BANK REGULATORS are seldom celebrities. But Ahsan Mansur, the governor of Bangladesh’s central bank, is an exception. Since he took over in August, after an autocratic prime minister, Sheikh Hasina, was overthrown by protests, his job has been to untangle the criminal mess she left behind.
Connected tycoons schemed with military-intelligence agents to loot the banking system, Mr Mansur says, siphoning away $21bn. The spooks helped the tycoons to forcibly take over banks and then issue loans to their new shareholders, which were not repaid, alleges Mr Mansur. On some occasions, he says, agents seized board members from their homes and forced them at gunpoint to resign.
The pillaging of Bangladeshi savings was a brazen example of a global scourge: state capture. This is when the powers and resources of the state are hijacked for the benefit of a few. It is a broader concept than corruption, since it includes acts that are not against the law, but should be. It can involve rewriting rules to benefit insiders, stuffing institutions with placemen, channelling favours to cronies, intimidating businesses into appeasing the powerful, and gutting checks and balances. The aim may be self-enrichment, or strengthening the captor’s grip on power, or both.
The threat appears to be growing. The world’s most disruptive state of late, Russia, is treated as private property by Vladimir Putin. China boasts of progress against corruption under President Xi Jinping. But according to an index created by Daniel Kaufmann of the Brookings Institution, a think-tank in Washington, state capture in China has grown worse (see chart). And the elephant in the room is America, which is at risk of two kinds of state capture: a president asserting powers he shouldn’t have, and moguls such as Elon Musk acquiring unwarranted sway.
Resisting state capture as it advances is hard enough. Reversing it once it has taken hold is an even taller order—but not an impossible one. Mr Kaufmann’s index offers a reasonable yardstick. It combines measures of the rule of law, the influence of special interests, the extent of corruption and also inequality, on the assumption that captured states funnel goodies to the few. It gives a snapshot only every three years, so there is much that it misses. But it suggests that changes at the top can make a big difference.
Ukraine got better after tossing out a Kremlin-backed kleptocrat in 2014. Conversely, Russia got much worse as Mr Putin became all-powerful: its score of 90 on Mr Kaufmann’s index is now matched only by a handful of gangster states.
Occasionally a reversal of state capture is sudden and revolutionary. In Estonia, after the collapse of the Soviet Union, a predatory communist dictatorship was swiftly replaced by a transparent, democratic system. Free competition replaced an “economy of favours”. But in most places, improvement is gradual and fragile.
Three more recent examples—Bangladesh, Poland and South Africa—offer a mixture of hope and caution. Each has ejected a leader or ruling party that was engaging in state capture: the autocratic Sheikh Hasina in Bangladesh in 2024, the Law and Justice (PiS) party in Poland in 2023, President Jacob Zuma of South Africa in 2018. Each country now has a government that is either determined to reform (Bangladesh, Poland) or at least has strong elements eager to do so (South Africa).
Several factors will affect their chances of success. Christopher Stone, a professor of public integrity at Oxford University, says they should not try to punish every wrongdoer. Prosecutions are time-consuming, costly and often fail. They should make examples of a few ringleaders, but focus on fixing rotten institutions.
When reforming, remember the “15-70-15” rule. Richard Pennington, a police reformer in New Orleans, reckoned that 15% of cops were the drivers of corruption, 70% went along with it and 15% were clean. This ratio applies at many institutions, argues Professor Stone. The best approach is to sack or sideline the filthy 15%, promote the honest staff and persuade the complicit middle that norms have changed. This can be done institution by institution, allowing measurable progress, which can generate support from voters for further reforms.
Bangladesh’s interim government is aiming at big fish, such as Sheikh Hasina, the former prime minister, who is fighting extradition from India. But it has resisted pressure from student leaders to ban her party, the Awami League. Instead it is planning reforms of the police, courts and other institutions captured by Sheikh Hasina’s cronies, before holding elections in late 2025 or early 2026.
At the central bank Mr Mansur also favours reform over revenge. He has blocked the sale of shares in a dozen banks, replaced their boards and begun a comprehensive audit. Meanwhile, he hopes to build a new financial architecture to boost business confidence and limit future opportunities for state capture. Proposed reforms would protect bank depositors and make political interference harder.
