An Elephant’s Graveyard? Democracy and Development in East Asia

Abstract

Using Przeworski et al.’s paradigmatic work on democracy and development as a touchstone, this review examines East Asia’s lessons for comparative politics. It focuses particularly on the challenges that China and South-East Asia present for modernization theory, a foundation stone of political science. In most of the rich world, including north-east Asian cases of modernization such as Korea and Taiwan, economic development and democratization have tended to go hand in hand. In South-East Asia, by contrast, almost none of the expected relationships between democracy and development seems to work. The most striking anomaly of all today is China, which appears to be moving ever further away from democratic reform as it grows richer. This disjuncture between theory and practice is explored, along with other, more positive, East Asian contributions to scholarship on democracy and development.

Adam Przeworski, Michael E. Alvarez, José Antonio Cheibub and Fernando Limongi, Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990 (Cambridge: Cambridge University Press, 2000).

Few books have had the impact on comparative politics of Democracy and Development, the landmark work by Adam Przeworski and three of his younger colleagues published at the turn of the millennium. Combining broad theorizing, empirical narratives and sophisticated statistical analysis, it examined the relationship between political regimes and prosperity at a global level through the four post-war decades until 1990. Hailed as a ‘tour de force’, ‘the best defence of democracy of its generation’ and ‘a model for how research should be done’, the book is one of the most substantial inductive defences of democracy ever published. ¹

Across time and space, Przeworski and his colleagues found that democratic governance outperformed its authoritarian counterparts on almost every measure of development. While economic development did not in itself generate democracy – an important, and contested, caveat – democracies performed better in economic management, growth and investment, particularly on a per capita basis (a result of another arresting finding, that dictatorships have much higher birth and death rates than democracies). Accordingly, to quote their oft-cited conclusion, there is not ‘a shred of evidence that democracy needs to be sacrificed on the altar of development’ (Przeworski et al. 2000: 271).

The book was influential not just in the scholarly world but also as a public policy case for democratization at a time when democracy was surging worldwide. Even apparently anomalous East Asian examples such as Korea, Taiwan and Indonesia – each of which experienced rapid economic development under autocratic rule – had turned democratic by the time the book was published. But today, the success of China and other Asian autocracies increasingly challenges our understanding of the democracy–development link and some of the other political science orthodoxies embraced by Przeworski and his co-authors. Hence the title of this review; is East Asia an ‘elephant’s graveyard’ for theories of democracy and development?

WHAT WE KNOW

Comparative studies of democratization have produced two types of ‘big and bounded’ generalizations: those having nearly universal application and those applying more to a given region (Bunce 2000). The relationship between economic development and democracy is perhaps the standout example of the first category, ever since (indeed, long before) Seymour Martin Lipset (1959) identified the mutually supportive relationship between political and economic development in the Western world. But most theory-building on democratization has drawn upon Western examples, especially Europe and Latin America, rather than other regions. ² Asian cases have been much less present in this literature, and in comparative politics scholarship more generally. ³ However, as the world’s most populous, diverse and economically dynamic region, and an increasingly democratic one too, Asia is in many ways an ideal testing-ground for theories of democracy and development (Yap 2015).

Perhaps the most venerable of these is the fundamental insight of modernization theory: that democracy and development go together, in a mutually reinforcing if not directly causal way. Despite some outliers, at both a regional and a global level, democracy and development remain empirically well correlated (Norris 2008). The relationship between the two is not axiomatic: oil-rich autocracies and poorer democracies are not hard to find (one reason the Middle East has always been an outlier). But at a comparative level, the basic insight of the ‘Lipset thesis’ – ‘the more well-to-do a nation, the greater the chances that it will sustain democracy’ (Lipset 1959: 75) – continues to hold up well. While the basic ideas go back to Aristotle, Lipset was perhaps the first to argue that improvements in material and social conditions themselves generate greater support for democracy. Since then, the collapse of communism and the ‘third wave’ of democratization has seen democracy’s reach extend globally, a process which, despite recent backsliding, has (mostly) confirmed the empirical relationship between wealth and freedom across both time and space (Huntington 1991: 59–72; Norris 2008: ch. 4).

