Economics As A Weapon

On February 26, 2022, two days after Russian tanks crossed into Ukraine, the European Commission, United States, United Kingdom, and Canada announced they would cut selected Russian banks from SWIFT, the network that connects more than 11,500 financial institutions across 220 countries.^[1] SWIFT is, at its core, a messaging service. Banks use it to tell each other where to send payments. The fact that losing access to a messaging service can shut down a country's economy tells you everything about how concentrated the system's dependencies really are.

The response was immediate — not from SWIFT itself, which didn't sever the seven banks until March 12, but from the broader sanctions package that accompanied it. Asset freezes, central bank reserve lockouts, payment network withdrawals. Lines formed at ATMs across Russia before the weekend was over. Moscow commuters tapped their phones at metro turnstiles and got nothing. Apple Pay, Google Pay, and Samsung Pay all stopped working because the bank that handled transit payments was on the sanctions list. The ruble lost 40% of its value in under two weeks. Migrant workers from Kyrgyzstan and Tajikistan, countries where remittances from Russia account for roughly 30% of GDP, watched their lifeline go dark.^[2]

But of everything in that sanctions package, SWIFT disconnection was the mechanism that mattered most structurally. A cooperative founded in 1973 had become the single point of failure for global finance, and someone had just demonstrated that it had an off-switch. Scholars Henry Farrell and Abraham Newman coined the term "weaponized interdependence" to describe exactly this: systems built for efficiency that become instruments of coercion because everyone depends on them and someone controls the gate.^[1]

Russia was not the first test case. That distinction belongs to Iran. In March 2012, SWIFT disconnected approximately 30 Iranian banks under EU sanctions. The stated target was Iran's nuclear program. Net oil export revenue fell from $95 billion in 2011 to $69 billion in 2012, a decline that accelerated as financial isolation compounded.^[3] Banks became so afraid of triggering secondary penalties that they refused to process even transactions the sanctions technically exempted. The architecture doesn't allow precision. You can't surgically disconnect a nuclear program while leaving the rest of an economy's financial relationships intact. SWIFT is a single network, and disconnection is all-or-nothing.

SWIFT severs access overnight. China's Belt and Road Initiative is more insidious. There's no single moment of disconnection. The leverage builds slowly, by design, until the borrower realizes the terms were the trap all along.

Hambantota Port sits on Sri Lanka's southern coast. China Harbour Engineering Company built it with roughly $1.26 billion in loans from China Eximbank, beginning in 2007. The first phase carried a 6.3% fixed interest rate, several times higher than World Bank concessional rates. The port underperformed. The debt compounded. In July 2017, the Sri Lankan government signed a 99-year concession handing China Merchants Port Holdings an 85% stake, plus development rights over 11.5 square kilometers of surrounding land. A New York Times investigation found $7.6 million had flowed from a China Harbour-controlled account to affiliates of the president's re-election campaign.^[4]

Hambantota is the most thoroughly documented case but not an outlier. AidData's 2021 study of over 13,000 Belt and Road projects across 165 countries found average Chinese loan interest of 4.2% versus 1% from OECD creditors, repayment periods under 10 years versus 28, and a 31-to-1 ratio of loans to grants. By 2022, 60% of China's overseas lending portfolio supported borrowers already in financial distress.^[5] The weapon doesn't need to fire every time. The possibility reshapes behavior.

The same dynamics operate at every scale. You don't need to be a government to weaponize a chokepoint. You just need to control one.

In December 2010, Visa, Mastercard, and PayPal cut off donations to WikiLeaks within days of each other.^[6] No charges had been filed. No court order existed. Payment processors simply decided the money should stop flowing, and because virtually all online payments route through the same handful of networks, that decision was functionally final. The blockade cost WikiLeaks an estimated 95% of its revenue.^[6] Whatever you think of WikiLeaks, the mechanism is the point: three companies made a call, and a legal organization's funding evaporated overnight.

WikiLeaks was a chokepoint at corporate scale. Amazon's destruction of Diapers.com was the slow version.

In 2009, Diapers.com was approaching $300 million in annual revenue. Amazon proposed an acquisition. The founders said no.^[7] Amazon dropped diaper prices 30% and deployed automated pricing bots that tracked Diapers.com in near-real-time. On September 8, 2010, Amazon launched steep diaper discounts timed for the exact morning the founders were flying to Seattle for acquisition discussions.^[7] Internal emails surfaced during a 2020 House Judiciary hearing showed an Amazon VP writing: "These guys are our #1 short term competitor... We need to match pricing on these guys no matter what the cost."^[8] Quidsi's executives calculated Amazon was losing $100 million every three months on diapers alone. The founders sold for $545 million. In 2017, Amazon shut down the entire operation and raised prices back up.^[8]

It's no-limit poker with asymmetric stacks. Amazon could shove all-in every hand because diapers were a rounding error on their balance sheet. Diapers.com was playing with their whole bankroll. But the deeper problem wasn't just size. When one platform becomes the marketplace a third of American e-commerce flows through, competitors don't have the option of simply selling somewhere else. The outcome was decided before a single hand was dealt.

When control over something people depend on gathers in one place, the capacity to weaponize it is already built. The structural counter isn't regulation or reform. Every one of these chokepoints exists because a system was designed with a central operator, and someone eventually used that position. SWIFT can sever a country's financial access because all payment messaging routes through one cooperative. Visa can shut down a legal organization's funding because online payments funnel through three processors. The architecture created the vulnerability. The only way to remove it is to build systems where no single entity controls the flow. Payment networks that settle transactions between participants directly, without a central authority that can be pressured, sanctioned, or simply decide the money should stop. That technology exists. Whether it gets built into infrastructure people actually use is a different question, but it's an engineering problem, not a theoretical one.