Is the U.N. the Organization the World Needs?
This past week, the U.S. military announced the official withdrawal of its forces from Iraq, hopefully signifying a close to the past decade’s most violent political crisis. This past week, with much less fanfare, Médecins Sans Frontières (MSF) announced that a humanitarian crisis is underway in the Central African Republic that may be associated with the highest mortality of any country in the world. What is most interesting in the juxtaposition of these two crises at opposite ends of the “news attention cycle” is that the UN has actively played a role in hiding the death toll in both. In theory, the UN’s raison d’être is world peace. Thus, the organization’s short-sighted interests in funding itself and appeasing member governments involved in conflicts should have all of us wondering, “Is this UN the organization the world needs?”
In Iraq, a variety of sources have put forward death estimates ranging from around 100,000 to one million (for example, see this Guardianmap or this survey by Opinion Research Business). The lower estimates tend to be tallies or lists acquired through medical sources or newspaper reports, a process referred to as “passive surveillance” in the public-health community. The higher estimates are more often from surveys or polls, which epidemiologists who work in war zones tend to see as more accurate (assuming they are conducted properly). During the first four years of the war, UN references to the number of dead tended to cite Iraqi government passive-surveillance tallies, despite credible surveys suggesting the actual numbers were four to 10 times higher. The highest profile of these was published in The Lancet in October 2006, and estimated that there had been 600,000 violent deaths since the invasion. This article produced considerable criticism from the Iraqi government, and an unusual condemnation from then-minister-of-health Ali al-Shammari, who said the government’s official estimate (<50,000) – the number the UN had been citing – was certainly too low, but that he thought the real death toll could be 100,000, maybe 150,000, but not the 600,000 reported in The Lancet.
The UN’s role in this story went from passive to active when the World Health Organization (WHO) helped minister al-Shammari’s staff plan, analyze, and publish a survey regarding the death toll in Iraq. The report, published in the New England Journal of Medicine (NEJM), was clearly designed to discredit the 2006 Lancet report. While the collected data covered a longer period, the published analysis was limited to the exact recall period of the Lancet article. The Lancet reported a 2.5-fold increase, and the NEJM survey reported a statistically similar 1.9-fold increase – but the overall excess deaths were not reported by the government and WHO authors. The focus of the NEJM study was only on violent deaths, which they estimated to be 151,000. This was almost identical to the upper limit of minister al-Shammari’s estimate from 2006.
Interestingly, after many months of virtually no interest in, or discussion of, the 2006 Lancet article in the U.S. press, on Jan. 4. 2008, the National Journal, whose reporters had interviewed the authors of The Lancet five and seven months before, wrote an article questioning its credibility and perhaps implying that the data had been fabricated. The various criticisms were rather weak and widely questioned, but the article inspired summary editorials in The Wall Street Journal (on Jan. 9, the exact day of the NEJM release), the Boston Globe, International Herald Tribune, and other papers in the days to follow. Thus, the Iraqi Ministry of Health’s data was not just released – it was delivered.
For decades, the world of public health has largely agreed that increases in overall mortality are the best way of documenting the effect of a crisis. The blatant hiding of the 400,000+ excess death toll in the NEJM, in the WHO press releases, and in interviews with WHO staff, is a pattern consistent with the UN’s practice of packaging the message of deaths to please a belligerent host country.
In 2005, the WHO funded and organized a survey regarding conditions in Northern Uganda reporting 1,000 excess deaths per week. The Ugandan president and minister of health argued against the findings, claiming the draft report was flawed and the study needed to be redone – a statement that WHO officials would not publically dispute. Similarly, in Darfur, UN surveys conducted with the Sudanese government produced mortality estimates far below those made by neutral parties. A primary reason the mortality measure was so low is that the Sudanese government only wanted the survey to cover the six preceding months, a period of relatively low violence, while some of the research team wanted to use a more standard one-year recall. In a particularly stunning event displaying how political the issue of mortality in crises can be, the debate over the design of the WHO Darfur study climaxed in a conference call between then-UN-secretary-general Kofi Annan and the director of the Centers for Disease Control and Prevention, Julie Gerberding (a laboratory scientist), to discuss what should be the appropriate recall period.
