26.4.03

SEPARATING EQUILIBRIUM. Mark Kleiman raises a question that Bill Sjostrom and I once had a conversation about: did steerage fares trade at a discount that reflected a somewhat higher probability of drowning in a shipwreck account the first- and second-class passengers had first crack at the lifeboats? (Anybody remember the camp song with the stanza that ends "So they sent them down below/where they were the first to go/It was sad when the great ship went down?" I am collecting variants of the chorus. So far I have "husbands and wives, little children lost their lives," "uncles and aunts, little babies wet their pants," and "fishies and turtles, little ladies lost their girdles." Other contributions welcome.)

The cost-benefit and discount argument is aptly parsed by Agoraphilia. Alas, a direct test of the hypothesis is going to be difficult. Shortly after lifeboats-for-all became policy, World War I, the postwar immigration laws, and the Depression put an end to the huddled masses disembarked at Ellis Island, which makes a comparison of fares, or of space devoted to the various classes using generally accepted econometric techniques difficult at best (noted indirectly by Amptoons.)

The Amptoons post notes that there were other failings that night. Indeed, Titanic commentators have noted the mindset of those years, which is that the sea lanes were going to be so busy that any ship in trouble would have help within easy steaming distance, that radio would permit quick transmittal of distress calls, and that ships would be so constructed as to sink slowly, if at all. Under those circumstances, the ship's own boats would be supplemented by boats from neighboring ships, and those boats would have the opportunity to make multiple trips. Those premises were mugged by reality on that April night. However, those were only the final indignities. Naval architects knew relatively little about the handling characteristics of extremely large ships. Titanic's center screw, which provided water flow over the rudder, only operated in the forward direction. Her rudder was small. An A Scow has more rudder relative to hull than Titanic did. Titanic's officer of the watch, William Murdoch, ordered a hard left turn and reversed all engines, in violation of all rules of reasonable piloting.

There is another dimension to the economics of providing lifeboats for all: will the crew be able to board all the passengers? Lusitania, torpedoed and sank in eighteen minutes. Empress of Ireland, holed in a collision in the Gulf of St. Lawrence, sank in fourteen minutes, took on a list that precluded use of the port lifeboats. In both cases, the sea claimed many dead despite the provision of lifeboats for all.

The preceding two cases come from George Hilton's Eastland: Legacy of the Titanic. (One of the greatest maritime disasters in U.S. history occurred in downtown Chicago as Western Electric employees and friends were boarding enroute to an employee picnic. (visit unofficial Eastland disaster site.)) Professor Hilton writes his book in recognition of an important fallacy of inference: reevaluation of prior probability on the basis of a single observation (are our homeland security folks paying attention?) His research turns up some thought-provoking material. Eastland had a very low metacentric height (very messy: if the idea of increasing profit by cutting prices seems counterintuitive, try worsening a list by counterflooding, i.e. filling the ballast tank on the opposite side to a list. I'm still working on it. Lasers have a metacentric height that you can adjust by hiking hard. Flat is fast.) A boat (we're on the Great Lakes now, people) with a low metacentric height is tender. Now, take a tender boat, install air-conditioning topside, renovate some floors with concrete, and pile a bunch of rafts on the boat deck. Board a large load of passengers through the starboard gangway, and counterflood to keep her stable, if you can.

No comments: