Capital and the Contested Nation
Part IV argued that capital holds one line when others are crossed: the contract, and the tissue of courts and law that enforces it. This essay asks what happens where nationhood itself is contested. Each of these places is an outcome of history, enshrined in law or in compromise, and still disputed. What paths did capital choose in each, and what does that tell us? We take five cases: South Africa, Northern Ireland, Cyprus, Catalonia and Israel. They differ profoundly. Only one classified its people by race in law, and comparing them does not make them equivalent. We are not examining the ideologies or the arguments. We are examining what capital, and the legal and institutional tissue that supports it, did and is doing in response to the structures it encountered in each case.
South Africa: Capital Defects
For most of its life, apartheid was profitable for the capital inside it and tolerable to the capital outside it. The gold mines ran on a migrant labour system that pass laws kept cheap and controlled. Manufacturing grew behind protection, and foreign banks lent against gold and a reliable debt record. Capital did not invent the system, but for decades it found the system workable. Most of what happened after 1977 is the story of that ceasing to be true.
A first crack came as a code of conduct, not a withdrawal. In 1977 Leon Sullivan, a Baptist minister who sat on General Motors’ board, published principles for American firms operating in South Africa: desegregated workplaces, equal pay for equal work, training and promotion for black employees. The logic was to stay and reform from inside. In June 1987 Sullivan concluded it had not worked and called on American companies to leave. The code’s lasting legacy was the tool itself, a written standard that investors could hold a company to. It would be used again elsewhere.
The decisive move was made by a bank reading its risk, not its conscience. The government declared a state of emergency in July 1985. By the end of that month Chase Manhattan had stopped rolling over its short-term loans to South African borrowers, and other banks followed. P.W. Botha’s “Rubicon” speech on 15 August promised reform and delivered none, and the rand collapsed. At the end of August the government closed the stock exchange and the foreign exchange market, and on 1 September it declared a standstill on repaying its foreign debt. Chase’s reasoning was about repayment risk. That risk had been repriced by unrest in the townships and by divestment campaigns that made South African exposure expensive to hold. Moral pressure won by turning into a price.
Two weeks later domestic capital crossed a line the state had drawn. On 13 September 1985 Gavin Relly, chairman of Anglo American, led a small group of businessmen and editors to a game lodge in Zambia to meet the ANC’s leadership in exile, Oliver Tambo among them. Botha criticised the trip, and Relly came home to a storm. This was not a conversion. It was a judgement that any future settlement would include the ANC, and that capital’s interests were better served by a seat at the table than by a regime that could no longer borrow.
Owners, managers, governments and workers did not move together. In November 1986 Barclays sold its stake in its South African subsidiary, the country’s largest commercial bank, to a consortium that included Anglo American. Mobile capital left, and immobile capital bought what it left behind. That year the US Congress passed sanctions over President Reagan’s veto, while the British government resisted all but token measures. Some American firms, IBM among them, left while keeping their products on South African shelves through licensing and distribution deals. Shell stayed and became the target of a consumer boycott. Black workers had their own channel: the Congress of South African Trade Unions, founded in 1985, made labour an actor in the crisis rather than its backdrop. Each of these agents weighed the same question on different scales.
The contract line held through the transition, and holding it was part of the price of the deal. The settlement of 1990–94 ended apartheid. It also protected property in the new constitution, kept the Reserve Bank independent and honoured the foreign debt; the last of the standstill debt was repaid on 15 August 2001. Capital helped bring the old order down and secured its own terms in the new one.
Return came with an exit attached. Sanctions lifted and foreign capital came back after 1994. In the same years five of the country’s biggest companies moved their headquarters and primary listings to London, with the new government’s permission: Billiton first, in 1997, then South African Breweries, Anglo American and Old Mutual in 1999, and Dimension Data in 2000. They wanted cheaper capital and room to grow, and Anglo presented the move as the route to larger capital markets and a simpler, more international structure after its merger with Minorco. The settlement that capital helped broker was also one it hedged against. Return and exit were the same decision, read from two sides.
