The Data Manipulation Scandal That Could Topple the Heads of the World Bank and IMF, Explained

As the annual meetings of the World Bank and the International Monetary Fund kick off next week, the Bretton Woods institutions are mired in scandal. On September 15, the World Bank published an independent investigation led by former US Attorney for Washington D.C. Ronald Machen and the law firm WilmerHale into the involvement of senior World Bank management in data manipulation.

The report alleges Bank officials altered data to benefit both China and Saudi Arabia in its flagship Doing Business report, which ranks countries on their business regulations. The allegations cover incidents spanning the tenures of two World Bank presidents, Obama-nominee Jim Kim and Trump-nominee David Malpass, with a leading role for Kristalina Georgieva who now runs the IMF.

Major shareholders including the US and UK have expressed concern. The Economist magazine has called for Georgieva to step down, and numerous op-eds have stressed the damage to the Bank and the Fund’s credibility.

This saga requires all sides to set aside some ideological prejudices. Doing Business has been the darling of right-wing economists and business-oriented authoritarian governments for nearly two decades. Progressives cheered when the Bank temporarily suspended the index after the first audit reports of data manipulation emerged. Now as it looks like blame may fall on Kristalina Georgieva, who’s been pushing the Fund to do more on climate and more for developing countries, the tables have turned. Eminent progressive economists have labeled this “a coup attempt at the IMF” (Joe Stiglitz) due only to “anti-China hysteria” (Jeff Sachs); Georgieva has won the endorsement of 16 African finance ministers, while big-name conservative economists like Anne Krueger have called for her resignation.

Meanwhile, World Bank President David Malpass has virtually escaped scrutiny, deepening suspicions among Georgieva’s supporters that a plot is underway to scapegoat the progressive woman leading the Fund while Trump’s man at the Bank skates away.

I want to set aside these broader political calculations and focus on the case at hand: the Doing Business scandal. Facts matter, and the credibility of the World Bank and IMF matters beyond current leadership.

So let’s review what we actually know about the data manipulation, how it arose, and who may be to blame, starting at the very beginning.

The Doing Business index steered billions in World Bank lending, and claimed to influence regulatory reforms around the world.

The Doing Business index, first published by the World Bank in 2002, ranked countries on several aspects of business regulation, from the time it took to get a construction permit or clear customs to more controversial measures like the strength of labor rights or corporate tax rates—both of which were coded as unambiguously bad things.

The index grew out of a series of academic papers surveying experts about business regulations around the world, led by Simeon Djankov at the Bank and Andrei Shleifer at Harvard together with several colleagues. The project had a strong anti-regulatory slant from the start, inspired according to Djankov by “the experience of centrally planned economies” in the former Soviet bloc and “the waste of entrepreneurial talent and resources as a result of overregulation.” Methodologically, Djankov claimed inspiration from Hernando de Soto’s popular 1989 book, The Other Path: The Economic Answer to Terrorism—a reference to Peru’s Marxist “Shining Path” guerilla movement—in which de Soto and colleagues catalogued the hassle of trying to register a formal business or get a land title in Peru, arguing that reducing red tape would help capitalism defeat Marxism.

For the World Bank, Doing Business was a blockbuster hit. While scores of Bank reports are literally never downloaded, Doing Business garnered frontpage coverage in major international newspapers. Paul Kagame set up a government unit to maximize Rwanda’s score. Narendra Modi campaigned for reelection in India on the back of India’s rising score. The World Bank boasted of 3,800 policy reforms to Doing Business-related indicators across over 100 countries (though attribution of reforms to the index was always implied more than outright claimed). The Bank itself used the index as the basis for either conditionality or outcome metrics in $15 billion in loans, and its influence on World Bank policy permeated much further than that.

Progressives railed against the index. Democrats on Capitol Hill eventually succeeded in forcing the Bank to drop the anti-labor regulation components of the index in 2011. And after the Rana Plaza factories collapsed in Bangladesh in 2013 killing over a thousand workers, the Bank softened its anti-regulatory stance to construction and added positive points for enforcing safety standards. But these political fights are not what eventually brought down Doing Business.

