Chiang Ching-kuo's Taiwan: From White Terror to Sunflower Democracy

Thesis

Chiang Ching-kuo’s period of effective and then formal leadership of Taiwan — as premier from 1972 and as president from 1978 until his death in January 1988 — is the Asian case in which developmental-authoritarian transformation and democratic transition were most nearly integrated as parts of a single political trajectory. Taiwan under the Kuomintang (KMT) government of Chiang Kai-shek (1949–1975) had been an authoritarian one-party state under martial law from 1949, with the White Terror producing an estimated 3,000 to 4,000 executions and tens of thousands of imprisonments through the 1950s and 1960s. Chiang Ching-kuo, the elder Chiang’s son, inherited this framework in 1975 and — over the following thirteen years — presided over Taiwan’s transition from developmental authoritarianism to democratic constitutional politics without regime collapse, civil war, or comprehensive institutional discontinuity. The 1987 lifting of martial law, the 1991 constitutional reforms, the 1996 direct presidential election, the 2000 alternation of power to the Democratic Progressive Party under Chen Shui-bian, and the subsequent alternations between the KMT and DPP through 2024 constitute the most successful managed transition from developmental authoritarianism to constitutional democracy of any Asian case.

The article’s argument is that Taiwan is the case in which the Asian model’s development-then-democratisation sequencing worked as its more sympathetic theoretical defenders (Meles Zenawi’s revolutionary democracy included) said it would. The developmental transformation was accomplished under authoritarian direction. The middle class the transformation produced generated the political demand for democratic reform. The authoritarian leadership, faced with that demand and calculating (correctly) that the transformation would survive democratisation, chose to transition rather than repress. The subsequent democratic period has produced continuing economic development, national security under continuing PRC threat, and a distinctive Taiwanese national identity that has emerged from the KMT’s original Chinese-nationalist framework through progressive political contestation. The costs of the transformation — the White Terror deaths, the martial law period’s political imprisonments, the 228 Incident of February 1947 in which the KMT’s initial post-liberation administration killed an estimated 18,000–28,000 Taiwanese — were paid by the pre-Chiang-Ching-kuo period. The returns to Taiwan since 1988 are what Ethiopia’s regimes have never produced.

The Kuomintang inheritance

The Kuomintang arrived in Taiwan in 1945 as the Chinese national government that had defeated Japan (with substantial American assistance) and was resuming administration of the island after fifty years of Japanese colonial rule. The 228 Incident of February 1947, in which the KMT’s initial administrators suppressed island-wide protests against corruption and mismanagement, established the pattern of KMT-Taiwanese relations that would persist through the subsequent decades: Chinese-mainland KMT elites governing a Taiwanese population from which they were culturally, linguistically, and politically distinct. The KMT’s 1949 flight to Taiwan after defeat in the Chinese Civil War brought approximately 1.2 million mainland Chinese to an island whose existing population was approximately 6 million. The 1949 KMT government imposed martial law on 20 May 1949, which would remain in force until 15 July 1987 — nearly forty years, one of the longest continuous martial-law periods in modern history.¹

Under Chiang Kai-shek’s leadership (1949–1975), Taiwan operated as an authoritarian one-party state with the KMT holding all significant political positions, mainland Chinese émigrés dominating the political and military leadership, and the native Taiwanese population subject to political restrictions that included, at various points, prohibition on the use of the Taiwanese Hokkien language in schools, restriction on Taiwanese political organisation, and systematic surveillance and repression through the security apparatus. The White Terror of the 1950s and 1960s produced its documented executions and imprisonments. The Formosa Incident of December 1979 (an anti-government demonstration in Kaohsiung that produced mass arrests and prosecutions) was one of the late episodes of the pattern.²

The pre-Chiang-Ching-kuo period accomplished substantial developmental work despite the political repression. Land reform (1949–1953) eliminated Taiwanese tenancy on the pre-war scale, destroying the political power of the pre-existing Taiwanese landowning class and providing the KMT with a mass smallholder rural base. Universal primary education was achieved by the 1960s. Industrial policy through the Council for Economic Planning and Development, established in various institutional forms from the 1950s, directed credit and technical assistance toward export-oriented sectors. Export processing zones established from 1965 (beginning with Kaohsiung) attracted foreign direct investment. The Taiwanese economy grew at rates comparable to Korea’s through the 1960s and 1970s.

Chiang Ching-kuo’s programme

Chiang Ching-kuo assumed effective leadership of Taiwan progressively through the 1970s: as vice premier from 1969, as premier from 1972 (during his father’s final illness), and as president from 1978 after his father’s death in April 1975 (with an intervening presidency of Yen Chia-kan). His formal presidency ran from 1978 until his death in January 1988. The programme he pursued had several distinctive features:

Ten Major Construction Projects (1974–1979). Chiang launched a large-scale infrastructure programme covering railways, highways, ports, steel mills, shipyards, petrochemical plants, and nuclear power. The programme was Keynesian in its immediate response to the 1973 oil crisis, developmental in its infrastructure logic, and identity-constructing in its framing as national achievement. The projects were completed on schedule and became the physical basis for Taiwan’s subsequent industrial development.³

Semiconductor and electronics industry development. The Industrial Technology Research Institute (ITRI), established in 1973, coordinated Taiwan’s entry into semiconductor manufacturing. Taiwan Semiconductor Manufacturing Company (TSMC) was founded in 1987 with substantial government backing and would, over the subsequent thirty-five years, become the world’s leading semiconductor foundry. The electronics industry more broadly — Acer, ASUS, Foxconn, and hundreds of smaller firms — became by the 2010s a globally strategic sector with implications for the geopolitical significance of Taiwan itself.