Northern Ireland: Disciplined, Then a Lobby for Peace
The Stormont state rested on an industrial base that hired along communal lines. Harland & Wolff, the linen mills and the Belfast engineering firms drew their skilled workforces overwhelmingly from the Protestant community. In 1969 the Cameron Commission, commissioned by the Stormont government itself, found unfair methods of housing allocation by certain local authorities, and complaints, “now well documented in fact”, of discrimination in local government appointments. Private capital and the state were not the same thing. Their hiring and allocation patterns reinforced each other.
That capital declined whatever anyone did politically. Shipbuilding and linen collapsed through the second half of the century. Harland & Wolff, which employed around 35,000 people at its peak, had about 125 workers when it went into administration in 2019. Part of the old order’s economic base did not need to be dismantled. It shrank on its own.
Law imposed a contract line from inside. The Fair Employment Act of 1976 outlawed discrimination on grounds of religion or political opinion. It relied largely on voluntary compliance and achieved little. The 1989 Act, in force from January 1990, had teeth: compulsory monitoring of workforce composition, a stronger commission and tribunal, and disqualification from public contracts and grants for firms in default. By 1993 only one company had ever been disqualified. The threat set the price, and it rarely had to be paid.
Outside capital pushed through pension funds. The MacBride Principles were launched in November 1984, driven by the Irish National Caucus and championed by New York City’s comptroller, Harrison Goldin. They set fair-employment standards for American companies with Northern Ireland operations and were modelled on the Sullivan Principles. The city’s pension funds filed shareholder resolutions, and state legislatures followed with investment and procurement rules. The British government argued that the principles would deter investment. The tool built for Johannesburg was now being used on Belfast.
Then business became a voice for settlement. In July 1996, after the Drumcree disturbances, seven business, union and economic bodies formed what became known as the Group of Seven: among them the CBI, the Chamber of Commerce, the Institute of Directors and the Northern Ireland Committee of the Irish Congress of Trade Unions. That October they invited all the parties in the talks to a meeting in Belfast to hear the economic case for a settlement. They did not write the 1998 Good Friday Agreement. They made clear that capital would reward one, and labour stood beside them saying so.
The settlement left sovereignty contested and let the economy run as if it weren’t. The Agreement rested on consent, and it deliberately left the constitutional question open. The European single market had already made the border close to invisible for trade. Brexit reopened that question, and the Northern Ireland Protocol, revised in 2023 as the Windsor Framework, answered it with a second ambiguity: Northern Ireland sits inside the UK market and the EU market for goods at once. A firm in Northern Ireland has unfettered access to both, which no firm in Great Britain or the Republic has. Ambiguity became an asset capital could use.
Cyprus: Capital Adapts to Partition
Since 1974 the island has been divided, and capital learned to work on both sides of the line. Turkey’s military intervention that summer is called an invasion by the Republic of Cyprus and most governments, and a peace operation by Turkey. It left the north under Turkish Cypriot administration. The Turkish Republic of Northern Cyprus, declared in 1983, is recognised only by Turkey. Neither side’s capital needed the dispute resolved to make money, and in that sense the dispute was affordable.
The south turned an unresolved conflict into a financial centre. Low taxes, English common law, double-taxation treaties and EU membership from 2004 drew in deposits, much of them Russian. By the early 2010s the banking system was seven or eight times the size of the economy. Most of that capital had no stake in the island’s politics. It was there for the treaties.
In 2013 the state turned on the capital that had made the island rich. Losses on Greek debt left the two largest banks insolvent. The eurozone’s rescue made depositors pay: Laiki Bank was wound down, and 47.5 per cent of uninsured deposits at the Bank of Cyprus were converted into shares. Capital controls followed, the first ever inside the euro, and stayed until 2015. The state turned on capital, but carried out under eurozone rules. Insured deposits were protected. Uninsured ones were not.
The north became close to a rentier economy. Without recognition there are no direct international flights, no access to most world markets and no conventional sovereign borrowing. Turkish transfers, the Turkish lira, universities, casinos and a property market built on contested titles filled the gap. Capital in the north depends on Ankara the way capital in the south depends on Brussels and its treaty network.
Then the courts carried property law across the Green Line. In 1974 Meletis Apostolides’s family fled their land at Lapithos in the north. A British couple, the Orams, later bought it and built a holiday home on it. He sued in the Republic’s courts and won. In April 2009 the European Court of Justice ruled that the judgment had to be recognised throughout the EU, and in January 2010 the English Court of Appeal enforced it. The Orams left the house empty. Ownership in the north could no longer be insulated from claims in the south. The case did not end the partition, but it raised the price of buying into it.