More relevant to the current scandal is another long-standing criticism of Doing Business, i.e., that it measured only de jure regulations, not their de facto enforcement. More accurately, it only asked hypothetical questions about the challenges facing a hypothetical firm which complied with every imaginable bureaucratic procedure. And it posed those hypotheticals to a group of experts in each country, not actual businesses. That made Doing Business quite subjective, and difficult to fact check. Expert responses weren’t blindly aggregated, they were debated and curated by World Bank staff, with input from country representatives. We’ve never seen any respondent-level data, and it’s unclear it even exists as such. Of course, a subjective process can and should still be impartial. But the process opened the door for the political interference that led to the current scandal.

The scandal arrived with a bang, but the World Bank strategy of flat denials seemed to work, at least for a while.

In 2018, World Bank Chief Economist Paul Romer announced political interference had tainted the Doing Business rankings. Two weeks later he was gone.

In a bombshell interview published on January 12, 2018, Paul Romer—then Chief Economist of the World Bank and future Nobel laureate—told the Wall Street Journal that he had lost faith in the integrity of the Bank’s Doing Business index. He lamented that the “political motivations of World Bank staff” had “tainted” the index, and pledged to go back, correct and reissue rankings for the previous four years. He apologized to the people of Chile, whose wild swings in the rankings he said had been driven by methodological changes at the Bank, not actual policy changes.

Kristalina Georgieva, the Bank’s CEO (i.e. No. 2) at the time, fired back publicly and denied the accusations. A few days later, Romer recanted some of his comments, clarifying that “I have not seen any sign of manipulation of the numbers published in Doing Business report or in any other Bank report.” That said, while there was no proof of political motivation, it appeared that his concerns about oddly timed methodological changes driving Chile’s numbers were well founded, and applied elsewhere as well, including India.

Twelve days after making his first accusations, Romer resigned from the World Bank on January 24, 2018. And for a few years the issue went away. Until another whistleblower within the Doing Business team came forward in 2020, provoking a chain of investigations culminating in the latest WilmerHaler report.

It turns out there may have been more to Romer’s accusations than perhaps even he realized at the time.

The latest investigation outlines detailed “factual findings” about Kristalina Georgieva’s personal management of data manipulation to benefit China.

The WilmerHale report offers a day-by-day timeline full of geopolitical intrigue.

In early October 2017, Chinese representatives at the Bank had started pressing the issue of their upcoming Doing Business score. China had some financial leverage over the Bank, which was about to launch a campaign for a capital increase and start fundraising for its concessional finance window, but as Jeff Sachs and others note, it’s possible to overstate this leverage.

In any case, then-president Jim Kim asked the Doing Business team for an update, and they reported that China would drop 7 places in the 2018 rankings. Officially, the report was signed off to go to the printers on October 16 with China ranked 85th. But Bank management pressed on.

Two days later, CEO Kristalina Georgieva called a meeting with the Doing Business team and took charge of the China issue. She nixed a proposal from President Jim Kim’s staff to merge Hong Kong with China in the analysis. Instead, the group settled on a plan to use the Chinese data from only one of the two cities surveyed, Beijing or Shanghai, rather than both.

The next day, October 19, the senior director with oversight of Doing Business, Shanta Devarajan, emailed Georgieva to say that the “best city” plan wouldn’t work, as that methodology would help other countries too.

That same day, Georgieva tasked Simeon Djankov, one of the founders of Doing Business, with finding a solution in the underlying data. Djankov worked with his team to identify three data points that involved sufficiently ambiguous judgment calls that they could justify altering the data. Switching each of these three points in China’s favor boosted its overall ranking by 7 places back to an overall rank of 78th, the same as the previous year.

Georgieva summoned Devarajan to her office the next day. He told her he could live with the “judgment calls” to accommodate geopolitical considerations, and Georgieva thanked him for doing his “bit for multilateralism.”