In Asia, however, the neat connection between modernization and democratization receives only partial support. India is one long-standing anomaly: as the largest but still one of the poorest democracies in the world in per capita terms, it has always been more democratic than theory would predict. But in East Asia, the focus of this article, the challenge comes more from the other direction – via the prevalence of increasingly wealthy, fast-modernizing states which in theory should be more democratic than they are in practice. Wildly successful but less-than-democratic Singapore is perhaps the best-known anomaly, but today it is China (whose economy has grown at close to 10 per cent per annum since the 1980s) that is the standout deviant case. Unlike Korea or Taiwan, contemporary China shows little sign of making a democratic transition on the back of its economic one. Indeed, China’s mix of open market economics and closed one-party politics stands as a resilient counter-example to the democracy–development link, and one which other authoritarian states appear keen to emulate (Kurlantzick 2013).

East Asia’s challenge to modernization theory is all the more important because of another important if little-noticed trend: China notwithstanding, the region is one of the few to register an increase in aggregate democracy levels over the past decade. More East Asians today live in democracies than ever before, the vast majority in developing countries. New or restored democracies, ranging from Mongolia in the north to East Timor in the south, have joined Korea, Taiwan, the Philippines and Indonesia (as well as the established democracy of Japan, and now potentially Burma too) as countries where governments are chosen and changed via the electoral process. Despite ‘backsliding’ in once-promising cases such as Malaysia and Thailand, and the institutionalization of ‘low quality’ democracy in the Philippines and Indonesia, this makes Asia something of a bright spot compared with the democratic recession taking place elsewhere (Diamond and Plattner 2015). As Freedom House has observed,

Over the past five years, the Asia-Pacific region has been the only one to record steady gains in political rights and civil liberties as measured by Freedom House. Although it is home to China, where over half the world’s Not Free population lives, and North Korea, the least free country in the world, a number of Asia-Pacific countries have made impressive gains in the institutions of electoral democracy – elections, political parties, pluralism – and in freedom of association. ⁴

All of this raises challenges for studies of democracy and development – and the eponymous book which provides a focus for this article. A model of both parsimony and precision, the book remains perhaps the single most authoritative test of the many empirical inquiries into the relationship between wealth and democracy. In defining, evaluating and extending the investigation of democratic modernization theory over time and space, Przeworski and his co-authors confirmed much of the scholarly orthodoxy: most importantly, that democracy is sustained by development – to the point that once a basic level of prosperity is reached, democracy becomes almost inviolable, once it is established. They also investigated other key questions: whether economic development is conducive to particular regime types, whether democracy fosters or hinders political welfare, democracy’s institutional scaffolding, and the broader social and economic preconditions for democracy.

Their results were, for the most part, a striking vindication of democracy’s strengths. At a global level, democracies had healthier, longer-lived populations, did a better job than autocracies of maintaining social peace and providing public goods, and were about the same in relation to long-run growth and investment. Economic growth and investment, demographic changes and fertility levels were all affected by regime type – again, with democracies superior on almost all measures. Democracy and development were thus intimately connected: once a country is minimally wealthy, they found, democracy was almost certain to survive ‘come hell or high water’. In short, there was no trade-off between democracy and development.

This was much more than a restatement of the Lipset thesis with better data. According to the modernization school, the combination of economic development, industrialization and education creates a society whose citizens are more assertive and politically aware. An educated and critical middle class which can act as a counterforce to authoritarian rule is crucial to democracy (Bollen and Jackman 1985; Burkhart and Lewis-Beck 1994; Huntington 1991; Lipset 1959). Przeworski and his colleagues confirmed much of this orthodoxy, but with a twist: they found that the increase in GDP per capita which presages the growth of a robust middle class does not in itself stimulate a move towards democratic governance, but rather that rising income levels reduce the risk of democratic backsliding in already-democratic countries.

This caveat is particularly relevant to East Asia, which contains the most important cases of sustained economic development under autocratic rule. As Francis Fukuyama (2013: 5) notes, ‘all the recent examples of successful authoritarian modernisation cluster in East Asia rather than other parts of the world . . . most of the arguments in favour of sequencing economic growth and law before a democratic opening are based on East Asian models’. This highlights one theme of this review article: the ongoing, and indeed increasing, divergence of modernization theory from democratic reality in East Asia. Given this, the autocratic resilience of economic leviathans such as China may reshape not just the global order but also scholarly assumptions about the link between democracy and development.

These assumptions are deep-rooted. When Lipset wrote in the late 1950s, almost all genuine democracies were Western (with Japan and India the main exceptions). Since then, democracy’s reach has expanded to almost every region. One of the strengths of Democracy and Development is that it sought to include the kind of variables most relevant to the experience of new democracies in the developing world: the institutions most likely to help or harm democratic persistence, the effects of ethnic and religious diversity, the impacts of political instability, labour participation and population growth, the timing and duration of democratic transitions, and so on. Published at the height of optimism about democracy’s munificent impacts and before the 9/11 terrorist attacks, Democracy and Development thus represented perhaps the high-water mark of a multi-decade effort examining the instrumental benefits of democracy.

