System reliability, disasters, and planning

Three different things have made me start thinking today about system reliability and planning and how that relates to the recent Katrina disaster. First, this morning on ABC’s Good Morning America I heard New Orleans Mayor Ray Nagin valiantly attempting to salvage his record and reputation which have received a certain amount of criticism in the aftermath of Hurricane Katrina. In response to Charlie Gibson’s question “Why didn’t you follow the published emergency plan for the city of New Orleans which called for the mayor being responsible for overseeing the evacuation of those who were unable to evacuate the city on their own?” Mayor Nagin’s answer was rather extraordinary. He said that they had been evacuated—to the emergency shelters like the SuperDome, the Conference Center, etc. I suppose this is what it means to live in a post-Clintonian world in which “it depends on what the meaning of is is”.

The second was a note that Austin Bay posted on his blog. The note was from William S. Gross, a consultant in emergency management issues with experience in domestic and foreign disaster relief. It doesn’t cover a lot of new ground but does reinforce the civics and logistics lessons we’ve been receiving and which the MSM, in particular, need to consider more seriously. I commend the article to your attention.

The third was a meeting I had this morning. In the meeting I was representing a client and the discussion was of reliability and security issues related to business relationships, disaster recovery, and business continuity planning (I’m seeing a lot of this in the wake of the ill-considered and cursed Sarbanes-Oxley). The folks I was meeting with were from a Fortune 50 company and my client was a vendor of theirs. They said they’d established target measurements for security and reliability for their business factor and that, in order to meet those targets, they expected all of their vendors to achieve those same measurements based on criteria they’d established e.g. let’s say they’d established a ten point scale with 1 the worst and 10 the best and they wanted to achieve an 8 in both areas. The method they were using to achieve this was to impel their vendors to achieve the same 8.

I pointed out that their approach was methodologically flawed: in general you can’t measure system reliability by simply averaging the reliability of the components—it’s a lot more complex than that. After struggling, vainly, for a minute or so to educate them on why this might be so and what realworld empirical studies had determined about actual measures of reliability based on component reliability, I realized that they simply didn’t have the attitude, background, or skills necessary for the task. They didn’t have the tools even to understand the issues.

After the meeting I continued to ponder this and how it related to the recent Katrina disaster. Most legislators are lawyers and lawyers by and large don’t have the tools for understanding the behavior of complex systems. These skills aren’t generally available in journalism majors, either. How in the world can they formulate a plan or evaluate a plan once it’s been formulated? I think the answer is that they can’t. But neither do their appear to be willing to just take the word of somebody who does have the tools, either. Even in government there is a role for fiduciary relationships.

All of this brings to my mind a series of questions. I don’t expect answers to any of them (or even that anyone will consider them seriously) but perhaps you might find them useful for thinking about the past disaster and even the inevitable future disasters.

1. If every plan by every level of government had been executed with 100% effectiveness, would it have saved any lives?

   Frankly, I doubt it. In particular I think that the plan of herding the poor, the sick, the incompetent, and the helpless into a sports stadium with career criminals, thugs, and nogoodniks was built to fail. As Chesterton put it, “Anything that’s not worth doing is not worth doing well”.

2. If the New Orleans and Mississippi government had done what they actually did and the federal government responded with 100% effectiveness, would it have saved any lives?

   It also might be interesting to identify what would be required for the federal government to achieve a high level of effectiveness in response and what the social and economic costs for that might be.

3. If the New Orleans city government did what they actually did and both the Mississippi and federal governments responded with 100% effectiveness, would it have saved any lives? Determine the costs and implications.
4. If the New Orleans city government and the federal government responded with 100% effectiveness and the Mississippi government did what it did, would it have saved any lives? Determine the costs and implications.
5. How would the effectiveness of the response of government at each level be quantified?
6. Who evaluates the emergency plans set up by local and state governments? Do they have the resources, skills, and attitude necessary for the task?