The Great Firewall and the Splinternet Economy

A protest, a shutdown, and a launch

In July 2009, ethnic violence between Uyghur and Han residents broke out in Urumqi, the capital of China's Xinjiang region, after a workplace dispute in Guangdong province turned into a wider grievance. The unrest left hundreds dead by official count and led the government to cut mobile and internet service across the region for months. It also gave Beijing cause to act against the platforms that had carried word of the unrest outward: Twitter and Facebook were blocked nationwide, and most of China's existing domestic microblogging services were shut down alongside them, clearing the field rather than simply policing it. The scale of the response set 2009 apart from the routine filtering the country had already been running for years: this was not a keyword pulled from a search result, but an entire category of foreign service made unreachable at once, in the same month a comparable domestic category was made to disappear.

The shutdown created a problem the government had to solve on its own terms. China's population had already discovered microblogging, and a market that size does not sit quiet for long. Within about a month, Sina, one of China's major internet portals, launched a new microblogging product built for the same purpose, and built inside the country's regulatory reach from day one.

Sina Weibo went live on August 14, 2009. Growth was not instant: it took until February 2011 to reach 100 million registered users, a stretch during which the product found its format and its early audience learned to use it in place of the platform they had lost. By 2013, Weibo had passed 500 million users, a figure corroborated across several industry accounts of the platform's early history. The timing is the detail worth sitting with: the space Twitter would have filled did not stay empty. It was filled, on purpose, by a domestic product with nowhere else to compete for the same users.

The machinery behind the wall

The system that made the 2009 blocks possible predates them by years. China's internet filtering apparatus began under the Golden Shield Project, an initiative that grew through the 2000s into the set of technical controls now known informally as the Great Firewall. Since 2013 it has been operated by the Cyberspace Administration of China, the body responsible for the country's internet policy and enforcement. The firewall works by inspecting the data packets crossing China's network borders and blocking traffic that matches banned keywords or known foreign domains, a method that filters content at the level of the network itself rather than relying on individual platforms to police themselves.

That distinction matters for what came next. A country that filters at the network level does not need a foreign platform's cooperation to remove it from the market; it can simply make the platform unreachable. Twitter and Facebook were never negotiated out of China. They were switched off, and no argument about market size or user demand could switch them back on. That is a different kind of market entry barrier than the one a domestic competitor usually has to clear, and it is the one Weibo and, later, WeChat were built behind.

The wall is not, and was never, a single unbroken barrier. Hong Kong and Macau are formally exempt under the "one country, two systems" arrangement that governed their return to Chinese sovereignty, and both retained open internet access that mainland China does not have. That exemption became politically consequential after Hong Kong's National Security Law gave authorities a legal basis to block specific websites, including ones documenting anti-government protests, inside a territory the firewall itself does not reach. The boundary of the system has moved with the politics around it, a useful caution against treating the firewall as fixed infrastructure rather than a policy applied, extended, and occasionally worked around.

The substitution economy

Weibo's growth solved half the problem the 2009 blocks created: it gave China's internet population somewhere to hold the public conversation Twitter would have hosted. Tencent, a company that had built its business on desktop messaging, solved the other half. It released WeChat in 2011, aimed less at replacing a single blocked service than at becoming the app a Chinese phone owner would open by default, for everything.

That ambition worked. By 2018, WeChat had grown into the world's largest standalone mobile app, with more than 1 billion monthly active users, most of them inside China. It got there by layering messaging, payments, a news feed, and a marketplace of small in-app programs run by outside businesses into one product, so that a user who never opened a browser could still pay a bill, order food, or buy from a small merchant without leaving the app. Each of those functions is, elsewhere in the world, a business Twitter, Facebook, PayPal, or Amazon might have competed for. In China, none of them had the chance to.

The economic consequence follows from the user numbers rather than from any single disclosed figure: a domestic platform carrying a billion monthly users' worth of advertising, retail transactions, and payment volume is a domestic platform capturing a market that, absent the block, foreign firms would have been competing to serve. Winning the audience was only the first step. Weibo and WeChat built the commercial infrastructure around it, the advertising sales relationships, the merchant tools, and the payment rails, inside two Chinese companies, Sina and Tencent, rather than inside Silicon Valley ones. That infrastructure compounds over a decade in a way a lost social media account does not.

The label China's press settled on for WeChat, the app for everything, is a useful shorthand for what the substitution actually built. A user paid a friend back for lunch, read the news, booked a doctor, and messaged a coworker inside one product, and a merchant who wanted to reach that user had, in practice, one gateway to do it through. Advertisers who once split budgets across several Western platforms found their Chinese spend consolidating toward whichever domestic app carried the audience, because there was no second gateway competing for the same attention. That consolidation is the mechanism behind the phrase value transfer: it did not require a single dramatic event, only a decade of ordinary transactions happening inside a Chinese company's product rather than an American one's.

The arrangement also kept the state's other objective intact. WeChat stores the data of mainland-registered accounts inside mainland China and is subject to content monitoring and disclosure to authorities on request. Accounts discussing sensitive subjects, including the Hong Kong protests, have been restricted even for users of Chinese descent living outside the country. The same platform that captured the commercial value the block created also gave the government a channel it could not have had over Twitter or Facebook, one it could monitor and, when it chose to, restrict.

