"We often hear about the danger of certain foreign airlines. Which are the safest and which should I avoid? What about budget carriers here in the USA"
Using data mostly from Airsafe.com, below appears a list of airlines that have gone fatality-free over the past 25 years. Certain small companies are omitted, though I've chosen to retain national flag carriers where applicable, regardless of size. All qualifying airlines have been in existence since at least 1980:
Several of the above have perfect records pre-dating 1980. Airsafe's own rankings reach back an additional ten years. I chose 1980 to best account for the changeover period from older, first-generation jets and propliners to what most would consider "modern" fleets.
You may or may not be surprised by some of the finishers -- second or third world operators not normally associated with safety. Whether the placement of an Air Zimbabwe, to pick one, attests to exemplary levels of oversight and professionalism is available for argument, and a mild caveat is due. Air Zimbabwe is a tiny outfit with, presently, four jets (two each 737s and 767s). Compare to American Airlines, with close to 800 ships and thousands of daily departures. Since 1980, American has outcrashed Air Zimbabwe 5-0 (including the September 11th aircraft), but plainly the comparison is unfair.
Nonetheless, any unblemished legacy lasting 25 years is impressive on its own accord, particularly when the setting is an underdeveloped nation with substandard facilities and infrastructure, debunking the customary wisdom that Western carriers present far and away better odds than everybody else's. And bear in mind that an exemption for even a single incident would expand the preceding list hugely, as would allowances for hijackings, hostile shoot-downs, or crashes involving fully or partly administered subsidiaries. A rundown of those counting one fatal mishap since 1980 takes in, just for starters, Royal Air Maroc, Kenya Airways and Mexicana.
Frankly, in certain regions I'd be more comfortable with a local carrier that knows its territory and the quirks of flying there. One example I love to cite is LAB —Lloyd Aereo Boliviano — the national airline of the poorest country in South America. Founded in 1925, LAB plies the treacherous peaks of the Andes in and out of La Paz, the planet's most highest altitude commercial airport. Since 1969, LAB has suffered two fatal crashes on scheduled passenger runs killing a total of 36 people. This is not a mainstay airline, but two crashes in 34 years amidst jagged mountains and the hazards of the high altiplano, is outstanding. Does that mean a company like LAB is safer, or even as safe, statistically, as our largest American or European airlines? No, not necessarily. The more accurate deduction is that both are trustworthy.
One of the most dubious reputations belong to Taiwan's China Airlines. According to Airsafe, China Airlines (not to be confused with the Beijing-based Air China) has recorded seven crashes since 1980. Some have called it "the most dangerous airline in the world." Seven wrecks in 25 years parses out to 0.28 annually, or roughly a crash every four years. That's comparatively awful, sure. A perusal of similarly sized companies reveals an average of one or two losses over the same quarter century span.
Revealing, yet therein is the rub. Seven versus two, in the context of tens of thousands of yearly departures, underscores how airline-to-airline comparisons are a game of statistical minutiae. China Airlines carries more than seven million passengers each year. Are your chances of injury or death greater aboard a China Airlines flight than aboard most others? Technically, yes. Practically speaking, no. Just as your chances of winning the lottery are only minutely improved by purchasing, say, three tickets instead of one.
Comparing airlines within the USA is no less an exercise in hairsplitting, whether American or AirTran; major or regional; jet or turboprop. Since the fall of 2001, our only black marks have been two commuter crashes in which 34 people were killed. Otherwise, Americans have taken to the skies over 1.5 billion times —more than two million daily— and lived to tell about it.
However, if you absolutely have to have statistics, the Web's most convenient source for data is probably Airsafe.com. Run by Dr. Todd Curtis, Airsafe is easy to navigate, accurate and professional —free of the sensationalist descriptions and gory disaster photos that mar the credibility of similar pages.
In some cases the site is too comprehensive for its own good, requiring some cautious dissection. Remember that raw accident numbers are deceptive without context. Airsafe's events totals offer equal weight to all incidents, regardless of the number of people killed, whether two or 200. They include hijackings, ground equipment accidents, and other ancillary casualties not always owed to a carrier's fault or negligence. Air China's tally, for example, accounts for the deaths of five people believed to have contracted SARS aboard one of its aircraft. And fatalities involving regional carriers are grouped with their mainline affiliates. The 2004 crash of a Corporate Airlines 19-seater is listed under American Airlines, with whom Corporate shares a name-only code-sharing arrangement. Also critically important, as we've already seen, is the size and scope—fleet size, volume of departures annually— of a particular airline.
Using a formula devised by MIT professor Arnold Barnett, Airsafe ciphers the variables of scale and proportion to come up with a "Rate" index. This method "sums the proportion of passengers killed and divides that by the underlying number of flights," providing a "passenger mortality risk per randomly chosen flight."
Although the Rate index is the site's most useful stat, here too the numbers aren't always what they seem. Air Zimbabwe's posted Rate of 11.54 is one of the highest (worst) of any airline. At the same time, as we've seen, Air Zimbabwe hasn't had an accident since one of its turboprops was shot down by rebels in 1979.
Personally I wouldn't bother vetting an airline's background unless headed for one of the more dodgy and unstable regions of the world. And even there, flying will normally present better odds than driving. For the survival minded flyer, there's not a lot of difference between who is "safest" and who is most "dangerous." The distinctions are, for most intents and purposes, academic. But if you're comforted by such reviews, and/or you're the type enamored of calculators and tinkering with decimal places, go ahead and bookmark Airsafe into your browser.
© Patrick Smith 2005. The above text is adapted from ASK THE PILOT. Portions have appeared previously on Salon.com