Przeworski et al.’s findings were widely circulated and boosted democracy’s proponents, including the governments of the United States and other Western countries. In Asia, however, things were different. During the 1950s and 1960s, almost all of Asia’s fastest-growing economies were non-democracies, with the success of the Asian ‘tigers’ – Korea, Taiwan, Malaysia, Singapore – particularly striking. The most rapid growth in the tiger economies took place under authoritarian rule, and indeed contributed to the eventual democratization of Korea and Taiwan (the only two dictatorships that started with annual per capita incomes under $1,000 in 1950 but were exceeding $5,000 by 1990). Observers of East Asia’s economic miracle, not least prominent Asian leaders such as Singapore’s Lee Kwan Yew and Malaysia’s Mahathir Mohamad, argued that democracy was not a natural form of government for the region (Thompson 2001). Some kind of benevolent authoritarianism was better for maintaining social peace, they argued, allowing meritocratic elite selection, stable long-term policy settings and faster economic growth.

Today these smaller Asian ‘tigers’ have been overshadowed by the economic might of China – the rapid rise of which has utterly transformed the global economy and, it appears, the future world order. China constitutes the single most dramatic poverty alleviation exercise in history, with several hundred million people having been lifted from extreme poverty, especially in rural areas, and many millions of others joining a fast-growing urban middle class. From 1978 to 2002, annual per capita GDP growth averaged 9.7 per cent, with a 10-fold increase in per capita GDP over the same period. Since then, growth has slowed to an average of 7 per cent, still a remarkable number for such a large country. Over this period, rural migration has seen the emergence of huge cities, mostly along the Pacific rim and immediate hinterland, which have increasingly become China’s trading window to the world and the key drivers of China’s economic and perhaps political future.

As predicted by modernization theory, the increasingly wealthy, mobile and aspirational nature of Chinese society saw growing demands for greater democracy over this period of rapid development, peaking with the 1989 Tiananmen student demonstrations in Beijing and hundreds of smaller protests around the country. The violent repression and rapid reimposition of central party control which followed postponed but did not permanently roll back either economic or social liberalization. It did, however, put sharp restrictions on incipient moves towards mass political participation and even more so any possibility of contestation for political power outside of the Chinese Community Party (CCP) itself. While the party has now largely abandoned the pretence of Marxist-Leninist ideology, it remains ruthless in its adherence to Marxist-Leninist political control. As long as the legitimacy of one-party rule is not questioned, however, citizens are relatively free to complain about other core problems of governance, environmental degradation and corruption.

As a result, China today stands out as exemplar of rapid economic development under autocratic government. In late 2014 the International Monetary Fund (IMF) announced that on purchasing power parity (ppp) terms China had overtaken the United States to become the world’s largest state-based economy, a transformation predominantly driven by export-led growth which has created a large emergent urban middle class. China’s 2015 per capita ppp income was $13,801 (on a par with Serbia and Colombia), with a projected 2020 level of over $20,000 – higher than the forecast for Brazil and Iran, and well above Poland and South Africa when they made their democratic transitions. Income levels in Shanghai were almost double this national average, as booming coastal cities prospered at the expense of the countryside.

Given what we know, or think we know, about the role of the middle classes – ‘no middle class, no democracy’ (Moore 1966) – this should be highly significant for democracy too. However, China’s burgeoning middle classes appear to have largely acquiesced, and indeed supported, the reality of economic but not political liberalization. Jie Chen (2014) has explored the attitudinal and behavioural orientation of China’s nascent middle class to democracy in three major Chinese cities, Beijing, Chengdu and Xi’an. He found that the new Chinese middle class tends to be more loyal to the CCP and less supportive of democratic values and institutions than their income levels would predict, particularly for those with institutional ties with the state – making the middle class more often an opponent rather than a supporter of democratic changes. Andrew Nathan (2016) recently came to a similar conclusion, arguing that it is the institutional configuration of China’s one-party system, state-dominated economy that explains the apparent ‘exceptionalism’ of attitudes towards democracy in the Chinese middle class.