A sovereignty model, exported

China's move did more than build two large companies. It demonstrated, in public and over a measurable decade, that a government could block the world's dominant social platforms and still end up with a functioning, commercially successful internet, one answerable to its own law rather than to a foreign platform's terms of service. That was a proof of concept other governments were positioned to notice.

Russia moved first among the major examples. President Vladimir Putin signed amendments to the country's personal data law on December 31, 2014, and the requirement took effect on September 1, 2015: companies collecting data on Russian citizens had to store that data on servers located inside Russia. Roskomnadzor, the country's communications regulator, was given the power to block sites that did not comply, and the law gave companies operating in Russia a hard choice: build the local storage the state demanded, or lose access to the country's users. The law did not attempt to build a Russian WeChat. It borrowed a narrower piece of the same idea, that a state's citizens' data, and by extension their online lives, should sit under that state's jurisdiction rather than a foreign company's.

Other governments have since reached for versions of the same logic, for reasons ranging from security to industrial policy to straightforward political control. India barred dozens of Chinese-owned apps in 2020, citing sovereignty and security concerns, a move that turned China's own playbook against a set of Chinese products. Iran has built out its own domestic network infrastructure with the stated aim of being able to operate independent of the global internet if it chooses to. Neither case is identical to China's, and each has its own specific law and history behind it, but the shared premise is the one the Great Firewall proved economically viable first: a national internet does not have to be the same internet everyone else uses, and a state that controls the medium controls what grows on it. The term that eventually gathered up this whole family of policies, digital sovereignty, entered common use only after China had already shown what the practice looked like in a working economy of well over a billion people, which is part of why the term carries as much weight as it now does in policy conversations far from Beijing.

That is the geopolitical half of the story, and it is worth stating plainly rather than as a subplot. What began as one country's response to a single unrest event hardened, over the following decade, into a template. The internet split, and the split was not an accident of engineering. It followed the border of political authority, one government's law at a time.

The durability of the wall

A wall built in 2009 to answer a specific protest has, more than fifteen years later, not come down. If anything, the gap it created has widened rather than closed. An Asia Society estimate cited in 2026 put China's top five domestic apps at roughly 1,000 times the traffic of the top five foreign apps the Great Firewall blocks, a figure that should be read as an estimate rather than a precise census, but one that points in a single clear direction: the substitution China engineered in 2009 did not fade as memory of the original block faded. It compounded.

That durability is itself informative. Plenty of policies meant to redirect a market lose their grip once the original justification recedes and users find workarounds. China's did not, largely because the substitute products got better on their own merits, not only because the originals stayed blocked. WeChat's built-in marketplace of mini-programs and its integrated payments, in particular, gave Chinese users a reason to stay even in a hypothetical world where Twitter and Facebook became available again. A wall that only blocks accomplishes less, over a decade, than a wall paired with a product people would choose anyway.

The same underlying logic, that whoever controls the medium controls who gets found, paid, and heard through it, has a live echo in how information now gets surfaced by AI systems. Search and answer engines trained and retrieved from different underlying stocks of content, an open global web on one side and a walled national one on the other, will not necessarily name the same businesses, tell the same stories, or surface the same facts when asked the same question. That divergence is still forming rather than settled, and it should be treated as a scenario to watch rather than a forecast, but the mechanism is the one this article has traced throughout: a medium's boundary decides what is inside it to be found.

What the split proved, and what it did not

The clean version of this story, that a government blocked two platforms and two domestic ones grew to replace them, is true and also incomplete. Every shift of this kind does two things at once: it opens access for some group that did not have it before, and it hands new concentrated control to whoever ends up running the replacement. China's case did both. Blocking Twitter and Facebook took away a channel millions of Chinese users had adopted for open, cross-border conversation, including the reporting that had drawn attention to the Urumqi unrest in the first place. It also, through Weibo and WeChat, gave hundreds of millions of people a genuinely useful set of communication and commerce tools, ones that in some respects, integrated payments especially, outran what Western platforms offered for years afterward. Neither half of that record cancels the other out.

The causal question is also less settled than the growth numbers alone suggest. Weibo and WeChat did not grow only because Twitter and Facebook were unavailable. China's internet population was, and is, the largest in the world; both products had state backing, well-funded corporate parents, and a mobile-first market that adopted smartphones on its own fast timeline. Crediting every user of a billion-plus-user app to a single 2009 policy decision would overstate what one block, on its own, can explain. What the evidence supports is narrower: the block removed the two most likely competitors from the field at the exact moment a domestic alternative needed room to establish itself, and that room mattered.

What the episode settled, more convincingly than any single growth statistic, is a question about power rather than about product quality: a state can, in fact, out-compete the open web's platforms inside its own borders, if it is willing to block the originals and invest in the substitute long enough for the substitute to become better on its own terms. That is the finding Russia's 2015 law borrowed a piece of, and it is the finding other governments have kept borrowing pieces of since. The internet that resulted is not one internet. It is several, drawn along the lines of political authority, and the line China drew first, in the weeks after a riot in a city few outside the country had heard of, is still the one other states measure their own against. Read that way, the story is less about two apps than about a lesson learned once and applied repeatedly: control the medium, and the economy and the politics that grow up around it will answer to whoever holds that control, in Beijing or wherever the next version of the same policy is written.