After hearing that the City of New Orleans didn’t have an adequate evacuation plan, I found the plan on the City of New Orleans website and read it. My first thought was that it didn’t sound unreasonable, so I wondered, “What sorts of plans do cities usually have?” In addition to the New Orleans plan, I looked at the emergency preparedness plans for Minneapolis, and St. Paul. I was surprised at how similar they were, and what they said about our understanding of citizenship.

In all three cases primary emphasis is placed on individual planning—having sufficient supplies of water and canned goods—and in the New Orleans case, much discussion of a “support network,” a.k.a. the “buddy system.” To some extent this makes sense. Chances that authority can get to you in some sorts of disasters may be very limited; in a crisis situation there’s incredible uncertainty of all sorts. So there’s an emphasis on voluntarism and a subtext of personal responsibility and leadership: Take control of your situation to the best of your ability and help others.

One can look at the way things unfolded—and there are aspects of the evacuation that anyone can see went wrong—and feel very cynical about the buddy system. But all the plans reflected familiar features of American political culture—self-help, ingenuity, and voluntarism. Of course, they also assume that people will be connected. In an urban area, however, there’s a tension between a desire for anonymity, being able to come and go as one pleases—an American sense of liberty on one hand—and the necessity of having people know whether you’re safe. So people make choices within this range of privacy and connection, thinking about how comfortable it is to have the neighbors paying attention to their comings and goings. The emergency preparedness plans give suggest what kind of citizen planners are imagining.

The next question, though, is what we do once the “buddies” are gathered together. What if we don’t have a boat or a car? Of course, we saw instances of resourcefulness and ingenuity, but there is a serious question about how much of a plan can be based on improvisational self-organization. When I read about looting grocery stores I did wonder if one analyst’s “looting” might not be another analyst’s “self-reliance.” As it turned out, the actual New Orleans plan not only depended on people being able to get to some provisional site, but also to evacuate the city. The part of the plan dealing with evacuation, not surprisingly, explains traffic flows on interstates, but offers little specific information about what to do if you reach public facilities like the Superdome or Convention Center. It’s unclear how citizens were to get information or what the mayor and FEMA are supposed to do. So, we find a detailed list concerning personal plans and less focus on public roles.

The discussion of evacuation plans for people with disabilities also raised questions about the plan and the people who would enact it. There is mention of taking stock of one’s situation on an “enclosed form.” Were surveys sent out to households? Were they to be returned to the city? That would suggest a level of organization beyond the buddy system and a relationship between planners in city government and the people the plan is being drawn up for. Or, were the forms to be given to the “support system?” What were the expectations of persons with disabilities if they’d filled out the form? There’s also a whole section on what to do with your pet. The advice was send your pet to a relative or friend on higher ground; send your pet to a hotel, if it accepts pets; or leave your pet in the highest part of the home, unleashed, but contained. Basically, the message is: Forget your pet. Of course we shouldn’t put German shepherds in lifeboats instead of people, but what of the reality that people are attached to their pets? During the calm—the eye of the hurricane—people apparently went back to get their pets, and then were stranded and drowned. Dead pets are now part of the enormous pollution problem. These are things I had never thought about, and that is the whole point here. You need to plan for real people based on actual information about their lives.

Overall, though, the plans were based on very similar assumptions about self-reliance and cooperation, and if I’d read them in August, I might have thought them more or less reasonable. I would never have predicted the situation at the Dome or the Convention Center. So here is a real puzzle in institutional breakdown. And I’m not sure we are yet seeing an adequate inquiry or diagnosis of that failure.

Part of the inquiry has involved evaluating federalism. We’re hearing a lot of talk of local inadequacies and the need for greater national command and control. I think this is a mistake. First, I’d suggest that “federalism” does not simply mean that national government devolves authority to lower levels, but can always take it up again—administrative decentralization is not federalism. A federal system properly speaking is uncentralized, with multiple centers of power or concurrent regimes and jurisdictional authority; it’s polycentricity. Such a structure includes citizens individually as units of authority who are able to hold citizen-officials accountable. This matrix of various types of authority also includes voluntary associations, and governments themselves take an associational form. The system is one of limited, distributed, and most importantly, shared constitutional authority. We should think in terms of larger and smaller arenas of action, not greater or lesser powers. Where powers are shared, we may expect contestation—the point is to make the contestation productive.

So, we evaluate a federal union’s strength by the relationships among the parts, not by the concentration of power in any one part. In an analysis of institutional failures, we shouldn’t think any of these parts is “the” problem, or view failures as availing themselves of only one solution. I find it incredible when people, after laying out how the government in Washington didn’t pay attention to the predictions about the city’s vulnerability, and pointing out all of the errors of various federal agencies, suggest a more centralized government response. If there’s corruption in New Orleans politics, do I think there’s none in Washington? One newspaper column used Hurricane Katrina as an occasion to show the importance of government. Unfortunately the author focused primarily on the issue of security—restoring law and order—and evoked the imagery of Thomas Hobbes’s Leviathan in describing “the” government solution. One wonders: Which part of the Washington response merits absolute power?

Leviathan as metaphor (in contrast to the actual book) is a poor place to turn for solutions. This and other misleading depictions of events limit our understanding. Thinking about an actual assessment is more critical and more difficult than assuming that bigger is better—or, for that matter, that small is beautiful. But there’s not a lot of political gain in taking time to study a cascade of errors as may have occurred in New Orleans. It is not only that we accept simple answers. People understandably want houses up, they want money for reconstruction. There are immediate needs and there is also the critical need for diagnostic assessments.

In the end, self-government depends on people solving some shared problems directly by their own initiative and, when compulsory associations (governments, in contrast to voluntary associations) are involved, holding officials accountable. We can have all the institutional means imaginable to do this, but they must be used. I’m interested in what is going on in New Orleans now. What have we learned? Are we paying attention? American voluntarism has certainly been one dimension of response. Will that response extend to thinking critically about governance?