What Beijing still needs is Hong Kong, not Xiānggǎng

Stephen Roach’s latest essay “From Hong Kong to Xiānggǎng” is provocative, asserting that Beijing’s post-2019 intervention amounted to a “hostile takeover” and that the city has become “Xiānggǎng”, merely another mainland city. He points to mainland issuers, tighter national security rules, departing foreign judges, a diminished Western expatriate presence and the greater visibility of Mandarin. As a political scientist who has travelled regularly between the mainland and Hong Kong for years, I find some of his observations plausible. They also echo a familiar public mood within Hong Kong since the 1997 handover: fear of the city’s “mainlandisation”.

But plausible observations can still lead to an implausibly absolute verdict. Hong Kong has changed; it has not changed species. Integration is not a magic trick in which one side disappears into the other. Shanghai and Shenzhen are formidable financial centres on the mainland, but neither combines an open capital account, a separate customs territory, common law and independent international economic status, which are institutionalised in the Hong Kong Basic Law and reproduced by local courts, banks, regulators and businesses.

Still ‘one country, two systems’

The contrast remains striking under “one country, two systems” (OCTS). The Fraser Institute’s 2025 Economic Freedom of the World report ranked Hong Kong first among 165 jurisdictions, while mainland China placed 108th. The World Justice Project’s 2025 Rule of Law Index ranked Hong Kong 24th out of 143 jurisdictions, compared with mainland China’s 92nd place. Hong Kong has certainly declined on several political and rights-related measures. Even so, the institutional distance between the city and the mainland remains far too wide to sustain the claim that Hong Kong has become just another mainland city.

For Beijing, the deeper challenge is that colonial rule and sovereign rule operate according to fundamentally different logics. London governed Hong Kong as an overseas possession. After 1997, Beijing had to incorporate the territory into its sovereign order without dismantling its capitalist system. OCTS was therefore a demanding experiment from the outset, not a promise to preserve late-colonial politics frozen in amber. Its central tension has always been how to sustain a highly autonomous and globally connected jurisdiction without allowing it to become an enclave beyond the effective authority of the sovereign state.

Yet “mainlandisation” was never Beijing’s preferred solution. The unrest of 2019 brought this tension to a breaking point. The national security legislation and electoral restructuring that followed unquestionably strengthened political integration and narrowed parts of Hong Kong’s political space. Their implication for the city’s civil society, public debate and international confidence deserves serious scrutiny.

Analytically, however, these measures were intended to make Hong Kong governable within the framework of Chinese sovereignty, namely “one country”, rather than to abolish its separate system. Even at the height of the unrest, Beijing stopped short of direct intervention that would have openly displaced Hong Kong’s semi-autonomous status. In essence, it redrew the boundary between “one country” and “two systems”; it did not erase it.

China gains if it keeps Hong Kong different

Hong Kong’s difference has instrumental value for Beijing. As long as Western and Chinese regulatory systems remain different and capitalist and socialist political economies continue to coexist, China will need a jurisdiction able to work across the divide, especially in a Cold War- like geopolitical rivalry. A thoroughly mainlandised Hong Kong, Roach’s “Xiānggǎng”, would be strategically pointless. Beijing already has Shanghai, Shenzhen, Guangzhou and Hainan. It does not need an expensive imitation of them on Victoria Harbour.

Roach sees the dominance of mainland issuers in Hong Kong’s market as proof that the market is no longer genuinely international. But Chinese firms raise international capital in Hong Kong precisely because Hong Kong is not institutionally identical to the mainland. In a word, a Chinese issuer base and an international capital platform can coexist. In 2025, Hong Kong hosted a record 11,070 companies with parent companies outside the city including mainland China, the US, Japan, the UK and Singapore, employing about 509,000 people. The network is truly more China-centred than before, but it is hardly sealed off from the world.

Beijing’s own language reflects the calculation above. Speaking in Hong Kong in 2022, Xi Jinping declared that there was “no reason to change” OCTS, that it should be upheld over the long term, and that Hong Kong’s distinctive status and advantages must be preserved. Four years later, those words carry even greater strategic weight. The more China encounters technological, financial and geopolitical barriers abroad, the more valuable a legally and economically differentiated Hong Kong becomes.

Mainland being ‘Hong-Kongised’?

There is also a deeper irony. While observers debate Hong Kong’s “mainlandisation”, Beijing has spent decades trying to endow selected mainland jurisdictions with Hong Kong-like functions. The Shanghai Free Trade Zone, for example, has experimented with freer cross-border flows and higher-standard trade rules, while the Hainan Free Trade Port is developing special customs arrangements, lower barriers and more convenient capital movement. Neither is becoming Hong Kong in any constitutional sense. Both represent attempts to transplant elements of Hong Kong’s openness and intermediary capacity into carefully controlled mainland settings.

This echoes Deng Xiaoping’s original developmental logic. In 1988, Deng observed that China had only one Hong Kong but intended, through further opening, to “build several more Hong Kongs on the mainland”. He also explained that the commitment to keep OCTS unchanged for 50 years was a deliberately vivid formulation: during the first 50 years, the policy could not be changed; after that, there would be no need to change it. In Deng’s developmental vision, China’s modernisation meant, in part, making the mainland more Hong Kong-like.

The more interesting long-term question, then, may be not how rapidly Hong Kong is being mainlandised, but how selectively the mainland is being “Hong-Kongised”. Hong Kong’s political integration with Beijing is unlikely to be reversed, but integration is not the same as assimilation. The city remains valuable precisely because it is both firmly part of China and institutionally distinct from the mainland. In this sense, Beijing does not need a “Xiānggǎng”. 

Related: From Hong Kong to Xiānggǎng | [Big read] The northbound drift: Hong Kong’s quiet exodus to the Greater Bay Area