A study of airline crash data over the period 1987-1996 shows that discriminating among airlines -- in favor of larger, established carriers over new start-ups -- to improve survival odds is a fruitless pursuit. Competing air services consistently entail virtually the same degree of mortality risk, according to Arnold Barnett of the Massachusetts Institute of Technology.

Here are some of his conclusions:

• Among established U.S. jet carriers, fatal crashes were distributed completely at random among the airlines.

• Post-deregulation, new airlines operated millions of flights in the U.S., with the ValuJet crash in 1996 being the only fatal accident among the start-ups -- not a statistical improbability over the decade period.

• Commuter propeller and prop-jet flights did exhibit higher mortality risks than jets -- but that may be caused by the two types of aircraft generally flying different types of routes and the gap may be narrowing.

• The data established that a passenger who took a First World domestic jet flight every day would on average go for 21,000 years before succumbing to a fatal crash -- 14,000 years for comparable international flights.

The corresponding statistics are 5,000 years for propeller/prop-jet U.S. commuter flights, 2,000 years for jet flights between the First and Third Worlds, and 1,500 years for Third World domestic jet flights.