Meritocratic bureaucracy consolidation. The technocratic ministries — the Council for Economic Planning, the Ministry of Economic Affairs, the Ministry of Finance, the Executive Yuan’s research offices — operated with substantial insulation from KMT political direction and with recruitment through competitive civil-service examinations. The Taiwanese technocracy of the 1970s and 1980s produced the sectoral analyses, the industrial-policy plans, and the infrastructural coordination that the developmental transformation required.

Progressive Taiwanisation of the KMT. Chiang Ching-kuo, unlike his father, promoted native Taiwanese into the KMT leadership. Lee Teng-hui, a native Taiwanese who would succeed Chiang as president in 1988, was appointed vice-president in 1984 as part of this deliberate ethnicisation of the ruling party. The move was politically calculated (it broadened the KMT’s political base within Taiwan) but was also identity-constructing (it began the redefinition of KMT-governed Taiwan as an inclusive Taiwanese political community rather than an exclusive Chinese-émigré project).

Progressive political liberalisation. From the early 1980s, Chiang initiated a controlled loosening of political restrictions. Opposition political activity, while formally illegal outside the KMT, was tolerated at levels that increased through the decade. The Democratic Progressive Party was founded in September 1986; the KMT government did not arrest its founders. The martial law was lifted in July 1987. The ban on new newspapers was lifted in 1988. The ban on political parties was lifted in 1989. Each of these moves was made under Chiang’s leadership or in the immediate transition period after his death, and each reflected the calculation that the KMT could survive democratic competition better than it could survive continued repression.

The 1987–2000 transition

The 1987 lifting of martial law began the formal transition to constitutional democracy that would run through the 2000 election. Its stages:

1987. Martial law lifted on 15 July. The Democratic Progressive Party operating openly. Substantial press freedom emerging.

1988. Chiang Ching-kuo died in January. Lee Teng-hui, the native Taiwanese vice-president, succeeded him and continued the reform trajectory.

1991. Constitutional reforms limited the National Assembly’s terms and permitted direct elections to that body from 1991. The “1947 constitution” — which had maintained fictional representation of mainland Chinese constituencies from before the 1949 KMT retreat — was substantially amended to reflect the Taiwan-only political reality.

1992. First fully-competitive legislative elections. KMT retained majority; DPP established as major opposition party.

1996. First direct presidential election. Lee Teng-hui won as KMT candidate. This was the first direct election of a Taiwanese head of state and marked the formal completion of the transition to constitutional democracy.

2000. Chen Shui-bian of the DPP won the presidential election. This was the first alternation of political power in Taiwan’s history and the confirmation that the democratic transition had produced durable constitutional politics rather than continued KMT dominance under democratic forms.

The transition was gradual, managed, and consensual. It did not produce constitutional collapse; it did not produce ethnic conflict between mainland Chinese émigré and native Taiwanese populations; it did not destroy the developmental base the KMT had built. The KMT lost the 2000 election and returned to opposition; the DPP governed and was subsequently voted out in 2008; alternations have continued through 2024. The political framework has been stable through changes of governing party.

The identity emergence

Taiwan’s national identity has emerged, through the democratic period, as distinctively Taiwanese rather than either Chinese-KMT or ethnic-Taiwanese in the pre-1949 sense. The specific features:

Democratic-constitutional identity content. Taiwan is one of Asia’s most democratic states, and Taiwanese identity has incorporated the democratic-constitutional content as one of its defining features. Public opinion polls since the 2000s consistently show that Taiwanese respondents regard democratic politics, civil liberties, and constitutional government as distinctive Taiwanese values that separate Taiwan from mainland China.

Standardised Mandarin plus recognised regional languages. Mandarin remains the national language and the primary language of education and administration. Taiwanese Hokkien, formerly suppressed, is now officially recognised, taught in schools, and used in public life. Hakka Chinese and indigenous Austronesian languages have similar recognition. The multilingual policy is a modification of the KMT’s original Mandarin-only framework but retains Mandarin as the national coordinating language.

Contested Chinese heritage. The relationship of Taiwanese identity to Chinese civilisational inheritance has been the central political question of the post-2000 period. The KMT position has been that Taiwan is part of a broader Chinese civilisational community while remaining politically distinct from the People’s Republic. The DPP position has emphasised Taiwanese identity as distinct from Chinese identity, with reference to the pre-1949 Japanese-colonial experience and the pre-Japanese Austronesian and Fujianese-settlement history. The 2014 Sunflower Movement, in which student activists occupied the legislature to protest a proposed trade agreement with mainland China, represented the emergence of a more assertive Taiwanese-nationalist position particularly among younger Taiwanese.⁴

Continuing security threat as identity focus. The People’s Republic’s continued formal claim to Taiwan and its progressive military modernisation have made national security a central component of Taiwanese identity. Taiwan’s Self-Defense Forces, Reserve Command structure, and civil defense preparation are all part of the daily experience of Taiwanese nationhood. The external threat, as in the Korean case, functions as an identity-construction mechanism even as it imposes continuing strategic costs.