Gas has added a new contested asset without changing the pattern. Offshore discoveries since 2011 brought ExxonMobil, Eni and others into waters the Republic licenses and Turkey disputes. In February 2018 Turkish warships stopped a drillship working for Eni from reaching its block off the island’s east coast, and the ship sailed for Morocco. The reserves are real. Commercial terms that both sides accept are not. Here too capital waits rather than forcing a settlement.
Catalonia: Capital Votes with Its Registered Office
In October 2017 contested nationhood arrived inside a eurozone democracy, and capital answered within days. The Catalan government held an independence referendum on 1 October, which Spain’s Constitutional Court had suspended and later ruled unconstitutional. Depositors began moving money out of Catalan banks. Their reasoning was simple: an independent Catalonia would leave the euro’s deposit protection and the European Central Bank’s supervision.
Any secession raises two questions that are often confused: whose currency, and whose debt. A new state can keep using a currency it does not issue, as Montenegro and Kosovo use the euro. What it cannot keep is the central bank behind that currency, the lender of last resort its banks turn to in a crisis. Debt runs the other way. The continuing state normally keeps its own debt and honours it, and how much the new state takes on is settled by negotiation, not by inheritance. A new state can refuse a share, but it needs to borrow from its first day, and refusal has a price. Catalonia had a further twist: a majority of the Generalitat’s own debt was owed to the Spanish state, through a liquidity fund created for the regions after the euro crisis. The government that wanted to leave was in debt to the government it wanted to leave.
The banks moved first, and the state made it easier for them. On 5 October Banco Sabadell’s board moved its registered office to Alicante. The next day the Spanish government approved an emergency decree letting a company’s board move its registered office within Spain without a shareholders’ vote, unless the company’s own statutes said otherwise. CaixaBank’s board resolved the same day to move to Valencia. The central government had lowered the cost of exit for capital leaving the region it was contesting. Thousands of companies followed over the following months.
The exit was mostly a signal, and a cheap one. A registered office is a legal address. Staff, headquarters buildings and operations largely stayed in Barcelona. The move cost little and said a lot: the region’s largest firms would not stay inside a state that might fall outside the euro. The declaration of independence on 27 October was met the same day by direct rule under Article 155 of the Spanish constitution. Capital’s verdict had already been given.
The return has been slow, partial and driven by each firm’s own interests. A pro-union regional government took office in 2024. In January 2025 Sabadell’s board voted to go home, saying the circumstances behind its move no longer applied. It was also fighting a hostile takeover bid from BBVA, and local political support was useful; the bid failed in October 2025, when barely a quarter of Sabadell’s shareholders accepted. The La Caixa Foundation and its holding company Criteria returned to Barcelona weeks after Sabadell’s decision. CaixaBank, in which they hold the largest stake, said it would stay in Valencia indefinitely. Owners went home and the listed bank stayed put.
Catalonia shows the asymmetry in its cleanest form. Exit was fast, collective and justified by risk. Return is individual, and each firm justifies it by its own advantage. The price of contested nationhood was set within a week, and nine years later it still hasn’t been fully refunded.
Israel: Capital, Conditionally Loyal
Israel is the only case still in motion, so this section will be careful to describe and not forecast. The election of 27 October 2026, certified on 4 November, gave the bloc led by Benjamin Netanyahu 53 seats and the bloc led by Gadi Eisenkot 48. Both are short of the 61 needed to govern. The Arab-led parties, the Joint List and Ra’am, won 19. On 7 November President Isaac Herzog announced that he would ask Eisenkot, recommended by 67 members of the Knesset against 46 for Netanyahu, to form a government. Eisenkot has 28 days, which the President can extend by 14; if no one succeeds, Israel votes again. Whoever governs inherits the same capital, the same instruments and the same open question.
Israeli capital is not one thing. Three kinds of capital want different things from the state. Tech and financial capital is mobile, globally listed and dependent on foreign investors and on predictable law. Security capital grew out of the state’s military institutions. State-dependent capital earns its returns from political allocation. In 2023 all three wanted something from the state at once.
