When Disaster Strikes - Part 3: The Sacrifice Calculus

The Impossible Choice

In the five days after Hurricane Katrina struck New Orleans, Dr. Anna Pou faced decisions no physician should have to make. At Memorial Medical Center, cut off from evacuation, without power for air conditioning or most medical equipment, she and her colleagues worked to keep patients alive.

When evacuation finally became possible, it came by helicopter—and helicopters could only take the mobile. Patients in the intensive care unit, on ventilators, too sick to move, could not be evacuated. Temperatures inside the hospital had reached over 100°F. Some patients were clearly dying.

According to later investigations, Dr. Pou and two nurses administered lethal doses of morphine and midazolam to at least four patients. They were arrested, though a grand jury declined to indict.

The case generated enormous controversy. Was it mercy? Murder? A war crime under conditions no one had prepared for?

But the more important question is why Dr. Pou faced that choice at all. The sacrifice calculus at Memorial Medical Center was the final step in a long chain of sacrifices—decisions made years and decades earlier that determined who would be at Memorial when the storm came, who would be unable to evacuate, and who would be left to die when systems failed.

Implicit Triage: How Decisions Are Made Before Disaster

The explicit triage of crisis—the doctor deciding who gets the ventilator, the rescue boat choosing which roof to approach—is dramatic and visible. But it represents only the final stage of a much longer process of selection.

Most of the “sacrifice calculus” is made implicitly, embedded in infrastructure, policy, and social structure.

Infrastructure as Frozen Triage

Where you build the levee determines who drowns. Where you locate the hospital determines who can reach care. Where you put the fire station determines whose house can be saved.

These decisions are made through political processes that favor the organized, the wealthy, and the vocal. The result is infrastructure that protects some areas and sacrifices others.

Consider the New Orleans levee system before Katrina:

• Levees protecting the French Quarter and Central Business District were built to withstand Category 3 hurricanes
• Levees protecting the Lower Ninth Ward were built to lower standards
• Levees protecting wealthy Lakeview were reinforced more recently than those protecting poor New Orleans East

These were not explicit decisions to sacrifice poor Black neighborhoods. They were bureaucratic decisions about budget allocation, engineering standards, and project prioritization. But they produced a geography of vulnerability that was perfectly predictable.

When the levees failed, they failed where they were weakest—and they were weakest where political power was weakest.

Evacuation Policy as Selection

“Mandatory evacuation” sounds universal, but implementation is selective.

In New Orleans before Katrina, the official evacuation plan assumed that everyone who needed to leave would leave by car. There was no provision for the estimated 100,000 residents who didn’t have cars or couldn’t afford to evacuate.

This wasn’t oversight. City officials knew that a significant percentage of the population couldn’t self-evacuate. The 2004 Hurricane Ivan evacuation had revealed the problem clearly. But developing a real solution would have required resources and would have raised uncomfortable questions about why so many residents were too poor to flee.

Instead, the plan simply assumed the problem away. Those who couldn’t evacuate were told to “shelter in place”—a policy that, when the levees broke, became a death sentence for hundreds.

Zoning as Predestination

Where you’re allowed to build—and where you’re allowed to build cheaply—determines who lives in disaster-prone locations.

Flood zones are mapped. Seismic hazards are known. Fire-prone wildland-urban interfaces are identifiable. Yet development continues in all of these areas, often with minimal mitigation requirements.

Why? Because restricting development would reduce property values, anger landowners, and impose costs on local governments. The benefits of risk reduction are diffuse and long-term; the costs are concentrated and immediate.

The result is predictable: the people who can afford only the cheapest land live in the most dangerous places. When disaster comes, they die at higher rates. This is not chance—it is the accumulated consequence of policy choices.

Explicit Triage: The Moment of Crisis

When disaster overwhelms capacity, explicit triage becomes necessary. Resources are insufficient to save everyone. Choices must be made.

Medical Triage

Medical triage originated in Napoleonic battlefield medicine: wounded soldiers were sorted by their likelihood of survival with treatment. Those too sick to save and those who would survive without help received less attention than those for whom treatment would make the difference.