In South Africa state capture was worst under Mr Zuma, who gutted watchdogs and enabled the looting of public funds. The ruling party, the African National Congress (ANC), narrowly voted to sack him in 2017, but efforts to hold him accountable for corruption stalled. When he was briefly jailed in 2021, his supporters rioted and brought the country to a standstill, costing perhaps 0.8% of GDP.
Cyril Ramaphosa, Mr Zuma’s successor, oversaw a clean-up of the tax authority and, with less success, the prosecution service, which had both been captured. Further momentum for reform came in 2024, when the ANC lost its majority and formed a coalition with the Democratic Alliance (DA), a liberal outfit with a relatively clean record. The DA now controls six ministries, including home affairs and public works.
Leon Schreiber, the young DA politician who runs home affairs, says it was in shocking shape when he took over. Crooked staff sold fake IDs, enabling widespread benefit fraud. Honest South Africans could not get documents quickly, but dozens of Libyan militiamen somehow received student visas to come and get private military training in South Africa.
Mr Schreiber started by sacking some of the worst offenders. “These are mafias,” he says. “The official who swaps a photo around for someone to get an ID that they don’t qualify for—they are not doing it [just] once.” Mr Schreiber talks of “consequences” for the most corrupt, “encouragement for those who…want to do the right thing, but have never felt safe doing it”, and clarity that graft will not be tolerated.
Digitisation helps, too. “We’re still issuing IDs from the 1960s with photos you can manually switch!” laments Mr Schreiber. He says he is trying to remove human discretion wherever it is not needed.
Pessimists note that the coalition is shaky and could easily collapse. Also, a policy of “black economic empowerment”, recently berated by Mr Musk, provides a smokescreen for politicians steering favours to cronies, and is not about to be abolished. Optimists retort that, since the king of state capture, Mr Zuma, was sacked, South Africa has cleaned up tax collection, identity and some public-works projects. This is not nothing.
State capture in Poland involved less financial corruption than in Bangladesh or South Africa, but was instead mainly an energetic power grab. Between 2015 and 2023 Poland’s ruling party, PiS, used illicit methods to gain control of the courts and the public media. Now a liberal government under Donald Tusk, a former eurocrat, is trying to undo the damage.
Prosecutions are not the most important tool. The new government has brought legal cases and launched parliamentary inquiries into abuses under PiS, from the improper use of public funds to unlawful snooping. Critics say these have made little headway, but “that is the nature of criminal trials,” shrugs Adam Bodnar, the justice minister.
More progress has been made with state media firms, but using questionable methods. When the new government tried to purge them of PiS propagandists, Poland’s president, Andrzej Duda, a PiS ally, vetoed their budgets. So the liberals got creative, putting the state media companies into liquidation and appointing interim bosses—who still run the show.
Reforming the courts has been even trickier. PiS stuffed the constitutional court with supporters, using illegal methods. It also asserted unconstitutional control of the body that appoints judges, and set up a mechanism to punish judges who questioned its nobbling of the legal system.
The biggest obstacle to reform is Mr Duda. When parliament passed two bills to restore impartiality to the constitutional court, Mr Duda referred them to the court itself, which sat on them. Reform will get easier if Mr Duda’s camp loses the presidential election in May, as polls suggest they will. But the window may be narrow. The next parliamentary election is due by 2027, and voters may tire of Mr Tusk’s government if it does not show more results. In the longer run, Stanley Bill of Cambridge University sees a trend towards “majoritarian democracy”, where successive governments invoke exceptional circumstances to justify intrusions into institutions. Norms may erode, he fears.
In many countries, would-be state captors tell a rousing story: that the deep state is a conspiracy, so a strong leader must smash it. Defenders of the rule of law often lack a similarly compelling tale—the safeguards that ensure institutional independence are typically arcane and dull. Somehow, they must find a counter-narrative, argues Mr Kaufmann.
“Public anger at state capture is really important to any effort to dislodge it,” agrees Professor Stone. Patriotism can help. Estonia’s reforms were popular partly because people saw them as part of a revolution against Russian imperialism. South Africans rallied against state capture partly because some of the culprits were foreign. (Mr Zuma’s bagmen were Indian-born.) But nationalism can cut both ways. Mr Putin and his imitators muzzle foreign-funded anti-corruption NGos on the grounds that they are “foreign agents”.
All the evidence suggests that reversing state capture is extremely hard. Far better not to let it take root in the first place. ■