The Capture Index ranks 25 countries across eight regions, three from each region except a four-seat South Asia home panel, and excluding the UN Security Council veto powers, by capture severity across the book's ledgers. Four dimensions are measured with verified 2024 to 2025 data: public money from Transparency International, rule of law and government effectiveness from the World Bank, and press freedom from Reporters Without Borders. The composite uses three; government effectiveness is shown but held out because it duplicates rule of law.
A measure of how far a republic has been captured: the distance between the rights it promises on paper and the access a citizen can actually use without a patron. Corruption indices score perceived bribery. This scores capture.
A higher score means deeper capture, so the desired direction is down. Read alongside the Diwan Index, the companion measure of recovery machinery, where the desired direction is up.
Three countries from each of eight regions, with South Asia, the book's home region, carrying a fourth seat, and the UN Security Council veto powers left out. Five follow the World Bank's funding-agency grouping. The World Bank's single Europe and Central Asia region is split into Europe, Eastern Europe, and Central Asia, since one region cannot fairly stand in for Denmark, Serbia, and Kyrgyzstan at once. Capture severity runs 0 to 100; higher means a wider gap between formal rights and usable access. Select a country to open its ledger.
Two of the four South Asian seats are the only places in this set where capture ran past diagnosis into rupture. Sri Lanka in 2022, Bangladesh in 2024. Both crowds removed the captured ruler. Neither, by removing him, removed the capture. That gap is what the book is about, and it is why the index carries a recovery twin.
The crisis read as bad luck and was not. A government cut taxes in 2019 and gutted its own revenue, borrowed abroad until the foreign debt passed the size of the whole economy, and imposed an overnight ban on synthetic fertilizer that collapsed the rice and tea the country lived on. In April 2022 Sri Lanka stopped paying that debt, the first sovereign default in its history. A captured fiscal ledger, revenue surrendered for political reasons and a flagship policy imposed with no institutional check, is the engine underneath every empty fuel queue.
The Aragalaya answered the queues. For four months crowds held Galle Face, forced out the finance minister and then the prime minister, and on the ninth of July walked into the President's House and took it. Gotabaya Rajapaksa fled to the Maldives, then Singapore, and resigned by email, the first Sri Lankan president to leave office mid-term.
Then the machinery reasserted itself. Parliament, still dominated by the Rajapaksa party, installed Ranil Wickremesinghe, a six-time prime minister and the establishment's own man, as president. He cleared the protest camps, arrested organizers, and delayed the local elections that would have measured the public mood. The ruler was gone, and the system that produced him chose his successor. Structural change waited two more years, until the 2024 ballot put Anura Kumara Dissanayake in the presidency and then handed his coalition 159 of 225 seats while the Rajapaksa party fell from 145 seats to three. Two years and a second mass verdict separated the palace falling from the system turning.
The spark was a job quota. Courts had reinstated a rule reserving 30 percent of government posts for the descendants of 1971 freedom fighters, which in practice meant posts handed out by lineage and loyalty rather than merit. A captured administrative ledger, public employment run as patronage, is what the students read in that quota, and they were right to. Behind it sat courts leaned on to bring the rule back, a press disciplined by the Digital Security Act and routine internet shutdowns, and a fifteen-year record of enforced disappearances and killings the state never answered for.
The crackdown turned a quota protest into a reckoning. Security forces and the ruling party's student wing fired on demonstrators. Hundreds died, by a later United Nations estimate as many as 1,400. On the fifth of August 2024 Sheikh Hasina resigned and fled by helicopter to India, and a crowd took the prime minister's residence. An interim government under the Nobel laureate Muhammad Yunus took over three days later.
A year on, the machinery is still standing. The Awami League is banned and its leader is in exile, yet the politicized bureaucracy, the partisan courts, the patronage that ran the job rolls, and the security apparatus that did the shooting remain in place, waiting for whoever wins the next election. By the first anniversary the mood had curdled to disillusionment. A student coalition fragmenting, reforms stalled, parties already jostling for the levers the crowd had just pried loose. Bangladesh is still inside the gap Sri Lanka took two years to cross.
An uprising is a diagnosis without a prescription. It reads, correctly, that the republic has been captured. It removes the most visible hand on the levers. It does not, by that removal, dismantle the levers, and so the levers wait for the next hand.
The book's framework supplies what both crowds lacked. Its diagnostic half names which engines were captured, fiscal and public money in Colombo, administrative discretion and justice in Dhaka, so the target becomes the machinery, not the man. The Diwan names what repair looks like, and the list is concrete: open the contract registry, publish the public debt, take recruitment out of party hands, disclose the court backlog, make the right to information answerable on a clock. Conditions to lock in during the short window when the captured elite is weakest, which is the only window when they can be locked in at all.
Activists organized around the ledgers would have entered that window with a transition agenda, not just an exit demand. The registry, the debt, the recruitment rule, written into the handover before a new establishment could settle into the empty chairs. In Dhaka, the depoliticized bureaucracy and the freed press fixed as the price of the interim government's mandate, not deferred to a reform commission that the next winner can outlast.
The framework does not promise a different ending. It names why both endings stalled. Removing a ruler is the visible half and the easy half. Rebuilding the ledgers is the half that decides whether the next government inherits a republic or a machine. The recovery twin exists for exactly this reason. Capture Index for the disease. Diwan for the cure, and the cure was never a better ruler.