Georgieva’s claim that this was a case of innocently double-checking data is contradicted by three separate internal and external investigations.

Georgieva has flatly denied the WilmerHale account. “I would ask staff to please check, double-check, triple-check, but never change, never manipulate what the data tells us,” she told an IMF staff townhall. Shanta Devarajan has similarly contested the WilmerHale portrayal, writing in a Twitter thread that Georgieva’s instruction “was to verify the China numbers, making sure that China received credit for the reforms they undertook, without compromising the integrity of Doing Business.” I should note that I have known Shanta Devarajan for several years, he’s a non-resident fellow at CGD, and I take him at his word here.

There are two different ways to read these denials: (a) there was no data manipulation, just innocent judgment calls, or (b) data manipulation happened, but nobody heard Kristalina Georgieva personally order it.

The first interpretation is suggested by two former World Bank chief economists, Nick Stern and Joe Stiglitz. But the idea these were legitimate judgment calls is contradicted not just by the WilmerHale report, but by two separate, previous reviews (not to mention at least two other former chief economists). A December 2020 management review by the bank’s own Development Economics Vice-Presidency concluded that the last-minute changes to China’s 2018 data and Saudi Arabia’s 2020 data, among others, “required correction” and were “not justified by the Doing Business methodology or by any new information provided to the Doing Business team.”

Furthermore, management called the shots. The World Bank Group Internal Audit office confirmed the findings of manipulation and noted that “DB team members reported undue pressure, both directly and indirectly by Bank management to manipulate data in 2017 during the DB18 production process [which involved manipulating China’s score] and in 2019 during the DB20 production process [where Saudi Arabia’s score was similarly manipulated].” The Auditor General goes on to note that “out of the 15 staff in the DB team interviewed by [World Bank Group Internal Audit], 9 staff indicated that they had been directly or indirectly pressured to manipulate data.” Eight complied.

Do we know Georgieva personally ordered staff to manipulate the data to help China?

Note that the WilmerHale report never claims that Georgieva uttered any such words. The allegation is that at her direction China was singled out for special attention after the report was finalized; this was done for political not technical motives; multiple alternative methodologies were tested at her request; those methodologies were discarded if they didn’t generate the “right” answer; only when the team got a result that improved China’s ranking was the report approved; and Georgieva personally oversaw each step of this process.

In short, Georgieva didn’t have to blurt out “just make it up!” (or anything close) to be responsible for manipulating data. Indeed, that’s rarely how data manipulation works.

The WilmerHale report documents equally troubling data manipulation to benefit Saudi Arabia, and leaves a lot of unanswered questions about David Malpass’s role.

The factual record in the case of Saudi Arabia is quite a bit murkier.

There is no case, at the moment, to accuse David Malpass of data manipulation, beyond the general notion that the buck stops at the top. But there are questions that shareholders should be asking during the Bank-Fund annual meetings.

Like China, Saudi Arabia was also being courted to contribute more to the World Bank’s concessional lending arm, and the Saudi’s had a unique fee-for-service relationship with the bank to improve their business climate score, known as “Reimbursable Advisory Services,” posing a potential conflict of interest flagged by auditors.

In the fall of 2019, Mohammed bin Salman (MBS) was looking to repair his image. For nearly a year after the murder of journalist Jamal Khashoggi in October 2018, the crown prince had gotten a cold shoulder from global leaders. Now Saudi Arabia was about to launch the initial public offering for its state oil company, Saudi Aramco, slated to be the largest IPO in world history. MBS needed some good business press.

The annual “Davos in the Desert” extravaganza in Riyadh in December 2019 was designed to announce that Saudi Arabia was back. The previous year, US Treasury Secretary Steven Mnuchin and World Bank president Jim Kim had cancelled. But in 2019 the US delegation was to be led by Jared Kushner, who would be joined by Mnuchin and former Trump campaign surrogate and now World Bank president, David Malpass.

On the eve of the Riyadh event, the World Bank launched the 2020 Doing Business rankings, and declared with much fanfare that the top reformer in the world was none other than Saudi Arabia.