In 2023 mobile capital resisted over predictability. The government brought forward its judicial overhaul in January, and some tech firms announced they would move funds abroad. In March the prime minister dismissed his defence minister for calling for a pause, and the Histadrut labour federation called a general strike with the open backing of leading business figures. That evening the legislation was paused. Owners, managers and workers moved together, the rarest alignment in this essay. In April Moody’s cut its outlook on Israel from positive to stable, citing a deterioration of governance. Capital slowed the reform but did not stop it. The Knesset passed the “reasonableness” amendment in July, and the Supreme Court struck it down in January 2024.
After 7 October the same capital deferred over security, and the rating agencies downgraded the country anyway. Protest gave way to mobilisation, and hundreds of thousands of reservists were called up, many of them from the tech workforce. In February 2024 Moody’s cut Israel’s rating for the first time in its history, citing the war’s effect on its political risk, its institutions and its public finances. The finance minister called the decision a “political manifesto.” This time the price came from bond markets and the agencies, not from the capital that had protested a year earlier. That capital stayed. Israeli bank shares rose sharply through the war, and the Tel Aviv 35 index gained 23.5 per cent in the first half of 2025, as foreign investors put in about NIS 9.6 billion.
Security capital grew out of the state’s military institutions; it does not stand outside them. Elbit is privately owned and publicly listed. Israel Aerospace Industries and Rafael are state-owned. The army sorts and trains the country’s technical talent, and its intelligence units have produced a generation of founders: Wiz, which Google bought for $32 billion, the largest exit in the country’s history, was founded by four veterans of Unit 8200. Beneath all of it sits US military aid, $38 billion over ten years under the 2016 agreement. This then is a rent that comes from geopolitics, not from taxation.
State-dependent capital earns its returns from political allocation, not from the market. This means settlement construction and infrastructure, and the stipends and draft exemptions that support full-time religious study. In June 2024 the Supreme Court ruled that the exemption had no legal basis. In July 2026 the outgoing coalition answered by passing a Basic Law declaring Torah study a foundational value of the state, which the opposition condemned; Yair Golan of the Democrats promised to repeal the coalition’s laws, and a petition asks the High Court to strike it down. This capital’s returns depend on who forms the next coalition, and its political representatives may help decide that: a government without the Arab-led parties would need ultra-Orthodox support, and the exemption is what those parties have always asked in return. The election matters more directly to this capital than to the other two kinds.
In the West Bank, where two legal systems apply to two populations in the same territory, the capital that operates there has accommodated the arrangement. Palestinians live under military orders. Israeli settlers carry Israeli civil law with them. In July 2024 the International Court of Justice, in an advisory opinion, found Israel’s continued presence in what the court calls the Occupied Palestinian Territory unlawful; Israel rejects the opinion. Israeli banks finance settlement housing and serve the settlements, and they argue that Israeli law obliges them to serve those customers. In 2020 the UN human rights office published a database of 112 businesses it found to be connected to settlement activity, 94 of them Israeli, among them the country’s largest banks. Accommodation by capital requires no sympathy, only that the numbers work, and here they have.
The price is being set from both sides at once. In August 2025 Norway’s sovereign wealth fund excluded five Israeli banks, citing their role in financing settlement housing, along with Caterpillar. More than 30 US states, meanwhile, have laws penalising companies that boycott Israel. Two episodes show both prices at work. In 2018 Airbnb said it would remove listings in settlements; after lawsuits it reversed the decision the following year. In 2021 Ben & Jerry’s board ended sales in what it called the Occupied Palestinian Territory. Several US states divested from its parent, Unilever, which sold the Israeli business to its local licensee, and the two companies ended up in court. In each case campaigners priced staying, governments priced leaving, and the people inside the firm disagreed about which price mattered.
Capital’s decision in Israel is still evolving. In our four other examples the politics followed the price once capital had settled on the outcome. In South Africa lenders priced apartheid as unaffordable. In Northern Ireland business priced peace above conflict. In Catalonia the banks priced independence as a risk to their depositors. In Cyprus capital priced partition as tolerable, and partition endured. In Israel capital is trying to price several things at once, through different instruments, and the prices do not yet concur. We do not yet know which one will prevail: a rating, a fund’s exclusion list, a state’s anti-boycott law, a court ruling or a coalition agreement.