Nonetheless, claims that China will fall into line with modernization theory are not hard to find. Bruce Gilley, who has long forecast a democratic transition in China, recently argued that ‘as China’s development levels approach $10,000 by 2020, the pressures for change will be immense . . . the “markets and modernization” that brought democratic change to the rest of Asia are well advanced in China’ (Gilley 2014: 140). An earlier prediction along the same lines was made by Stanford’s Henry Rowen, who predicted that China would begin its own democratic transition in 2015:

When will China become a democracy? The answer is around the year 2015 . . . [when] China’s per capita GDP will be between $7,000 and $8,000 (in 1995 dollars) . . . Several scholars have suggested that the transition to stable democracy correlates with mean incomes between $5,000 and $6,000, and becomes impregnable at the $7,000 level. There is a compelling logic behind the statistical relationship. (Rowen 1996: 61, 67)

A decade later, favourably citing Democracy and Development as a supporting source, Rowen doubled-down on that prediction, estimating that even if China’s GDP growth slowed in 2015, it would still by 2025 be at about $5,500 per person ($12,000 ppp), by which time it would have made a full transition to democracy (Rowen 2007). Another who made 2015 the starting point was Zhengxu Wang (2008: 198–202), who used survey results of changing social values to argue that China’s citizens would begin to demand democracy in large numbers between 2015 and 2020.

Well, I am writing this review article in 2016 and, much as I would like to see it, there are few if any signs of these predictions bearing fruit. The income level at which Rowen predicted China would make a ‘full transition to democracy’ was actually reached in 2012, around the time that Xi Jinping assumed the leadership of the Communist Party and the presidency of the People’s Republic. Since then, ‘Xi who must be obeyed’ has launched a major crackdown on political freedoms and the rule of law, which has seen the reversal or abandonment of many (small) steps towards greater openness – in areas such as village elections, legal rights and media freedoms – taken by his predecessors. Today, China appears to be both richer and further away from democracy than ever, with informed observers increasingly gloomy about the trajectory of even modest democratic reforms started by more liberal leaders such as Jiang Zemin (Fewsmith 2013).

In one way, however, China’s trajectory supports Democracy and Development’s refinement of modernization theory: the contention that an increase in GDP per capita does not foster democratic transitions, but rather that rising income levels reduce the likelihood that democracies will revert to authoritarian rule. Przeworski and his colleagues thought that democracies can emerge anywhere and at any level of development, but are much more likely to survive in countries with a minimal level of development (per capita incomes over $4,000 at the time at which they were writing). Once established in a developed country (essentially, at per capita income levels over $6,000), ‘the miracle occurs’ and democracy endures regardless of exogenous challenges.

This delinking of economic development from democratic transitions remains contentious and was critiqued by a number of later studies, many of which argued that there is indeed an income threshold at which we should expect transitions to occur. Boix and Stokes (2003) challenged what they typified as ‘endogenous democratization’ on both theoretical and empirical grounds. Epstein et al. (2006) showed that higher incomes per capita did, in fact, significantly increase the likelihood of democratic transitions once ‘partial’ or ‘unconsolidated’ democracies were properly classified. Given that this middle category between democracies and autocracies encompasses the very states for which the findings of modernization theory are most relevant, this was a significant caveat, supporting earlier large-N work by scholars such as Barro (1999), who argued that prospects for democracy rise in step with the share of income held by the middle class. The reality of China’s resilient autocracy in the face of spectacular economic growth is thus more of an anomaly than a casual reading of Przeworski et al. may suggest.

But China is not East Asia’s only challenge to theories of democracy and development. Across a number of other areas of democratic modernization theory – including the literature on democratic transitions, the role of the middle classes, the impact of ethnic and religious diversity, and the choice of political institutions – the region’s experience consistently contradicts the scholarly orthodoxy. South-East Asia in particular truly seems to be a graveyard for some of the largest lumbering theories of our discipline, with not just modernization theory but much of the literature on democratic transitions and preconditions finding little support.

SOUTH-EAST ASIA’S CHALLENGE

South-East Asia is a useful test-ground for assessing the democracy–development nexus, with great variation on both measures. The region contains some of the world’s richest (Singapore) and poorest (East Timor) states, as well as a full spread of regime types: electoral democracies in East Timor, the Philippines and Indonesia, soft-authoritarian ‘quasi-democracies’ in Singapore and Malaysia, resilient Communist regimes in Laos and Vietnam, a military junta in Thailand, an absolute monarchy in Brunei, and even an ongoing democratic transition in Burma. Strikingly, democracy is weak or absent in the wealthiest states of Singapore, Brunei and Malaysia, but present, to varying degrees, in most of the poorest ones, Indonesia, Timor-Leste, the Philippines and now Burma too – which also rank low on aggregate measures of educational attainment, literacy, maternal health and other human development indicators.

