideas we once would have rejected out of hand as “reasonable” only after we have been exposed to other people whom we respect or identify with who believe them. (Though I have not seen any research on it, I suspect that this applies equally well to people we only know through their written words. Once we respect an author or come to identify with her—even if she is dead—we will take her augments for uncomfortable ideas more seriously).

What I have seen research on is the role political leadership plays in all of this. There has been a lot of work in political psychology on the way in which leaders shape political preferences. [7] What they describe accords with what I have laid out above. Respected leaders can bring about great shifts in public opinion by shifting to a new position. Why? Because they are one of the few individuals who large numbers of people trust and identify with. If the leader of a party or movement adopts a new position, it will automatically seem ‘reasonable’ to most of those who identify with or trust that leader.

As a concrete example of what this looks like in real life, imagine a 50-something evangelical who has lent his support to Trump. He is a Republican because of the abortion issue. He reads the newspaper, but does not have any special knowledge or interest in the economics and law of international trade. He traditionally supported free trade because that is what all of the people he trusts supports. Their arguments instinctively sound the most plausible out there because they are tied in with people and explained in language he already endorses. This fellow rejects critics of free trade out of hand because those critics are associated with the opposing coalition. But then, a change! Trump moves towards trade war. Our man now has a strong reason to reconsider his position. Trump is a man whose instincts he trusts. He respects Trump’s positions, even when he disagrees with them. The bounds of what was ‘reasonable’ just shifted. Now he is much more willing to consider arguments against the existing trade regime regardless of source. If Trump is against it, there is probably a good reason for it, and it is worth hearing those reasons out.

At this point (and this is my personal sense for where things stand in 2019) both the pro- and the anti- side of the trade issue will strike the average Republican as ‘reasonable’ positions to hold. But if protectionism continues on long enough, those who care about it most will abandon the GOP coalition. Free trade will be associated with the out-group, and our evangelical friend (who would not particularly care one way or the other were he left to his own devices) will harden his heart against the WTO. Expect this process to occur even quicker if the Democrat running for President in 2020 makes tearing down tariffs a centerpiece of his or her campaign.[8]

This brings us back to the question we started with. Was the sudden swing towards independence and Taiwanese identity in the ’90s something new to Taiwanese politics, or just something newly uncovered once martial law had been dismantled?

I find it unlikely that a large section of the Taiwanese population secretly wanted Taiwanese independence in the 1970s. The remarkable changes we saw in the late ’80s and early ’90s can be explained just as easily with the mechanisms that lead to rapid changes in public opinion elsewhere. There was clearly a social division in martial law Taiwan. The new immigrants from the mainland, their children, government workers, and KMT clients had natural attachment to the sitting regime. Other parts of Taiwanese society were isolated from it. They feared it or were frustrated with it. The Tangwai movement was able to use this to their advantage. However, while a small cadre of intellectuals did conceive of Taiwanese identity in nationalist terms, their ideas were just that: ideas. Censorship and police control was effective at keeping those ideas away the social identity of masses of disaffected Taiwanese.

Once the censorship and political warfare systems were dismantled, however, there was an opening. Then anti-KMT Taiwanese were exposed to Taiwanese nationalism in an anti-KMT context. Ideas that had previously been limited to an underground elite could now be voiced to the audience best primed to hear them. The people expressing these ideas were recognized as the intellectual and political elite of the anti-KMT side. Their position quickly became reasonable, and then immensely popular. Without censorship and thought control, ideas were free to harden into identities. Taiwanese nationalism was finally married to the natural constituency most receptive to it.

If there had not been an existing social cleavage between the KMT clients and the ‘old’ Taiwanese this would not have been possible. The new, combative political identities of the ’90s needed both of these things to emerge: a social cleavage in the population, and a set of ideas capable of politicizing that division. A subversive political platform with no constituency is powerless. A constituency with no platform is just another group of people. The role of a functioning system of thought control is to quarantine political ideas away from divisive social identities.

Return to the problem I addressed a few weeks ago: why do so many Chinese students become more nationalistic and supportive of the Communist Party when they travel abroad for work or study? How is this possible, even though they have left censorship behind? Because for them, the in-group that defines what is reasonable and what is not is other Chinese students! In 2019, the vast majority of Chinese students live in a social world comprised almost entirely of other Chinese. That is their in-group. That is the relevant population of people they respect. That is the coalition they are a part of. And when these same students are exposed to ideas which criticize the Party, it is usually in a context where the Chinese students feel personally attacked. They jump to the defense of who they are, just like Democrats and Republicans do when they face off with each other.

