On January 11 the World Bank sent letters to Cambodia's Prime Minister, Hun Sen, and its Finance Minister, Keat Chhon, threatening to suspend all Bank lending to their country. Unless the Cambodian government took "urgent steps" to fulfill auditing requirements for and to repay funds illegally obtained through a program designed to downsize the bloated military, almost US$300 million - an amount roughly comparable to half of Cambodia's annual budget - of loans would be frozen in a month.

Just before that deadline and the arrival in Phnom Penh of Bank President James Wolfensohn two weeks ago, the Cambodian government met two of the Bank's demands and agreed to meet the third by mid-May.

On the surface this would appear to look like progress in the quest for transparency and accountability, particularly in a project that has regularly been described as "an embarrassment for the Cambodian government." Yet the failures are hardly exclusively Cambodian. They are also the failures of the Bank, whose insistence on a program of dubious efficacy, and whose opaque management and tolerance of gross corruption, make it equally culpable.

The Bank's "Demobilization and Reintegration Project," which began its public activities in 1999, was supposed to decrease military spending by shrinking the armed forces and transferring the "peace dividend" to desperately under-funded social welfare ministries. The project aimed to remove thousands of unnecessary soldiers from the rolls and assist them in settling back into civilian life. Regional commanders were to submit lists of elderly, ill, or otherwise unnecessary soldiers, and those soldiers would receive a small amount of cash, some household goods, and the means to move into civilian lines of work. A database would be developed to track both current and demobilized soldiers.

But not only were these basic premises flawed, so too was the agenda of the World Bank's Post-Conflict Unit, whose primary interest lay in promoting itself within the Bank. That Unit accepted an obviously inflated number of soldiers from the Cambodian civilian heads of the project, who saw in a larger number an opportunity for extraordinary profit. Dozens of experts on Cambodian politics, military reform, social welfare, and human rights cautioned the Bank against the project, citing concerns about its efficacy and specifically its vulnerability to official corruption. Despite these concerns, the Bank pushed ahead.

Over the next five years, the project encountered every problem predicted by those experts. Funds began to flow before viable auditing mechanisms were in place.

When a Bank project consultant began uncovering evidence of corruption in early 2000, he was the victim of an armed attack; his supervisors did nothing. As clear evidence of fraudulent procurement of motorbikes for the project surfaced in July 2002, the Bank staff member responsible for disbursing funds shrugged it off, saying, "As long as 60 percent of the money gets where it's supposed to go, that's good enough for me." The Bank did not publicly announce the fraud for another year and did not sanction anyone until November 2004. Suggestions regarding greater transparency were routinely ignored, and occasional get-tough meetings and correspondence between the Bank and government rep-resentatives had no effect because the Bank regularly backed down for "fear the project would stop."

A yearlong political stalemate followed Cambodia's July 2003 national elections, and Bank officials allowed the project to stagnate. They claimed nothing could be done until a government formed, though no other Bank projects stood still. When a government did form in 2004, one of its first pieces of draft legislation proposed mandatory military service - a move that clearly showed its disinterest in slimming the armed forces but that elicited no comment from the Bank. Throughout 2004, Bank officials droned on at meetings about the importance of "lessons learned" from problematic projects and even launched a study of several Cambodia projects' vulnerability to fiduciary risks. But somehow the study did not include either the demobilization project or the equally controversial forestry management project.

It is true that the project finally withered because the Cambodian government refused to meet the Bank's demands for greater accountability. Yet this is precisely the resistance that should have prevented the project from ever beginning. That Bank officials have taken so long to respond to these concerns and reflect on their own mistakes calls into question their capacity for swift responses and administrative and financial transparency.

For an organization pur-portedly devoted to alleviating poverty, the Bank's lingering threat to suspend all loans - an action that would invariably affect Cambodia's credit rating and its relationship with other donors - seems not only disproportionate but also positively callous. The country's current economic situation is at best precarious. The expiration on January 1 of Cambodia's pre-ferential access to the US market for garments puts at risk an estimated 20 percent of GNP and a quarter of a million jobs. According to the United Nations Development Programme, an estimated one million Cam-bodians, about a tenth of an overwhelmingly agrarian pop-ulation, are now landless. Socio-economic indicators suggest alarming trends in nutrition and infant mortality. Yet the Bank seems willing to jeopardize the foreign aid that subsidizes most of the meager social services.

Who should suffer as a result of this project's failures? Obviously the corrupt Cam-bodian officials in question should be barred from future projects. If the World Bank only concerns itself with 60 percent of the funds, that is what Cambodia should repay. But given the Bank's complicity, its staff members in Bangkok, Phnom Penh, Singapore, and Washington - those who launched the project, kept it going, and ignored the problems - should also be barred from future projects, and they should repay the remaining 40 percent of the funds. They should match the mid-May deadline for the final auditing documents from Phnom Penh with a full disclosure of all the project documents. For the Bank to shirk these responsibilities while making ordinary Cam-bodians bear the burden is nothing less than grotesque.

Sophie Richardson is an analyst of Cambodian Politics and has worked previously for Human Rights Watch, the National Democratic Institute and the International Crisis Group. This article was originally printed in The Nation on March 1, 2005.