All things considered, the MSF press release this week suggesting that the death rates in large portions of the Central African Republic (CAR) are two-to-six times the regional norms and the official UN estimates is not a new message. A group from the University of California at Berkeley published a similar finding last year. What is most damning for the UN is that UNICEF requested, helped design, organized, and conducted a survey that found similar results two years ago. The survey, which I was involved in, was designed to document human-rights abuses, and was not a typical design for a mortality survey, producing potentially confusing “sets” of mortality estimates. However, the findings were viewed as controversial within UNICEF primarily because, just months before, UNICEF had, in conjunction with the host government, released a Multiple Indicator Cluster Survey (MICS) suggestinga death rate one-quarter of the rate that our survey arrived at. This extremely expensive MICS survey, conducted with, and by, government employees, is a primary input to the official estimate that the death rate in CAR is 1.4 /1,000 population / month. Thus, our survey (completed at less than five per cent of the cost, over just a three-month period), was clashing with a huge multi-year UN MICS survey process that globally spends tens of millions of dollars per year, keeps governmental partners employed and happy, and justifies the employment of dozens of people. While it was clearly the UN’s responsibility, only MSF took up the task of investigating the report. It is likely that the official UN statistics have underestimated deaths by approximately 300,000 people since the report’s release two years ago – deaths largely from malaria, respiratory infections, and diarrhea, which could have been prevented.
In the past few decades, colleagues have repeatedly told me about negotiations they have taken part in between the UN and individual governments over what the “official” death rates would be. Had these negotiations been about how many children had really been sexually abused versus how many should be admitted to the public, there would be major public outcry and demands of firings and imprisonment when the story broke. But, in the hyper-politicized context of the UN, people are rewarded for placating host governments and keeping their activities funded even if it means the de facto covering-up of thousands of deaths resulting from government incompetence, or, in the case of CAR, worse.
There is a profound global need for multilateral mechanisms for exploring the truth. The UN’s International Civil Aviation Organization inspires little debate or controversy while developing air-safety regulations and accident-inspection guidelines, and while compiling safety statistics, in spite of the contentious and litigious nature of their work. No UN work is more political than that of the International Atomic Energy Agency (IAEA), the nuclear-energy watchdog group that won the Nobel Peace Prize in 2005. Both organizations reliably produce highly regarded technical information amid political firestorms. However, both agencies are largely removed from compromising partnerships with the national governments, and, compared to the rest of the UN, the technical staff are well buffered from the UN leadership and Security Council.
When people die at the hands of their government, either as an indirect or direct result of violence, or from incompetence, the world is disserved by a UN that can find political advantage in hiding or ignoring these deaths. The operational service delivery agencies of the UN cannot operate effectively without close relations with their governmental partners, and thus must constantly be tempted by political advantage. In Uganda, Sudan, Iraq, and CAR, this temptation has served the world and the UN poorly over the past decade. There are two logical ways to address this problem: end all UN relief programs in conflict-affected countries, or create a firewall between violence and suffering-related data-generation and the operational arms of the UN.
While for many, removing the UN from conflict areas seems like abandoning the most needy, the effect of removing agencies like UNICEF and the UN Development Programme from conflict settings would likely be an increase in services to the most underserved. This is because the UN is exceptionally expensive and inefficient compared to local NGOs, international NGOs, church groups, and others that would likely adsorb the funds that are currently being sent to the UN. Adding to the inefficiency is that the UN is exceedingly risk-averse, and will not travel as broadly or reliably in insecure settings compared to the independent relief agencies. That said, there are also noteworthy downsides to the exclusion of operational UN agencies from conflict settings. Certain services such as airline safety and disease surveillance must operate across country boundaries, and unfalteringly over time, in order to produce optimal results. In settings like Libya last summer or Syria today, to cut off programs for just a few months of conflict would result in massive material and achievement losses and strained relations with national partners abandoned in their moment of greatest need.
This leaves the second option – an independent body firewalled from the political influences of the UN and member states, more like the IAEA. An ad hoc process directed by a permanent UN or International Criminal Court staff, but made up of outside experts with no links to the nation under study or the larger UN, may be easier and less expensive to maintain. Whatever the solution might be, the present UN system for detecting excess deaths and suffering is profoundly broken.
The episode of the WHO’s mortality obfuscation in Iraq is particularly problematic from a political perspective. The pre-emptive invasion was declared illegal by Kofi Annan, and was orchestrated by two permanent members of the UN Security Council. Those two countries were not sanctioned, and their leaders have been the most out of step with the facts in downplaying the death toll from the war. Thus, others like the UN who downplay the death toll appear to many as stooges of these countries. In August of 2003, when the UN headquarters in Baghdad were bombed, the UN Security Council called it a “terrorist, criminal act,” and the world’s emotional response largely aligned with that sentiment. Virtually no one in the West considered it a legitimate act of war, or in any way justified. In 1999, NATO bombed a Serbian radio and television station, killing 16 civilians. This past summer, it conducted similar strikes in Libya. These episodes are presumably justified because being a politically driven mechanism that spreads information makes one a fair target. Given how systematically it has acted as a tool of propaganda when it comes to the issue of wrongful deaths in Iraq, CAR, and Uganda, the UN needs to get out of the business of producing conflict-related evidence, and keep at bay the perception that it is a politically driven information-spreading mechanism.
Photo Courtesy Reuters.