This framework—utilitarian, focused on maximizing lives saved—has been refined and formalized. Modern triage protocols assign categories: immediate treatment, delayed treatment, minimal treatment, expectant (dying).

But even “objective” triage protocols embed value judgments:

What counts as “saveable”? Elderly patients, patients with disabilities, and patients with chronic conditions may be classified as less likely to survive—and therefore receive less priority—even when treatment could help them.

What resources are assumed? Triage protocols assume certain resource levels. Change the resources, and the triage changes. During COVID-19, the question of how to allocate ventilators raised profound questions about whose lives counted more.

Who applies the protocols? Training, bias, and implicit assumptions shape how protocols are applied. Studies consistently show disparities in how triage decisions are made across racial and socioeconomic groups.

Rescue Prioritization

When the Coast Guard helicopter can only reach one rooftop at a time, which rooftop comes first?

Officially, rescue prioritization should follow need: the most endangered first. In practice, multiple factors shape rescue patterns:

Visibility: Rooftops with people actively waving, standing where they can be seen, get rescued first. Those trapped inside—elderly, disabled, those who retreated upstairs—are less visible.

Accessibility: Areas that are easier to reach get reached first. Remote areas, areas with overhead obstructions, areas with hostile conditions wait longer.

Communication: Those who can call for help, who have working phones, who can reach official channels, get prioritized. Those without communication are invisible to the system.

Social capital: In extreme cases, personal connections matter. The well-connected can call officials, can get prioritized, can jump queues. The unknown wait.

After Katrina, analysis of rescue patterns showed that higher-income areas were reached faster than lower-income areas, controlling for other factors. The official system was not designed to prioritize wealth—but the implementation advantages of wealth (better communication, more visible housing, more connections) shaped outcomes.

The Age-Old Sacrifice Pattern

Across cultures and centuries, certain groups consistently bear the highest disaster mortality: the elderly, the disabled, the poor, and the politically marginalized.

The Elderly

In almost every disaster, the elderly die at disproportionate rates. They are less mobile, more likely to have chronic conditions, more likely to live alone, more dependent on systems that fail in crisis.

During the 2003 European heat wave, over 70,000 people died across the continent—predominantly elderly people living alone without air conditioning. French mortality was particularly severe, with over 14,000 deaths, largely among the elderly.

The French response included recriminations: families were blamed for abandoning elderly relatives to take August vacations. But the deeper issue was structural: an aging population, reduced social connection, inadequate heat emergency planning, and a healthcare system not designed for heat crisis.

The elderly died not because families were callous, but because social structures made the elderly invisible until they started dying.

The Disabled

Disability dramatically increases disaster mortality. Evacuation systems assume mobility. Shelter systems assume self-care capacity. Communication systems assume access to standard channels.

When New Orleans was evacuated after Katrina, people in wheelchairs faced extraordinary barriers. Buses were not accessible. Shelters were not equipped for people with disabilities. Many people with disabilities simply couldn’t evacuate and didn’t.

A 2005 study found that people with disabilities were two to four times more likely to die in disasters than the general population. This is not about the disabilities themselves—it’s about systems designed without consideration for disability.

The Poor

Poverty is the most consistent predictor of disaster mortality. The poor live in more vulnerable housing, in more hazardous locations, with less access to information and resources for evacuation.

But poverty also interacts with every other vulnerability. Poor elderly people die at higher rates than wealthy elderly people. Poor disabled people are more vulnerable than wealthy disabled people. Poverty compounds every other risk factor.

This is not inevitable. It reflects policy choices: the decision not to provide evacuation assistance, not to build affordable housing in safe areas, not to invest in infrastructure for low-income neighborhoods.

The Politically Marginalized

Groups without political voice—immigrants, ethnic minorities, the unhoused, the incarcerated—consistently fare worse in disasters.

After the 1906 San Francisco earthquake, Chinatown was devastated, and city officials attempted to use the disaster to relocate Chinese residents permanently out of valuable downtown real estate. Only intense political resistance prevented this “reform.”

