Enough about lasers

Lasers are in the news again, as the US Navy has successfully tested a solid-state high energy laser fired from a warship. Overcoming violently bobbing seas, the laser managed to set its target's engine on fire from miles distant, a fact attested by a subsequent YouTube video:

Shortly following the test, commentators jumped to suggest an immediate application of the weapon: hunting pirates (see, for instance, US Navy's laser test could put heat on pirates). It is the second time in recent months that lasers have been proposed as anti-pirate weapons, the first being BAE's development of a laser distraction system in January.

Believe me, I get it: shooting pirates with lasers would be really cool (a view shared by the BBC News anchor who asked his guest, an expert on maritime security, if the pirates would be able to defend themselves from such a weapon by holding up a giant mirror). But the challenge faced by the international naval forces off Somalia is being within miles of the pirates when they attack, not what weapons to use once they have them in their sights. Put simply, you don't need space age weaponry to defeat poorly-trained brigands armed with a motley collection of aging assault rifles and the odd RPG launcher. I'll admit, it's an amusing image: that of a wretched, half-starved Somali pirate, a US warship bearing down on him, secretly relieved that his ordeal has come to an end and resignedly preparing to toss his gun overboard and surrender himself, only to look on in perplexed bewilderment as the outboard motor on his beat up fishing skiff suddenly erupts into spontaneous flame.

This is a one-sided arms race if there's ever been one.

The numbers lie

Charles Lane, the former editor of The New Republic who helped expose the Stephen Glass scandal, once told an interviewer that the "mystique acquired by words printed on paper" was what had permitted Glass's journalistic fabrications, though in hindsight utterly outlandish, to pass unquestioned for years. Glass was hubrisitic enough to invent fictitious people, buildings, even government agencies-- such flagrant falsehoods, in other words, that his downfall was preordained. Statistics are far more insidious offenders; they hide their lies in numbers, mask their troubled histories behind ciphers on a page.

An indelible professor of mine at McGill University, who taught a class titled The Underground Economy, once described being at a cocktail-and-canapé attended by sundry UN bigwigs in New York. Spotting the then-head of the UN Office on Drugs and Crime (UNODC), my professor approached and began a conversation. The UNODC had recently upped its estimate of the annual worth of the global illicit drug trade by some $50 billion, and my professor wished to know the basis for this lofty revision. Flustered and defensive, the UN chief stormed out of the room. He later returned in a calmer state, admitting to my prof that word had come down from the Secretary General's office to make the drug trade a greater priority over the coming year; cooking the stats seemed a good first step. This anecdote has stuck in my mind ever since, and remains one of the single most important lessons from my university years.

My eternal mistrust of statistics has proved especially appropriate over the course of my current assignment. Shortly after the Somali pirates hit the international news pages in late 2008, for instance, Kenyan foreign affairs minister Moses Wetangula claimed that the pirates had been paid $150 million over the previous year, a baseless figure that he (correctly) deemed was sufficiently large to earn him some ink in the international press (the actual number was around $30 million). A more recent dubious figure making the press rounds is $238 million, the amount allegedly earned by the pirates in 2010, according to a report by the think tank One Earth Future (OEF). Yet diligent scrutiny reveals this figure to be barely more credible than Wetangula's concoction. The estimate, I determined, had been calculated by multiplying a rough estimate of the average pirate ransom ($5.4 million)—a number derived midway through the year from a sample size of less than ten—by the total number of ships hijacked in 2010 (44). While sloppy in its overall methodology, the report also contained a more rudimentary flaw in logic: the assumption that every ship hijacked in 2010 was ransomed during the same year. In reality, the vast majority were not released until 2011; indeed, as of this writing many are still anchored within sight of the Somali shore.

Immediately suspicious of this report, I decided to attempt my own estimate. I set about researching each vessel hijacked since 2008 and the amount paid to release it, supplying my own guesswork where ransoms had not been publicly disclosed. The result? The cash paid to pirates in 2010 fell in the range of $65-$85 million, or about one third the OEF estimate. My (admittedly rough) initial calculations are available here.

Junk statistics are everywhere, but they tend to seek out certain victims, like Somali piracy, where facts are scarce and credible authorities few. It is environments like these in which junk stats thrive, breeding and propagating themselves like so many germs on a handkerchief. The media is only too willing to serve as a transmission vector, eagerly quoting whichever individual or organization conjures up the largest (and most sensational) figures; even The Economist, generally assiduous with its stats, quotes this mammoth estimate in its latest piece on piracy. In the age of Internet journalism, as stories are churned out increasingly rapidly and with commensurately less oversight, the danger posed by junk stats has become more immediate.

The printed word is too often treated as sacrosanct, imbued with an authority that overrules our own judgment. We must always remember that people, with their own flaws, biases, and agendas, write the words. And the numbers.