Global Terrorism Statistics Debated

Russian servicemen inspect the wreckage of a plane that crashed after taking off from  Moscow  in August 2004, along with another plane, after explosives on board were detonated.
Russian servicemen inspect the wreckage of a plane that crashed after taking off from Moscow in August 2004, along with another plane, after explosives on board were detonated. (By Sergei Karpukhin -- Reuters)
By Susan B. Glasser
Washington Post Staff Writer
Sunday, May 1, 2005

Late on the night of Aug. 24 last year, two Russian airplanes disappeared nearly simultaneously from radar screens not long after taking off from a Moscow airport. Both crashed when Chechen women blew up explosives hidden on board, killing nearly 100 people in the first multiple-plane terrorist incident since the Sept. 11, 2001, attacks in the United States.

But the U.S. government considers only one of the downed planes the result of an international terrorist attack, because two Israeli citizens were on board one of them while the other explosion killed only Russian passengers. It was, said the senior intelligence official responsible for compiling the U.S. statistics, "the poster child for what is wrong" with the annual report monitoring global terrorism that the United States has put out since the 1980s. "It simply makes no sense," said John O. Brennan, acting head of the new National Counterterrorism Center (NCTC).

For years, the statistical annex attached to the State Department's annual "Patterns of Global Terrorism" report was a respected but low-profile government publication, pored over by academics, debated by a small circle of policy wonks, a "seminal" if flawed work in the opinion of one specialist who regularly used it. But that was before Sept. 11. Now, the report has become so controversial, its findings so politically charged, that for the second year in a row the State Department has been involved in an embarrassing public dispute over the statistics.

Last year, the department was forced to withdraw the report and admit that its initial version vastly understated what turned out to be a record high number of terrorist attacks. This year, government analysts determined that attacks had gone up once again -- three times more, in fact, to a high of 651 attacks that resulted in 1,907 deaths. Rather than publish that information, the State Department decided to strip the annual terrorism report of the numbers and hand responsibility to Brennan's new NCTC. Faced with an outcry once the redacted statistics showing a surge in terrorism leaked out, the NCTC last week released the numbers, but then said the methodology that produced the statistics was so hopelessly flawed the numbers should not be relied upon to make any conclusions.

All of which has left official Washington debating a key question: How to measure progress, or lack thereof, in the Bush administration's war on terrorism, now that the government's top analysts have deemed their own report unreliable? President Bush, quizzed on the apparent upsurge of global terrorism at his prime-time news conference Thursday, attributed the increase to aggressive U.S. action. "We've made the decision to defeat the terrorists abroad so we don't have to face them here at home," he said. "And when you engage the terrorists abroad, it causes activity and action."

But other top officials dispute that there has been any measurable increase at all in terrorism, saying the increase is mostly a result of more aggressive effort by the NCTC to identify terrorist acts and include them in the statistics. "It doesn't tell us anything about the war on terror," said Philip D. Zelikow, counselor to Secretary of State Condoleezza Rice. Zelikow, who served as executive director of the Sept. 11 commission that investigated the attacks on the United States, recommended that Rice pull the statistics from the report. They "are simply not valid for any inference about the progress, good or bad, of American policy," he told reporters at a hastily called news conference to release the already leaked numbers.

Congressional Democrats have assailed that interpretation, pointing out, as Rep. Henry A. Waxman (Calif.) did in a letter to Rice on Friday, that no fewer than five other reports on terrorism in 2004 also found sharp increases. A new Congressional Research Service survey, for example, concluded that al Qaeda and its associated groups launched seven attacks that killed at least 220 in 2004 -- up from four attacks killing 104 in 2003. The NCTC may have made a more rigorous attempt to include terrorist acts in its report, Waxman wrote, but "it seems inconceivable that the administration missed two-thirds of the international terrorist attacks that occurred in 2003."

Amid the small circle of analysts inside and out of the government who have followed the terrorism numbers for years, the controversy has revived debate over how to compile useful statistics.

"Good riddance to meaningless rubbish," is the way security analyst Anthony H. Cordesman greeted the decision in an e-mail circulated around Washington, in which he highlighted numerous problems with a report that never found a useful way to distinguish between a "freedom fighter" and a "terrorist," overemphasized attacks on U.S. interests, and made arbitrary distinctions, such as in the Russian airliner explosions, between domestic and international acts of terrorism.

"The problem is we've had meaningless statistics on terrorism for a very long time. But they've only dropped it now because it didn't produce the cut in terrorism they wanted," Cordesman said in an interview. "My hope is there'll be enough pressure on them to do something right rather than simply stopping doing something wrong."

Dennis Pluchinsky, a former State Department terrorism analyst, agreed that the numbers are far less useful than they might seem in assessing efforts to combat terrorism. "They talk about the war on terror and the increase of international incidents taking place, but they don't weigh the statistics -- 9/11 and someone throwing a Molotov cocktail is treated as the same thing," he said.

Many analysts agree there were problems with the method used to generate the government's numbers of "significant" international terrorist attacks. Among those identified by Brennan: how to identify what constitutes "significant" property damage, whether deadly acts by homegrown terrorist groups with links to international terrorism should be counted and how to decide what is a terrorist act in a war zone. Iraq has proved a particularly vexing test, and government counters, Brennan said, "found it virtually impossible to distinguish between insurgency and terrorism" there, settling instead for the imperfect solution of a report that includes attacks that killed or wounded non-Iraqi civilians in Iraq but excludes many other incidents in which terrorist-style tactics were employed. Overall, the NCTC identified nine times more terrorist attacks in Iraq in 2004 than in 2003.

"Inevitably there are some judgment calls that go into deciding what is a terrorist event," said Alan B. Krueger, a Princeton economics professor who helped reveal the flawed statistics used in last year's report. "But it is astonishing to me that three years into the 'war on terrorism' there is not more interest by the administration in keeping track of terrorist events."

Flawed or not, some politicians said the government should not abandon the one annual report card it had. "How can we hold ourselves accountable for achieving benchmarks of progress in this struggle," asked Sen. Russell Feingold (D-Wis.) in a letter to Rice, "if we have no clear idea of what exactly it is that would constitute success?"

© 2005 The Washington Post Company