I live in Florida so I've been tracking Katrina and have watched New Orleans dying well after the storm struck and then it hit me...
New Orleans depended upon a brittle infrastructure.
I once wrote on the subject of CyberDiversity and ranted about competition being necessary in order to ensure flexibility-- though this is lost if a monopoly or cartel forms. (A gross asymmetry like Apple vs Microsoft qualifies as a monopoly in spirit if not the letter.)
Inefficiencies conferred by having multiple competing organizations provides far more resiliency when conditions are, to put it mildly, "sub-optimal", then any single "efficient" organization.
We saw that in Louisiana, didn't we?
Admittedly the principles aren't immediately obvious to the eye. No, first we had to go down to the cellar. With a torch...
Oh, sorry, got lost a bit, there.
When we try to maximize efficiency (by not spending a penny more than absolutely necessary) any organization will likely embrace a "single point of failure" in order to minimize the expense of getting it right. When this is done they will not have any more capacity than that absolutely needed for the "normal" course of business.
So in a city near the sea sitting below sea level it made efficiency-minded economic sense to not only size the pumping resources for normal levels of rainfall (and tolerate some flooding when it rains harder than usual) but to also use as few canals as possible to move the water elsewhere.
OK, so the pumps would have caught up, eventually, though likely not before flooding did a fair amount of damage.
Of course, there was very little hardening done to these systems, despite their criticality.
But, then, hardening costs money, doesn't it? Why invest in lining the canal with enough concrete to minimize the risk of breakage?
(As an aside, if, instead, an airliner as a cruise missile had been used instead... or a set of truck bombs... Water is New Orleans' Achilles' Heel. Of course now we have a "movie plot" to defend against, won't we?)
(I also believe that other localities have weak links in their infrastructure that, if damaged, will reduce (if not remove) livability. The problem is that we live with these for so long that we forget that it's a threat, all due to misplaced complacency. "The squeaky hinge gets the oil", for instance, rendering us reactive rather than proactive. We end up not managing risks in a reasonable fashion and triaging on the wrong set of criteria.)
For New Orleans, though, the break in the levee of the canal carrying much of the pumping stations' output to Lake Ponchartrain was the actual death blow to the whole city as a city: A single point of failure that rendered the pumping capacity irrelevant.
In this case we can see that efficiency encourages embracing brittle systems. Expending extra monies to have "extra capacity" beyond what most experts thought was needed would be considered wasteful.
I'm not sure any of the former residents of New Orleans would have considered extra pumping stations, extra canals *and* hardening of the levees to be a waste of public monies.
But, then, brittle systems are usually exposed only in hindsight. Under normal circumstances those whose job it is to manage risk are balancing it against a "bottom line"... i.e. "money" and "economic efficiency".
Bruce Schneier (see link ) has commented on security measures, both virtual and real. He terms some systems, especially those we've come to rely on in the airports, as "brittle". One hopes that security systems are more resilient in failure. My understanding is that resiliency requires more "depth"... which is why the military use the term "defense in depth".
As Bruce Scheier remarks, When a brittle system fails, it's a comprehensive (if not catastrophic) failure. It's all-or-nothing, a very binary set of outcomes. Resiliency requires some thought to fallback and elasticity, allowing a rebound. It can be argued that Flight 93 was an example of resiliency in action.
Resiliency usually implies enough spare capacity (that "defense in depth" thing) which seems anathema to those counting the beans. Spare capacity implies a level of redundancy that isn't "economical".
If commerce is a form of conflict, why aren't there any reserves? (Churchill, on asking the French high command in Paris about the "mobile reserves" during 1940, was told "Aucune"... as in there were none. This is probably one of the best examples of why we need spare capacity; they didn't have it and France was overrun.)
So it's not economically efficient to have spare capacity. Redundancy is a bad word when dealing with efficiency, at least when measured against "productivity per dollar".
Well, at least this quarter.
But efficiency has many adherents, especially in publicly traded corporations. Look at the efficiencies in having several refineries on (or near) the shoreline-- yes, they are convenient for shipping and, on the gulf coast of Louisiana, out of sight of the folks who don't want them in their backyards, but economic efficiency also encourages making them as large as possible. Large facilities tend to trade off scalability. When a disaster (or other damaging incident) occurs the resources needed to restore operation are greater, too, and, I suspect, more time is needed to get it right.
Again, a brittle failure. Too many eggs in too small a target... uhhh... basket.
So after Katrina we don't need to care about the price of crude oil since we can't make enough refined products (gasoline, kerosine, etc) to keep up with demand.
Well, at least until the refineries can be put back in operation.
