Case Study 8.1: The Great Firewall and Bitcoin — China's Attempts to Control the Network

Background

No country has attempted to control Bitcoin as aggressively, persistently, and through as many different mechanisms as China. The story of Bitcoin in China is a decade-long case study in network-level censorship resistance — one that reveals both the remarkable resilience of peer-to-peer architectures and the very real limits of that resilience when a state deploys its full technical and regulatory apparatus.

China's relationship with Bitcoin has passed through several distinct phases, each escalating in severity and each met with a different set of adaptations from the Bitcoin community. Understanding this history is essential for any serious assessment of Bitcoin's censorship resistance properties, because China represented the most significant real-world stress test the network has faced.

Phase 1: The Banking Ban (2013-2017)

In December 2013, the People's Bank of China (PBoC) and five other government agencies issued a joint notice titled "Notice on Preventing Bitcoin Financial Risks." The notice classified Bitcoin as a "virtual commodity" rather than a currency and prohibited financial institutions and payment companies from conducting Bitcoin-related business. Banks were forbidden from opening accounts for Bitcoin exchanges, processing Bitcoin payments, or offering Bitcoin-related financial products.

The immediate effect was a sharp price decline and the closure of some exchange services. However, the notice explicitly did not prohibit individuals from holding or trading Bitcoin. Chinese exchanges adapted by using workarounds: personal bank accounts, voucher systems, and peer-to-peer trading. Trading volume quickly recovered and eventually far exceeded pre-ban levels. By 2016, Chinese exchanges (particularly OKCoin, Huobi, and BTCC) dominated global Bitcoin trading volume, at times accounting for over 90% of reported spot trading.

From a network perspective, the 2013 ban had minimal impact. Bitcoin's P2P network continued to operate in China because the ban targeted financial intermediaries, not the protocol itself. Chinese nodes could still connect to the global network, relay transactions, and mine blocks. The Great Firewall of China (GFW) — the country's sophisticated internet censorship infrastructure — did not target Bitcoin protocol traffic.

Key lesson: Banning institutional intermediaries (banks, regulated exchanges) is insufficient to stop a P2P network. Users route around intermediary restrictions by creating new intermediaries or trading peer-to-peer.

Phase 2: Exchange Regulation and the ICO Ban (2017)

In September 2017, China escalated dramatically. The PBoC ordered all domestic cryptocurrency exchanges to cease operations. Simultaneously, regulators banned Initial Coin Offerings (ICOs), classifying them as illegal fundraising. This time, the ban was enforced: major exchanges like Huobi and OKCoin moved their operations offshore (to Hong Kong, Singapore, and later the Seychelles), while some smaller exchanges simply shut down.

The exchange ban was more disruptive than the 2013 banking ban because it targeted the primary venues where Chinese users bought and sold Bitcoin. However, Bitcoin trading among Chinese users did not stop. It migrated to:

  • Over-the-counter (OTC) desks operated by the relocated exchanges, where buyers and sellers were matched directly and settled via bank transfer or payment apps like Alipay and WeChat Pay.
  • Peer-to-peer platforms like LocalBitcoins, where users could arrange direct trades.
  • VPN-accessed foreign exchanges, where Chinese users created accounts using foreign identification or without KYC (know your customer) requirements.

The Bitcoin network itself was again unaffected at the protocol level. Chinese miners continued to produce blocks. Chinese full nodes continued to validate and relay. The exchange ban was an economic and access-layer restriction, not a network-layer one.

Key lesson: Banning trading venues pushes activity to harder-to-regulate channels (OTC, P2P, foreign platforms). The Bitcoin network continues to function because its consensus mechanism does not depend on exchange infrastructure.

Phase 3: The Mining Ban (2021)

The most consequential Chinese action against Bitcoin came in May-June 2021, when the State Council's Financial Stability and Development Committee announced a crackdown on Bitcoin mining. Multiple provinces — including Inner Mongolia, Sichuan, Xinjiang, and Yunnan, which together hosted an estimated 50-65% of global Bitcoin hash power — ordered mining operations to shut down.

This was unprecedented. For the first time, a government action directly affected Bitcoin's consensus layer (mining), not just its access layer (exchanges). The results were dramatic:

Immediate effects: - Bitcoin's total hash rate dropped by approximately 50% over June-July 2021, from roughly 180 EH/s to 90 EH/s. - The difficulty adjustment mechanism activated as designed: difficulty dropped roughly 28% in the July 3, 2021 adjustment and continued falling over subsequent adjustments. - Block times temporarily increased to an average of approximately 14-16 minutes (compared to the 10-minute target) before difficulty adjustments normalized them. - Some miners relocated their equipment to other countries (Kazakhstan, the United States, Russia, Canada). This relocation took weeks to months.

What DID NOT happen: - The Bitcoin network did not stop. Blocks continued to be produced, just more slowly until difficulty adjusted. - No transactions were censored, reversed, or invalidated. - The protocol continued to function exactly as designed. - By December 2021 — roughly six months later — Bitcoin's hash rate had fully recovered and exceeded pre-ban levels, now distributed more globally.