Singapore remains a major anomaly for democratic modernization theory. The former entrepôt-turned-city-state at the tip of the Malay peninsula is one of the great postcolonial success stories in the world today. Easily South-East Asia’s richest and most developed state, Singapore’s per capita GDP of over US$85,000 is the world’s third-highest in ppp terms. As a long-standing soft-authoritarian ‘competitive authoritarian’ regime under the rule of the same party since independence, Singapore is, in Larry Diamond’s (2012: 7) apt formulation, ‘the most economically developed non-democracy in the history of the world’. The Singapore model relies on a combination of strong rule of law with limits on political competition in a regime model usually characterized as a quasi-democracy or soft autocracy (Ortmann 2015). This provides for regular elections (which the incumbent People’s Action Party invariably wins, with some opposition permitted), but with judicially enforced defamation action routinely used to cow too-vocal critics of the government.

Malaysia too uses the courts and particularly the Sedition Act of 1948 to maintain incumbent control. With a 2015 per capita GDP of over $25,000 in ppp terms, Malaysia has a higher income than most third-wave countries did when they made their transitions to democracy. As Diamond notes,

Malaysia’s score on the UNDP’s Human Development Index (HDI) – which, in measuring not only per capita income but also levels of health and education, is arguably a truer measure of development – is now significantly higher than the levels in Brazil, Chile, Mexico, and even Hungary, Poland, and Ukraine when they made their respective transitions to democracy. From the standpoint of modernization theory, then, Malaysia is also ripe for a democratic transition. (Diamond 2012: 12)

But theory has again been confounded. As in China, a hoped-for reformist leader in Prime Minister Najib Razak has proved to be anything but, clamping down on civil liberties and jailing opponents to maintain the regime’s grip on power as his popularity has crumbled.

Both Singapore and Malaysia thus represent apparently resilient counter-cases in which high levels of human development and government capacity are combined with illiberal, quasi-democratic politics – and which, in Malaysia, is becoming more repressive over time. As standout examples of ‘competitive authoritarianism’ (Levitsky and Way 2010), both allow opposition contestation in elections without ever experiencing a change of government. Rather, dominant governing parties use electoral gerrymandering, compliant judiciaries and colonial-era internal security acts to buttress their incumbency. Critics of the government are routinely prosecuted in defamation actions which stifle freedom of expression and sap the energy of even relatively benign opposition voices.

While the resilience of such cases undermines the claims of democratic modernization theory, other Asian examples suggest alternative routes to democratic development. In some cases, what were once Asia’s strongest authoritarian parties embraced democratization as a means to stay in power. By conceding democratic openings to rising actors at times of high political uncertainty, adept ruling parties can increase their prospects in the post-authoritarian polity. Dan Slater and Joe Wong argue that this scenario has played out, albeit in very different ways, in at least three Asian developmental states where ruling parties have democratized from positions of strength: Taiwan, South Korea and (more questionably) Indonesia. In such cases, the declining popularity of long-ruling hegemonic parties saw them open up space for opposition movements, but also allowed them a route back to power (Slater and Wong 2013).

Other scholars see ‘electorally led’ democratization as a more viable model of transition from authoritarian rule, claiming that holding repeated elections over time, even in less than open political environments, can itself promote democratic prospects (Lindberg 2009). While the idea of ‘democratization by elections’ is undermined by Singapore and Malaysia, the recent experience of Burma does provide some support. There, the process of reform which culminated in the 2015 transition was clearly scaffolded by elections – not just the 2015 polls but also prior, partial and less than competitive elections in 2010 and 2012, which while a long way from free and fair did, it is now clear, contribute to meaningful democratic progress. The political openings that have occurred since show how flawed but partially competitive elections can themselves lead to greater democratization.

The tightly controlled 2010 elections, the first in Burma for 20 years, followed an incumbent-led process of constitutional change which guaranteed, among other things, a permanent role for the armed forces in Burma’s legislature and government. These elections were only marginally competitive, with the opposition National League for Democracy (NLD) barred from competing and their leader, Aung San Suu Kyi, still under house arrest. The ruling junta’s Union Solidarity and Development Party won a tainted victory, but these elections also brought new ethnic and minority parties into parliament for the first time and kept momentum for further progress. By-elections held in 2012, following further liberalization, were dominated by the NLD and saw Suu Kyi elected for the first time. This made more credible the government’s commitment to hold national elections in 2015 and encouraged opposition groups to commit to the electoral process rather than pursue extra-constitutional avenues.