Supply versus demand. With an administration that thinks that rising prices will impose de facto rather than de jure rationing.
Sure, it'd work if there were alternatives-- but, for the majority of the USA, the alternatives to private automobiles are inadequate at best.
Of course these higher energy prices will have other effects, as we've seen in the oil shocks of the 1970s. The effects of higher fuel prices will be ripple through the rest of the economy as these cost get factored into foodstuffs and the other products we need for daily life.
This may be temporary, of course, but such would assume that the greed of corporations controlling the flow of energy is subordinate to a sense of ethics.
Changing the subject again, before I go off on a tangent about greed and politics, which go together better than sex and violence, let's now look at some of the other impacts of Katrina.
There is now a non-trivial dislocation in the US transportation network. This may also be fairly temporary, but I suspect there'll be an "echo" effect up and down the supply chains that used to pass through the Gulf of Mexico.
New Orleans as a shipping port is, for the time being, not going to be very usable, or able to handle the volume it has in the past. Sure, I don't believe the port itself flooded, but I suspect the routes through the Mississippi Delta may need to be re-surveyed. Add to that the labor to operate the port won't be as conveniently located, either.
Rail operations in the Southern US are disrupted, albeit to a minor degree, so cargo-carrying trains have to be re-routed around this area and the switch-yards closest to New Orleans won't be usable until conditions improve.
Railroads are efficient carriers but they are also subject to brittleness since the roadbeds, being able to carry so much weight, won't have as much redundancy as needed for maximum connectivity. Railroads are not as resilient as the InterNet.
And that's part of the point: redundancy implies resiliency though it doesn't guarantee it. As long as the redundancy is spread around, of course. The DC-10's own redundant hydraulic systems all routed through the same portion of the wing structures, so a single event, like an engine ripping itself out of the wing, can cause a a single point of failure to take out all of these redundant systems.
It is perhaps a good thing that aeronautical regulations require redundancy that, if left to themselves, aeronautical manufacturers would be pressured by their bean-counters (who claim to represent stockholders) to leave out.
The effects of these transportation disruptions are not that great when viewed directly, of course; anything can be worked around if you provide enough time.
The less obvious impact will be seen elsewhere, in the basic fabric of business. It won't be a great disruption and is unlikely to be no more than a minor rumble as things "settle back down" to a "steady state", but...
There is another fad encouraged by the pursuit of maximizing economic efficiency: JIT.
"Just In Time" inventory. Inventory that is not in a store-room but "in transit". Nothing in the way of buffering.
Pipelined operations. With very little buffering.
Disrupt the pipeline-- at the source or even mid-way in the flow-- and the delivery rate at the consuming end will drop.
Heck, the delivery rate will bounce all over the place, though likely averaging out to a rate that is "planned".
In the process of Recovery from Katrina, between the energy costs to operate these pipelines (trains, trucks and planes) and the time disruptions at delivery, we have a lot of businesses in the southern tier that likely haven't "run dry" yet. When they do it'll be... upsetting.
JIT, in my belief, is yet another brittle system. It's wonderfully brilliant. It's wonderfully efficient. There's no waste of "products" sitting on a shelf at either end of the supply chain.
In a perfect world it's perfect.
Sadly, we do not live in a perfect world.
Up until now the impacts on JIT have usually been labor oriented; strikes and the like have provided all of the real impact, most blameable on human agencies.
I'm believing now that "efficiency" encourages brittleness.
Brittle thinking, brittle systems, brittle people.
With the layoffs various companies have suffered there is so little "spare capacity" (though I have noticed that non-productive bureaucrats and bean-counters seem never to be affected by layoff plans) that just operating around people who want to go on vacation imposes its own set of dislocations.
It's strange how I've heard layoffs in the UK as being "declared redundant". For the sake of economic efficiency and maximum profitability, all spare capacity has to go.
Why can't they just cook the books?
Oh, yeah, CEOs and CFOs don't like the idea of being declared as criminals.
If your work has you declare a "back up" person who can theoretically cover your work, either of you (usually the higher paid) can be declared "redundant".
Right-sizing is about eliminating redundancy in a system, which reduces any "defense in depth" by removing resiliency. Over time, such systems get more and more fragile as larger and larger workloads get placed on people since "they're carrying it all right for the last two weeks".
Of course there are limits to how productive a human being can be; we're not robots and we don't function well when we can't get enough rest.
No one wants to pay for spare capacity.
Somehow, I'm starting to think this will be the epitaph of the human race as we try to embrace more and more brittle mindsets in a desperate urge to reach new levels of economic efficiency and productivity...
... and all of this is in pursuit of profits.