During Hurricane Maria in Puerto Rico, the death toll was initially reported as 64. Later studies estimated over 3,000 excess deaths. The low initial count reflected Puerto Rico’s political status—a territory whose residents cannot vote for president, whose concerns receive less federal attention.

The Moral Architecture of Sacrifice

Societies tell themselves stories about disaster sacrifice—stories that make the pattern acceptable, even invisible.

The Naturalness Narrative

“It was a natural disaster. Nature doesn’t discriminate.”

This narrative obscures the political choices that determined who was vulnerable. The earthquake doesn’t discriminate, but building codes do. The hurricane doesn’t discriminate, but levee investment does. The famine doesn’t discriminate, but food distribution systems do.

By framing disasters as natural, we convert political choices into natural fate—and make the sacrifice pattern seem inevitable rather than chosen.

The Personal Responsibility Narrative

“They chose to stay. They chose to live there. They chose not to prepare.”

This narrative converts structural constraint into individual choice. Did the Lower Ninth Ward resident “choose” to live in a flood zone, or was that the only place she could afford housing? Did the nursing home resident “choose” to stay, or was there no evacuation plan for people who couldn’t walk?

Personal responsibility narratives make the sacrifice of the vulnerable seem like the consequence of their own decisions—absolving the systems that constrained those decisions.

The Tragedy Narrative

“It was a tragedy. No one wanted this to happen.”

This narrative acknowledges harm while diffusing responsibility. Tragedies just happen—they have no authors, no decision-makers, no alternatives.

But the sacrifice calculus has authors. It has decision-makers. And there were alternatives. The tragedy narrative makes those alternatives invisible.

Can We Choose Differently?

The sacrifice calculus is not fixed by nature. It is the product of choices—choices that could be made differently.

Making Implicit Triage Explicit

The first step is acknowledging that triage happens before disaster, embedded in infrastructure and policy. Making these choices explicit—discussing openly who the levee protects and who it doesn’t—allows democratic deliberation about values that are currently hidden.

Some communities are beginning this process. Climate adaptation planning increasingly involves explicit discussion of which areas will be protected and which will be abandoned. Managed retreat programs acknowledge that some places cannot be made safe at any reasonable cost.

This is difficult. No politician wants to tell constituents they’re being sacrificed. But the alternative—implicit sacrifice without acknowledgment—is worse for those being sacrificed and less accountable for everyone.

Redesigning for Resilience

Infrastructure can be redesigned to reduce sacrifice. Universal evacuation planning—not just for those with cars—can reduce who gets left behind. Hospital design that considers power failure can reduce who dies when systems fail. Building codes that don’t discriminate by neighborhood can reduce who loses their home.

This requires investment and political will. The costs are real. But so are the lives saved.

Changing Who Decides

The sacrifice calculus is shaped by who makes decisions. When the vulnerable are excluded from planning, their needs are predictably ignored.

Disability rights activists have transformed emergency planning in many jurisdictions simply by being at the table. Community organizations in flood-prone areas have demanded and sometimes won infrastructure investment.

Including the potentially sacrificed in decision-making doesn’t guarantee their protection—but excluding them guarantees their sacrifice.

The Calculus We Accept

Every society makes sacrifice decisions. The question is not whether to make them, but how—and who bears the cost.

Currently, the sacrifice calculus follows predictable patterns: the elderly, the disabled, the poor, and the politically marginalized die at higher rates. This is not fate. It is policy.

We could choose differently. We could invest in universal evacuation. We could build infrastructure that protects everyone. We could design systems that don’t require lethal triage decisions.

We don’t. That choice—the choice not to change the calculus—is itself a sacrifice decision. It accepts that some lives matter less than others.

When Dr. Pou administered those injections at Memorial Medical Center, she was making the final decision in a long chain of sacrifices. But she was not the only one making choices. All of us, through the systems we accept and the policies we tolerate, are complicit in the calculus.

The disaster forces the choice into the open. But the choice was made long before the disaster came.