The November 2015 elections, when they came, proved to be a watershed. Amidst intense international interest and expectations, the NLD secured an overwhelming victory, winning 80 per cent of available seats and catapulting a party of ex-political prisoners and protesters into government. As Burma’s first genuinely open elections in over 50 years, they presaged one of the most dramatic shifts from autocratic to democratic rule seen anywhere in East Asia. But even with its overwhelming victory, the NLD still has to cope with numerous constitutional impediments bequeathed by the former junta – including the military’s ongoing grip on a quarter of legislative seats, contrived citizenship restrictions on the presidency (aimed directly at Suu Kyi) and a cumbersome mixed presidential–military governance structure.

For political scientists who believe in rational actor models of politics, Burma also offers an unexplained puzzle: why did the incumbent regime agree to free elections which they surely knew they would lose? And why did the military also acquiesce to the diminution of power inherent in the change of government? After all, we assume that politicians are rational office-seekers, not turkeys voting for Christmas. But the regime’s willingness to hold free elections has seen most of their recent leadership now out of office. Even taking into account international pressure from not just Western states but also their ASEAN neighbours, this appears to be a rare case of self-abnegating leadership from a departing autocratic regime.

This flouting of rational actor models is not the least of the challenges South-East Asia presents to democratic theory. Take, for instance, the question of ethnic, religious, linguistic or other forms of social diversity, which have long been considered a challenge for successful democratization – and indeed broader modernization too. Democracy is often held to be more likely in smaller, more homogeneous states than in large, diverse ones (Dahl and Tufte 1973). It is thought to be particularly problematic in highly diverse societies with deep ethnic or cultural divisions (Rabushka and Shepsle 1972), and less than compatible with some kinds of religions, particularly Islam, due to the difficulty in separating church and state under Islamic law (Barro 1999; Huntington 1991).

While Przeworski and his colleagues found supporting evidence for most of these democratic ‘preconditions’ globally, they fail dismally in South-East Asia. The region’s standout democracy, Indonesia, is a Muslim-majority country of over 250 million people, spread over thousands of islands with hundreds of different ethno-linguistic groups. Like its democratic neighbours, the Philippines and East Timor, it combines national-level electoral democracy with deep social and religious divisions, widespread poverty and acute challenges of national governance. Burma too is an ethnic kaleidoscope, with seven ethnic states, over 100 official ethnic minorities and a multitude of identity schisms (including an increasingly deep divide between the Buddhist majority and the Muslim minority, particularly the Rohinga), and a long and ongoing history of minority ethnic insurgency.

These four South-East Asian democracies thus counter much conventional political science wisdom. All are ethnically fragmented, despite the scholarly consensus that such divisions make democracy more difficult. All are under-developed, with poor HDI indicators and per capita GDP well below that of China, Thailand and even Mongolia. All combine presidential or semi-presidential systems of government, despite the well-known problems of this model (Gerring et al. 2005; Linz 1990). Indeed, most of East Asia’s newest democracies (Taiwan, Mongolia, East Timor and now Burma) have adopted variants of semi-presidentialism, despite what Elgie and Moestrup (2007: 237) characterize as the ‘consensus that young democracies should avoid this type of institutional arrangement as the in-built conflict between president and prime minister may damage the prospects for successful democratization’ – highlighting again the region’s divergence from the expectations of the political science literature. ⁵

Perhaps the most important anomaly of all concerns the attitude of the region’s middle classes towards democracy. Experts on Singapore and Malaysia have long decried the quiescent and apathetic attitude displayed by most of the middle class in the face of flagrant rigging of the democratic rulebook by incumbents (Case 2002). Even in genuine democracies, there is increasing evidence that South-East Asia’s middle classes are less committed to democracy than political theory would have us believe (Dore et al. 2014). In Thailand, for instance, the Bangkok-based elite have confounded democratic theory by showing themselves to be actively hostile to majority rule, both before and after the military coup of 2006 (Jones 2005). Particularly since Thaksin Shinawatra rewrote the rule book for winning elected office, the Thai middle classes’ distaste for majority rule has been on full display, at one stage occupying the main airport to campaign against the elected government and demonstrating repeatedly against one-man, one-vote democracy. Combined with evidence from China discussed earlier, this raises real questions for assumptions that the middle class will promote or at least support democracy. Increasingly in East Asia, this does not appear to be consistent with the facts.