Capital Prices the Status Quo
Capital took no side in any of these cases; it was the test each side had to pass. The Machinery of Legitimation Part II ended on that claim: capital is not on any side, it is the mechanism by which every side gets tested. The five cases show what the test looks like when nationhood itself is the question. Capital prices the status quo, and politics moves when the price changes. Where capital left, as in South Africa and Catalonia, the status quo became unaffordable and politics moved. Where it adapted, as in Cyprus, the status quo stayed affordable and nothing moved. Northern Ireland sits between the two: capital was disciplined, then offered a dividend for change.
By “prices” we mean something narrower than a market quote. We mean the cost to capital of carrying on under a given arrangement: what it pays to borrow, which contracts it can win, what it can sell and to whom, and what legal exposure it carries. Capital responds to that cost by staying, leaving, returning or speaking up. No one sets the price. Different owners, managers, shareholders and workers read it differently, and it moves.
Every case ran on pressure, structure and opportunity, in a different mix each time. Pressure raised the cost of the status quo: the townships and the divestment campaigns, insolvent banks, a referendum, a judicial overhaul and a war. Structure decided what was possible: the single market, ECB supervision, non-recognition, a Green Line, two legal systems in one territory. Opportunity was the move left open to capital: a change of address that cost almost nothing, a flight to Lusaka, dual market access under Windsor.
Part II’s four moves all appear here, and a fifth appears with them: return. Exit: Chase and Barclays, and the Catalan registered offices. Midwifery, though not of the kind Part II described: at Lusaka and in the Group of Seven, capital helped bring a settlement into being rather than a coup. Accommodation: apartheid’s capital before 1985, both sides of the Green Line, and the West Bank. The state turning on capital: the Cyprus bail-in. Return is the move Part II never asked about, and it behaves differently from exit. Exit is fast, collective and justified by risk. Return is slow, selective and justified by each firm’s own advantage. Sabadell went home in the middle of a takeover fight, CaixaBank stayed in Valencia, and South Africa’s largest companies came back to a new order and moved their listings to London.
Mobile capital exits; capital that cannot move uses its voice. Hirschman’s distinction maps onto the cases almost exactly. Banks holding short-term loans, and firms whose registered office is only an address, could leave in days, and did. Anglo American could not move its gold mines, so it flew to Lusaka, and later bought the stake Barclays sold. Northern Ireland’s employers could not move their customers or their workforce, so they lobbied. In 2023 Israel’s tech sector did both at once, threatening to move money abroad while marching and striking at home.
Look closely at these cases and it is the legal structures beneath capital that carry each transition. Loan agreements, codes of conduct, statutes, treaties, court rulings, emergency decrees, registered offices, listings, licences, a bail-in, a framework. These are the agreed arrangements without which capital’s choice could not function. They were made by people, and they could be made differently.
Pressure as price: morality is rarely decisive until it becomes a cost, and the cost can be set from either side. Sullivan’s principles were a moral project, and on their own they did not end apartheid; they mattered once divestment made South African exposure expensive to hold. MacBride worked through pension funds, and the 1989 Act through public contracts. In Israel campaigners make staying expensive while state legislatures make leaving expensive, and a single firm can face both prices at once.
The agents inside capital do not always agree. Relly went to Lusaka against his government’s wishes and to criticism at home. Barclays’ board faced protests outside its British branches while its South African subsidiary’s prospects worsened. In Northern Ireland employers and unions spoke together, and in Israel in 2023 workers struck with business leaders’ backing. Ben & Jerry’s board and Unilever ended up in court. In Catalonia the La Caixa Foundation went home while the bank in which it holds the largest stake stayed away. “Capital” is a convenient word for many hands, and the hands do not always pull the same way.
These five cases are illustrative, not exhaustive. Other places would serve, and each case here could fill an essay of its own. We have not tried to catalogue every event behind each exit and each return, only to show the mechanism at work in enough settings, and different enough ones, for the reader to recognise it elsewhere.
Methodology note: this piece was written collaboratively between a human and an AI — the human providing the instincts, provocations, editorial judgement and voice; the AI providing research synthesis, intellectual scaffolding and drafting. Full method at [athomehefeelslikeatourist.blog].
Leave a Reply