The index is built from the book's own framework. It does not invent new primary data. It reframes established international indicators through the capture lens, one ledger at a time.
The Corruption Perceptions Index scores how much bribery experts perceive in a country's public sector. Useful, and used here as one input. It does not capture the book's subject.
Capture is the arrangement where a republic keeps its rights formally intact while making them unusable without help. The law is open. The office receives applications. The court hears cases. The gap appears in delay, discretion, opacity, and fear, the instruments that convert a public duty into a private favor. A captured state does not abolish the right. It makes the right expensive enough that a patron becomes necessary.
So the index measures the distance between de jure right and de facto access across the institutional ledgers where capture operates. The composited dimensions diverge in ways a corruption score alone would hide: South Africa, Brazil, and Ukraine all sit lower than their corruption scores suggest because their press is freer, while Vietnam sits higher because its press is among the world's least free.
Five regions follow the World Bank's geographic grouping, the standard funding-agency framework: Sub-Saharan Africa, the Middle East and North Africa, South Asia, East Asia and the Pacific, and Latin America and the Caribbean. The high-income North America region is omitted, since removing the United States leaves only Canada.
The World Bank's single Europe and Central Asia region spans too wide a governance range for three countries to represent, so it splits into three. Europe takes one country each from the North, South, and West: Denmark, Italy, Germany. Eastern Europe takes one each from the Balkans, the Visegrád group, and the post-Soviet Slavic space: Serbia, Hungary, Ukraine. Central Asia stands on its own: Kazakhstan, Uzbekistan, Kyrgyzstan. The split earns its place in the data. The three regions average 20, 53, and 71, a clean eastbound capture gradient the lumped region would have hidden, and Central Asia reads as severe as South Asia.
The five UN Security Council veto powers are excluded by design. Germany takes the Western Europe slot France would otherwise hold, and Ukraine the post-Soviet slot Russia would hold. An index that ranked the United States, United Kingdom, France, Russia, or China would invite a sovereignty deflection and would not fit the funding-recipient frame the index is built for. Denmark stays in as the low-capture reference at the clean end.
South Asia carries a fourth seat, the only region that does. The book is anchored there, forty chapters deep, and a home region earns resolution the others do not need. That fourth seat also buys the one exhibit no other region can show: a complete capture lifecycle, from entrenched capture in Pakistan, through capture inside a capable state in India, to the two ruptures where capture ran to collapse, Sri Lanka in 2022 and Bangladesh in 2024. The asymmetry is deliberate and labeled, not an accident of counting.
Each ledger maps to a chapter cluster in the book and to a named, institution-trusted indicator. Four are measured. Two need building.
| Ledger | Book | Indicator source | Status |
|---|---|---|---|
| Land & inheritance | Ch 11, 23, 36 | Bespoke: PRINDEX tenure security · OECD SIGI inheritance | bespoke axis |
| Fiscal sovereignty | Ch 12, 35 | Bespoke: general-government effective tax · revenue disclosure | bespoke axis |
| Public money & procurement | Ch 13, 14, 15 | Transparency Intl CPI 2024 | composite |
| Justice & rule of law | Ch 16 | World Bank WGI Rule of Law 2024 | composite |
| Administrative discretion & local voice | Ch 17, 18, 19 | World Bank WGI Govt Effectiveness 2024 · 0.93 with justice | measured |
| Narrative freedom | Ch 20 | RSF World Press Freedom Index 2025 | composite |
The administrative ledger was added and measured for all twenty-five, using the World Bank's Government Effectiveness indicator on the same scale as Rule of Law. It correlates with the justice ledger at 0.93. Government Effectiveness restates rule of law rather than adding a fourth reading, so it is shown on each country card but held out of the composite, which stays on the three least-redundant measured ledgers. Folding in a measure that duplicates one already counted would double-weight governance perception and inflate the appearance of rigor.
Fiscal sovereignty hit a wall that off-the-shelf data cannot clear. The only clean cross-country tax series is the World Bank's central-government figure, and it books Germany at 11 percent and India at 6.7 percent because their states and social funds collect the rest. A measure that makes a federal democracy read as a tax haven cannot carry a capture ledger. General-government effective tax, the figure that would work, is not published comparably for all twenty-five in any single table.
Land met the same wall from the other side. The property-rights indices that are cleanly available are built from judicial independence, expropriation risk, and contract enforcement, which is the rule-of-law ledger under a different name. The book's land question, who holds the title and whether a daughter receives her share, lives in tenure-security and inheritance data that no governance index collects.
The result is the case for the Diwan stated in data. Existing indices measure the governance-perception cluster well, and redundantly. The two axes that separate one capture engine from another, fiscal and land, are exactly the two that have to be built. They are the Diwan's first two instruments.
Measuring capture is half the instrument. The book's reform argument supplies the other half: the machinery that lets a right stand without a patron. The Diwan Index, a companion measure named for the register that once recorded what the state owed its people, will score that machinery. A falling Capture Index and a rising Diwan Index then describe the same progress from two directions.
It will score function, not whether a law merely exists, because the gap between a law on the books and a law that works is the book's whole subject.
Four dimensions are measured from verified 2024–2025 sources, three of them composited and one (administrative) shown but held out for redundancy. Two ledgers, land and fiscal, await bespoke instruments, and no score is published here as final.
The two bespoke ledgers, and the administrative figures' role, require sourced methodology sign-off before this index carries the credibility an institutional citation needs.