The costs

The Taiwan case, unlike the Meiji Japan or Korea cases, carries the specific documented cost of the White Terror and the 228 Incident. The 228 Incident of February 1947 killed an estimated 18,000 to 28,000 Taiwanese in the initial KMT response to island-wide protests; the subsequent decades of White Terror produced additional executions and imprisonments. These events are, in operational terms, the price the Taiwanese population paid for the KMT’s authoritarian consolidation and subsequent developmental transformation.

The Taiwanese Transitional Justice Commission, established in 2018 and completing its initial phase in 2022, has documented the White Terror period’s victims and their circumstances. The government has issued formal apologies, provided compensation to victims and families, and constructed the 228 Peace Memorial Park and Museum in Taipei as public memorialisation. The transitional justice process has been contested — the DPP-led implementation has been criticised by KMT-aligned opinion as politically motivated, and the specific compensation and accountability decisions have been controversial — but the process has produced substantial documentation and recognition of the historical costs.⁵

The specific quantitative comparison with the Ethiopian case: Taiwan’s total political deaths under the KMT authoritarian period, including the 228 Incident and the White Terror, are estimated at approximately 30,000 to 50,000 across four decades. Ethiopia’s total political deaths across the Derg Red Terror, the wars with Eritrea and Somalia, the EPRDF-era conflicts, the Tigray war, and the current Amhara and Oromia conflicts are estimated at well over one million across four decades. The Taiwanese population paid a substantial price. The Ethiopian population has paid a far larger price. The Taiwanese population has received in return a mature democracy, a globally-leading semiconductor economy, and a distinctive national identity. The Ethiopian population has received the specific conditions this series exists to examine.

What the Taiwan case shows

Three points from the Taiwan case are relevant to the Ethiopian argument.

First, that the development-then-democratisation sequencing can work when executed with sustained political will across leadership succession. The Chiang Kai-shek to Chiang Ching-kuo to Lee Teng-hui succession maintained the KMT’s developmental framework while producing progressively increased political openness. This is the sequencing that Meles Zenawi’s revolutionary democracy invoked as its theoretical template and that his successor Hailemariam Desalegn was supposed to implement. The Ethiopian equivalent did not happen; the EPRDF’s post-Meles leadership produced neither democratisation nor sustained developmental momentum. Taiwan’s case shows the sequencing is possible; Ethiopia’s case shows that possibility is not automatic.

Second, that ethnic-linguistic diversity can be accommodated within a national identity without institutionalising the diversity as constitutional sovereignty. Taiwan’s mainland Chinese, native Taiwanese Hokkien, Hakka, and indigenous populations have not been given constitutional secession rights or ethnic-federal regional states. They have been given recognition, language rights, cultural support, and progressively equal political participation. The result has been the emergence of a Taiwanese national identity that incorporates the diversity without being paralysed by it. Ethiopia’s ethnic-federal framework, by contrast, has institutionalised the diversity in ways that have made pan-national identity construction structurally difficult.

Third, that developmental authoritarianism can end without regime collapse if the developmental transformation has produced a middle class with political consciousness and organisational capacity, and if the authoritarian leadership makes the calculation that transition is safer than continued repression. Chiang Ching-kuo made this calculation. Lee Teng-hui executed the transition. The Taiwanese population had been prepared by two decades of developmental transformation to participate in constitutional politics. Ethiopia has, on multiple occasions, faced comparable transition moments — 1974, 1991, 2005, 2018 — and has not, on any of them, produced the managed transition that Taiwan produced. The reasons are examined in Part VI.

What comes next

Article 9 takes up China under Deng: the case in which developmental transformation was pursued without democratisation and continues in that mode into 2026. Article 10 synthesises the Tigers’ shared features. Part IV then examines Singapore as the central multi-ethnic case.

---

References

Footnotes

1. On the KMT arrival, the 228 Incident, and the White Terror, see Denny Roy, Taiwan: A Political History (Cornell, 2003); Steven E. Phillips, Between Assimilation and Independence: The Taiwanese Encounter Nationalist China, 1945–1950 (Stanford, 2003). ↩

2. On the Formosa Incident, see James Wei, Taiwan Independence Movement: The Origins of Taiwan’s Democracy (Routledge, 2019). ↩

3. On the Ten Major Construction Projects, see the Council for Economic Planning and Development historical archives. ↩

4. On the Sunflower Movement, see Ming-sho Ho, Challenging Beijing’s Mandate of Heaven: Taiwan’s Sunflower Movement and Hong Kong’s Umbrella Movement (Temple, 2019). ↩

5. On transitional justice, see the Transitional Justice Commission of Taiwan, Final Report (2022). ↩