The only problem was that the data had been manipulated (again).

The WilmerHale report notes that the original calculation had Jordan coming out as the top reformer, but Simeon Djankov—the Doing Business founder who was also at the center of the China scandal—ordered the team to find a way to alter the data to knock Jordan out of the top place, and to add points to Saudi Arabia’s score. The method used to help Saudi Arabia inadvertently helped the UAE as well, but didn’t affect its overall ranking.

The motive for this change, and even who ordered it, is contested among the individuals interviewed. But the WilmerHale report does note that “we identified no evidence suggesting that the Office of the President or any members of the Board were involved in the data changes affecting Saudi Arabia and UAE in the 2020 report.”

The question that remains is whether they tried to find out. If you search for Malpass’s name in the report, it’s nowhere to be found.

In the case of Jim Kim and Kristalina Georgieva, it is apparent that the WilmerHale team had incredible levels of access to people and electronic communications, and even building-access records to determine who attended which meeting. But none of that seems to have been brought to bear in the case of Saudi Arabia and Malpass, to which WilmerHale devotes less than a page of its report. Did WilmerHale interview David Malpass? Did they interview his chief of staff and other aides? Did investigators have access to his emails or phone records?

By this point Kristalina Georgieva and Shanta Devarajan had left the Bank. Simeon Djankov remained, and Malpass was clearly engaged with the Doing Business issue. Bloomberg revealed last week that he plotted methodological changes that would drop China in the rankings, and he was exchanging correspondence with congressional Democrats (via Secretary Mnuchin) about reforms to the index. How deep his role extended, we don’t know.

For now, all we know is that MBS got away with murder, then launched the world’s largest IPO, and in between the World Bank delivered him a PR gift built on fake data. But so far David Malpass hasn’t had to answer any hard questions on the matter.

Doing Business is dead. Whether Georgieva or Malpass face any accountability likely hinges on the US stance at next week’s annual Bank-Fund meetings.

On September 16, the day the WilmerHale report came out, the World Bank announced it was discontinuing Doing Business.

Beyond that, what are the chances of any real accountability for either Georgieva or Malpass?

The White House and Secretary Yellen hold the key votes. The Europeans are unlikely to sacrifice the European head of the IMF for fear of jeopardizing their monopoly on the managing director’s office. Many possible coalitions of smaller shareholders contain countries which are financially vulnerable vis-a-vis the Fund, and hardly in a position to lead an assertive push for accountability.

On the World Bank side, since the current occupant of the president’s office is a Trump appointee, the White House and Treasury are unlikely to shed many tears over him. The administration’s international agenda around climate change and global vaccination efforts would likely be advanced if he had to be replaced.

In both cases then, the Biden administration and congressional Democrats have little to lose and lots to gain from an impartial pursuit of the truth. But it’s unclear if the administration is willing to expend the energy to unseat either or both of them right now, especially without an Undersecretary of Treasury in place, who would normally lead US engagement on leadership at the Fund and the Bank.

On both principled and pragmatic grounds, calls to ignore deliberate manipulation of data because one or the other official agrees with someone’s policy preferences should be ignored. Impunity for misdeeds by the heads of the World Bank and IMF will set back multilateralism. Progressives should be willing to let one of their own fall at the IMF if the factual record as we currently understand it holds up. And similarly, they should be demanding to know a lot more about what happened with Saudi Arabia under David Malpass’s watch.

Beyond that, what can the Bank do to restore its credibility? Paul Romer has proposed outsourcing the research department entirely. Devesh Kapur and my colleague Arvind Subramanian have a forthcoming op-ed arguing this scandal reinforces the case for meritocratic hiring, and the end to the US-EU deal to monopolize leadership of the Bank and Fund. In a recent external panel review of Doing Business chaired by Mauricio Cárdenas, a group I participated in proposed a variety of more modest internal safeguards and transparency measures. But those measures may be little match for deliberate manipulation ordered from the very top. For that, there seems to be little alternative but to gather all the facts and hold those responsible accountable.