Network topology impact: The mining ban had a notable effect on Bitcoin's network topology. Before the ban, a significant fraction of full nodes and miners were located in China, creating a dense cluster within the network graph. After the ban, this cluster largely dissolved. The geographic distribution of nodes became more even, which arguably improved Bitcoin's censorship resistance by reducing dependence on any single jurisdiction.

However, the migration also revealed vulnerabilities. During the transition period, hash power was more concentrated among the remaining (non-Chinese) miners, temporarily increasing the influence of mining pools like Foundry USA, Antpool (which had partially relocated), and F2Pool. The network's dependence on a small number of ASIC manufacturers (Bitmain, MicroBT, Canaan) was also exposed, as the migration required large-scale hardware logistics.

The Role of the Great Firewall

Throughout all three phases, one question persisted: did the Great Firewall of China actively interfere with Bitcoin P2P traffic?

Research by academics and the Bitcoin community has produced nuanced findings:

Latency injection. Studies in 2017-2020 found evidence that the GFW introduced artificial latency (delays) into Bitcoin P2P connections crossing the Chinese border. While Bitcoin traffic was not blocked outright, connections between Chinese and non-Chinese nodes experienced higher latency than geographically comparable non-Bitcoin connections. This latency could be a side effect of the GFW's deep packet inspection (DPI) system analyzing Bitcoin traffic rather than a deliberate degradation, but the effect was measurable.

Intermittent disruption. Some Chinese node operators reported periodic connection drops on port 8333 (Bitcoin's default port) that were consistent with firewall interference. These disruptions were not persistent enough to prevent Chinese nodes from maintaining connectivity but did reduce the quality and reliability of connections.

Block propagation effects. During periods of apparent GFW interference, blocks produced by Chinese miners took measurably longer to reach non-Chinese nodes, and vice versa. In a 2020 paper, researchers estimated that GFW-induced latency increased the orphan (stale) block rate for Chinese miners by a small but statistically significant amount, effectively imposing a "censorship tax" on Chinese mining without fully blocking it.

Circumvention. Chinese Bitcoin users and miners extensively used VPNs, Tor, and alternative ports to circumvent GFW interference. The Bitcoin community developed relay bridges specifically designed to maintain connectivity between Chinese and non-Chinese nodes. Mining pools operated dedicated relay infrastructure to ensure their blocks could propagate globally regardless of GFW conditions.

Lessons for Censorship Resistance

China's multi-year attempt to control Bitcoin provides several critical lessons:

1. Layered attacks require layered defenses. China attacked Bitcoin at multiple layers: banking access, exchange operations, mining, and network traffic. Each layer required a different defense mechanism. Bitcoin's protocol-level censorship resistance (P2P gossip, difficulty adjustment) handled the mining ban well, but the exchange and banking bans required social and economic adaptations (OTC trading, offshore platforms) that are outside the protocol's scope.

2. The difficulty adjustment is Bitcoin's most underappreciated feature. When 50% of hash power disappeared overnight, the difficulty adjustment ensured that blocks continued to be produced. Block times increased temporarily but normalized within weeks. Without this mechanism, the mining ban could have caused a catastrophic spiral: slower blocks would lead to a worse user experience, which could lead to further hash power departure, which would slow blocks further. The difficulty adjustment circuit-breaker prevented this.

3. Geographic diversification emerged from adversity. The forced redistribution of mining from China to multiple countries arguably made Bitcoin more robust. Before the ban, a single government could theoretically pressure a majority of hash power. After the ban, no single government controls more than roughly 35-40% of hash power.

4. Network-level censorship is hard but not impossible. The GFW's ability to degrade (if not fully block) Bitcoin P2P traffic demonstrates that sophisticated network censorship can impose costs on Bitcoin participants. The network's defenses (Tor, VPNs, encrypted transport) mitigate but do not fully eliminate these costs.

5. Bitcoin survived, but access was reduced. For the average Chinese citizen, using Bitcoin became significantly more difficult after the successive bans. While the network continued to function globally, Chinese users lost easy access to exchanges, mining operations, and reliable node connectivity. Bitcoin's censorship resistance protected the network, but it did not protect the users in the censoring jurisdiction from bearing costs.

Discussion Questions

  1. The Bitcoin network survived China's mining ban with minimal disruption at the protocol level. Does this prove that Bitcoin is "unstoppable," or does it prove that China's approach was insufficiently aggressive? What would a government need to do to actually stop Bitcoin?

  2. After the mining ban, Chinese miners relocated to multiple countries, diversifying the geographic distribution of hash power. Was this an unintended benefit of China's policy? Could a future coordinated ban by multiple countries have a more lasting effect?

  3. The Great Firewall's apparent degradation (rather than outright blocking) of Bitcoin traffic represents a "soft censorship" approach. Is soft censorship (making something harder, slower, and less reliable) more effective than hard censorship (outright blocking) against a system like Bitcoin? Why or why not?

  4. Bitcoin's protocol continued to function throughout all phases of Chinese regulation, but Chinese users faced increasing difficulty accessing the network. If a censorship-resistant protocol is usable only by technically sophisticated users willing to use VPNs and OTC trading, does it meaningfully serve its stated purpose of providing permissionless money?

  5. Compare China's approach to Bitcoin regulation with El Salvador's approach (adopting Bitcoin as legal tender). What do these opposite policies reveal about the range of possible government responses to decentralized networks?