A DIFFERENT APPROACH

Taken together, then, it is clear that East Asia challenges some well-worn scholarly findings about democracy and development. So far, so disappointing for political science. But these are mostly negative findings, dogs that did not bark. What positive contributions can East Asia make to our theoretical and empirical understanding of the democracy–development relationship?

One emerging stream of the scholarly literature that does attempt to grapple directly with this question can be found at the intersection of political science and economics, again with the South-East Asian experience central. In this literature, political regime type appears to be less important to economic outcomes than the particular political institutions adopted by new democracies. Because they influence both the electoral incentives facing politicians and the broader shape of a country’s party system, these institutional choices – electoral systems, party laws, executive structures and the like – can directly influence the public goods and policies needed ‘to get growth going’ (Rock 2013).

Some economists argue that countries with centralized executives, majoritarian elections and integrated party systems will both theoretically and empirically economically out-perform those with proportional elections and more fragmented executive and party systems (Persson and Tabellini 2000, 2005). ⁶ Likewise, the ‘developmental state’ literature on the optimum political arrangements for economic reform in new democracies cite the benefits of aggregative political institutions, majoritarian electoral processes and broad ‘catch-all’ parties or coalitions (Haggard and Kaufman 1995). My own work has shown how many East Asian states have followed this path as part of a deliberate exercise in political engineering (Reilly 2006). Indeed, the combination of plurality or mixed-member majoritarian electoral systems, limits on party fragmentation, and stability-inducing executive formation rules is so strong that it can be said to constitute a distinctive ‘Asian model’ of democratic design (Reilly 2007).

What have been the policy effects of these reforms? With both elite and public opinion in Asia focused on the need for ‘pro-development’ politics, active efforts to promote more centrist and stable government have been widespread, via electoral reform (as in Japan), constitutional change (as in Thailand or the Philippines) or political party laws (as in Indonesia). Such reforms typically seek to encourage more stable and cohesive politics and better governance than was previously the case. In Japan, for example, former Liberal Democratic Party (LDP) Secretary General Ichiro Ozawa argued that reformed electoral laws and the development of a two-party system was necessary for the country’s long-term survival (Sakamoto 1999). Similarly, former Indonesian president Megawati and other national leaders argued that limitations on party proliferation were necessary to protect Indonesia’s national integrity and prospects for development (Hillman 2010). Thai reformers too saw political reform as key to their country’s economic revival following the devastating 1997 Asian economic crisis (MacIntyre 2003).

What have been the effects of this shift towards an ‘Asian model’ of majoritarian electoral democracy? At the country level, studies have highlighted impacts in the shape of more programmatic policy offerings in cases as diverse as Japan (where electoral reform has led to a decline in particularism), Taiwan (looking at the Kuomintang’s median voter strategy on the national identity question), Thailand (in relation to health policy) and, most recently, in relation to pro-poor initiatives in Indonesia as well. ⁷ These case studies lend credence to the theoretical insight that reforms to electoral and party systems can have public policy consequences with flow-on impacts for economic development, as incentives for political elites to deliver particularistic policies are supplanted by the need to deliver broader public goods to the electorate as a whole.

What about the specific issue of democracy and development? The most comprehensive comparative investigation of the economic impacts of democratic reform has been conducted by Rock (2013), who tested the implications of political reform across 11 East Asian democratic and autocratic polities — Cambodia, China, Indonesia, South Korea, Laos, Malaysia, the Philippines, Singapore, Taiwan, Thailand and Vietnam. Using regression analysis, he found that there was no trade-off between growth and democracy for those newly democratic states which adopted the ‘Asian model’ of aggregative and majoritarian institutions, in comparison to the growth rates of the autocracies such as China, Laos and Vietnam or semi-democracies such as Singapore and Malaysia. As a result, Rock (2013: 1) found that ‘the contribution to growth from majoritarian democratic institutions in East Asia is as large as that from the region’s developmentally oriented authoritarian governments’.

The specific mixture of institutions Rock investigated were those I identified as key to political engineering in Asia: majoritarian electoral systems, stable and insulated executives and consolidated party systems. Rock (2013: 18) found that ‘when it comes to the impact of political institutions on growth, the devil really is in the details’: it was these kinds of reforms, more than macro-level regime distinctions, that were key to politicians appealing to broader groups of the electorate and delivering growth-promoting policies. Those democratizing states which adopted more majoritarian reforms were able to sustain the pro-growth coalitions initially developed by autocratic governments, whereas other institutional configurations (including, surprisingly, semi-democracy and measures of the quality of democracy) were less successful – suggesting that ‘East Asia’s majoritarian democracies are as growth enhancing as the region’s developmental autocracies’ (Rock 2013: 14).