But there is the larger take-away from all of this. If I am correct, then Chinese public opinion now does not reflect what it would be in a more open political system. This is not because Chinese public opinion polling is all bad (though on many issues I think the numbers are clearly suspect), but because authoritarian control has frozen the underlying forces that naturally create divisive political identities in more democratic societies. In countries like Taiwan, Germany, India, or the United States, social cleavages naturally mature into political cleavages. Were China an open society the same thing would eventually happen there.

This is one reason I am a far more bullish on the prospects of integrating a Chinese democracy into a peaceful international order than I am a China ruled by the Communists. I do not have any illusions about where the state of Chinese public opinion is at the current moment. The majority of Chinese want to swallow up Taiwan, ascend to the heights of global superpower, and win a few bloody battles along the way to restore the country’s lost honor. That is simply how things are. Pretending that those ideals are entirely fostered by the Communist Party is not helpful. However, the important thing is not the ideals being fostered by the Communists, but the ideals being smothered by them. An open China is a China whose social divisions would slowly become meaningful political identities. Some of those identities would be tied to liberalism and an affinity with the cultures of Japan and the West. Those are people we could work with—and they would be people that Chinese committed to a more muscular take on ‘national rejuvenation’ would have to compromise with to get things done.

[1] Dean Karalekas, Identity and Transformation: Perceptions of Civil-Military Relations in the Republic of China (Taiwan) (PhD Dissertation, National Chengchi University, 2016), p. 15-16.

[2] Tang Wenfang, “The “Surprise” of Authoritarian Resilience in China,” American Affairs Vol 2, No 1 (Spring 2018).

[3] “Upton Sinclair,” Wikiquotes (accessed 16 September 2019)

[4] Eric Gould and Esteban Klor, “Party hacks and true believers: The effect of party affiliation on political preference,” Journal of Comparative Economics Vol 47, Issue 3, (September 2019), pp. 5, 20, 24-25. The excerpts are from the prepint, which can be accessed here.

For an experimental test of the same phenomenon, see Geoffrey L Cohen, “Party over Policy: The Dominating Impact of Group Influence on Political Beliefs,” Journal of Personality and Social Psychology vol 85, no. 5 (2003), pp. 808-822.

[5] Dan Sperber and Hugo Mercier. The Enigma Of Reason (Harvard: Harvard University Press, 2019).

[6] John Tooby and Leda Cosmides, ““Groups in mind: Coalitional psychology and the roots of war and morality,” In Høgh-Olesen, Henrik, ed., Human Morality and Sociality: Evolutionary and Comparative Perspectives (New York: Palgrave Macmillan, 2010), pp.191-234; Henri Tajfel and John C Turner, “The Social Identity Theory of Intergroup Behavior,” in S. Worschel, ed., Psychology of Intergroup Relations (Chicago: Nelson-hall,1986).

For research that explicitly investigates the role of coalition psychology in modern electoral politics, see David Pietraszewski, Oliver Scott Curry, Michael Bang Petersen, Leda Cosmides, and John Tooby, “Constituents of Political Cognition: Race, Party Politics, and the Alliance Detection System,” Cognition Vol 140 (July 2015), pp. 24-39.

[7] Druckman, James N., Erik Peterson, and Rune Slothuus, ““How Elite Partisan Polarization Affects Public Opinion Formation,” American Political Science Review vol 107, iss 1 (2013), pp. 57-79; Martin Bisgaard and Rune Slothuus, “Partisan Elites as Culprits? How Party Cues Shape Partisan Perceptual Gaps“, American Journal of Political Science, vol 62, iss 2 (2018), pp. 456-469; Alexandra Guisinger and Elizabeth N. Saunders, “Mapping the Boundaries of Elite Cues: How Elites Shape Mass Opinion Across International Issues,” International Studies Quarterly 61, no. 2 (June 2017), pp. 425–41

Of course it is not always a one way transfer from leader-to-voter. Other sources that an individual may trust are just as good, and often changes in ideology begin at the grass-roots before percolating up to the top. Joshua Kertzer and Thomas Zeitzoff have a good summary of this in the lit review section of their “A Bottom‐Up Theory of Public Opinion about Foreign Policy
,” American Journal of Political Science, vol 61, iss 3 (2017).

[8] Shanto Iyengar, Gaurav Sood, and Yphtach Lelkes, “Affect, Not Ideology: A Social Identity Perspective on Polarization,” Public Opinion Quarterly 76, no. 3 (2012): 405–31