How bizarre. The World Health Organization has declared swine flu a pandemic, signaling that governments worldwide should launch emergency response plans.

The mildest pandemics of the 20th century killed at least a million people worldwide, according to the WHO's data, while old-fashioned seasonal flu strikes every nation yearly and kills an estimated 250,000 to 500,000. As of last week, when the swine-flu pandemic was declared, the H1N1 virus had killed a total of 144 people. In Mexico, where the outbreak began and has been most severe, the cases peaked quickly, in just four weeks.

A pandemic declaration will be costly at a time when we can least afford it, and it could prompt severe restrictions on human activities (think China). Perhaps most important, the declaration could render the term "flu pandemic" essentially meaningless, risking lethal public complacency if a bona fide one hits.

So how can the WHO say swine flu qualifies as a pandemic? And why?

The WHO definition of "influenza pandemic" once required "several simultaneous epidemics worldwide, with enormous numbers of deaths and illness." But in 2005, the organization promulgated a definition that virtually ignores the number of cases and completely ignores deaths.

Now it requires "sustained chains of human-to-human transmission leading to community-wide outbreaks" in two parts of the world, with this addition: The cause must be an animal or human-animal flu virus; the latter is known as "genetic reassortment."

Under this definition, "community-wide outbreaks" of swine flu in two South American countries and somewhere in China could qualify as a pandemic - no deaths required. And a purely human flu that killed 20 million people would not qualify.

The obvious presumption is that viruses with animal genes pose a greater threat. But that's "a matter of faith more than science," says James Chin, an epidemiologist at the University of California, Berkeley, who was in charge of surveillance and control of communicable diseases at the WHO in the late 1980s.

Indeed, the available science indicates this presumption is false. The WHO first warned of an H5N1 avian-flu pandemic in 2004, projecting up to 150 million deaths. Yet a 2007 study found H5N1 - although first detected in 1959 - was many mutations away from becoming readily transmissible among humans.

We're also repeatedly warned that if H5N1 reassorted with human flu, it would produce a combination with the alleged severity of the bird virus and the infectiousness of the human one. Yet a 2006 Centers for Disease Control and Prevention study found the opposite: A genetically engineered reassortment given to ferrets - the best animal models for human flu - produced milder and less infectious flu than did pure H5N1. Reassortment didn't create a "super-flu," but rather a 98-pound weakling.

As to human-pig hybrid viruses, a 1976 swine-flu outbreak that took place in New Jersey in the dead of winter, when the flu is most contagious, infected just 230 soldiers, killing one, on a crowded Army base.

At some level, the WHO knows its definition is faulty. The online WHO Handbook for Journalists still states, "A pandemic virus can emerge" by adapting "during human infections." And the WHO has warned that one way avian flu could become pandemic is through a purely human mutation.

But it's also true that the worst influenza pandemic in history, the Spanish flu of 1918-19, did involve an animal-human hybrid virus, and that episode has become a popular obsession. Never mind that "each subsequent novel flu that sweeps through the world has been milder," Chin said. "The public-health community keeps waiting for the second coming" of Spanish flu, he said.

So, in 2005, the WHO apparently redefined "flu pandemic" to reflect popular angst as well as its own.

A less generous observation is that the WHO's "when, not if" avian-flu pandemic failed to materialize, and it has now put swine flu in its stead. In her pandemic announcement last week, WHO's director general, Margaret Chan, declared, "The world can now reap the benefits of investments over the last five years in pandemic preparedness."

Not really. Swine flu, being no more contagious and far milder than seasonal flu, calls for absolutely no actions that wouldn't apply to seasonal varieties.

"I think they're going to have to go back to the drawing board," Chin said of the WHO definition. For now, we'll pay in told and untold ways, because the WHO has cried "havoc" and let slip the dogs of pandemic.

Michael Fumento is director of the Independent Journalism Project and specializes in science and health issues. This appeared in the Los Angeles Times.