Asia’s political engineering strategies were animated, above all, by the search for political stability and economic growth, making their experience particularly important for the democracy–development debate. Small variations in democratic forms can and did have big policy consequences. Indonesia’s post-Suharto party reforms, for instance, advantaged large nationally focused parties of government and blocked smaller regional parties – a move which ‘prevented Indonesia’s national institutions from becoming a battleground for organized regional and ethnic interests, and limited the extent to which local institutions can be captured by ethnically exclusivist movements’ (Aspinall 2010: 26). Under such circumstances, it is too costly and insufficiently electorally rewarding to try to deliver private or club goods to localized bases of voter support. Rather, a more efficient strategy is to deliver broad-based public goods – in the shape of policy, not pork – to broad social groups as a means of securing voter support (Keefer 2011).

CONCLUSION

The conclusion to Democracy and Development observed that ‘because China and India add up to one-half of us, they deserve special attention’ (Przeworski et al. 2000: 272), noting that despite one being a democracy and the other a dictatorship, they had almost equivalent per capita income levels in 1990, the final year of that study. Today, 25 years later, China’s income levels are more than twice India’s, and its system of competitive markets and closed politics appears far more secure than it did in 1990 too, raising difficult questions for those who believe that democracy and development will run in parallel.

Nonetheless, China’s leaders clearly do not feel secure and are looking for ways to further entrench their system. In particular, Chinese officials have closely studied the success of Singapore, the only country in the region to achieve advanced economic industrialization without undergoing substantial political liberalization. The key lesson that China seeks to learn is how to combine authoritarian rule with ‘good governance’ under some form of meritocratic one-party rule. In reality, as Ortmann and Thompson (2016) observe, the Chinese misread the Singapore experience, focusing on its more repressive elements without sufficient attention to the rule of law and free (if not fair) elections in allowing most if not all liberties to be protected and some opposition to the ruling party, both impossible in China.

The extreme differentials between a continent-sized ancient empire of over a billion people and a small city-state just 50 years old also makes genuine lesson-learning difficult. Unlikely as it may seem, a better South-East Asian comparison for China may be Indonesia – a large, complex and unruly country which nevertheless managed to make a successful transition from autocratic to democratic rule without losing its territorial integrity (the departure of East Timor aside), experiencing social revolution (although Suharto’s fall in 1998 did see large-scale riots in Jakarta and elsewhere) or upending the constitution (which has been successively reformed rather than being replaced). Indonesia’s reformers also managed to ease the military out of parliament, where they had been given a quarter of all seats by former General Suharto, and back to barracks, where they remain.

With its unitary but decentralized presidential model, Indonesia’s political architecture may be relevant for a more democratic China at some stage in the future. This is particularly the case in terms of its management of political pluralism. Any move towards a more democratic China would confront similar issues to those Indonesia faced in 1999: how to create a competitive party system but avoid the dangers of fragmentation, ethnic mobilization and separatism. As in China, order, stability and national integrity were paramount concerns for Indonesia’s political leaders. Their choice to allow party formation but also restrict electoral competition only to those parties able to demonstrate a nationwide organizational basis across Indonesia’s 3,000-mile long archipelago made potentially separatist parties unviable. Combined with laws requiring broad cross-national support for presidential candidates, and an electoral system that helps larger parties more than small ones, the effects of these ‘centripetal’ rules has been to limit political fragmentation and create a more nationally focused political system than would otherwise have been the case (Reilly 2016).

For now, Xi Jinping has dismissed the whole concept of competitive elections as ‘Western ideas’ (although he has also committed to the goal of ‘democracy with Chinese characteristics’ by 2049, a date both weirdly specific and utterly distant). Any moves in a more democratic direction, even reinvigoration of the now-stalled political openings at the rural and village level, would of course be welcome. More ambitious steps in the Singapore direction of limited electoral competition would be a huge advance, providing a safety valve for social unrest and injecting some real, if conscribed, political accountability. There is no sign of anything like this happening at present – but the lesson from most, if not all, of the democratic modernization literature is that China will, sooner rather than later, need some concessions towards liberalism if it is to continue to develop. However, it is also true that such statements have been made many times before and, given Asia’s habit of challenging key political science orthodoxies, it will not surprise if China continues to do so.