Evacuate or Shelter in Place?

Most preparedness questions are about quantities. This one is about direction. When something goes wrong, you will either leave your home or stay inside it, and everything you have stored, packed, or planned serves one of those two answers. People assume the choice will be obvious in the moment. Often it is not, because the moment is loud, the information is partial, and the window is short. Ready.gov's framing is the useful one: work out both plans in advance, precisely because you will not have time to invent either one live. The decision itself, when it arrives, reduces to three questions.

Question one: is the hazard actually coming to you?

Start with geometry. Some hazards travel toward a specific place, and the only thing that matters is whether that place includes your address. A wildfire has a direction. A flood follows the drainage. A chemical release drifts downwind, not upwind. For hazards like these, the honest question is not "how bad is it" but "is its path and my location the same point on the map". If the answer is yes or even probably, distance is the fix, and distance means leaving early, while the roads are still roads. Other events have no path at all: a regional storm, a grid failure, a heat wave. They are simply everywhere, and driving does not take you out of them so much as it takes you away from your supplies.

Question two: is outside more dangerous than inside?

The second question compares the two environments directly. Evacuation always means hours of exposure: roads, weather, traffic, fuel stops, other people making the same decision at the same time. Sheltering means betting on your building and your supplies. In a fast-onset local hazard, inside loses that comparison badly, because the building itself is what the hazard is coming for. In a widespread event where the danger is the conditions outdoors, an ice storm, a hurricane already making landfall, a period of dangerous air, the building is the asset and the road is the risk. The general shape ready.gov describes holds up well: evacuation favors fast-onset local hazards, and shelter favors widespread events where outside is the dangerous place.

Question three: what are officials saying?

The third question outranks your read of the first two. Emergency managers can see the fire line, the river gauge, and the plume model; you cannot. When officials order an evacuation, the framework is done and the answer is go, promptly, because the order accounts for how long leaving takes when everyone leaves at once. When they say shelter in place, that is equally load-bearing: it usually means the roads are the hazard. The failure mode to avoid is treating official instructions as one input among several. They are the tiebreaker and often the whole decision. Your job before the emergency is making sure you will actually receive them, which is a big part of why an emergency communication plan exists.

Walkthrough: a wildfire on the ridge

Smoke is visible from the yard and the fire is a few miles out with wind pushing it your way. Question one: yes, it has a path, and you may be on it. Question two: outside is dangerous, but inside is worse, because a house does not negotiate with a fire front. Question three: an evacuation warning is out for the zone next to yours. Every branch of the framework points the same direction. You leave now, ahead of the order if you are ready, because the cost of leaving early and being wrong is a hotel night, and the cost of leaving late is the one this framework exists to prevent.

Walkthrough: an ice storm takes the grid down

Same framework, opposite answer. Question one: the storm has no path to dodge; it is sitting on the whole region. Question two: outside is the hazard, iced roads, downed lines, and every mile driven is exposure, while inside you have food, water, blankets, and walls. Question three: officials are telling people to stay off the roads. The framework says shelter, and the event becomes a supply problem rather than a movement problem: keeping warm, keeping fed, and managing the refrigerator on USDA's clock of about four hours for the fridge and about forty-eight for a full freezer. The power outage planning guide covers that whole sequence.

What each answer requires you to already have

The reason to run this framework on a calm afternoon is that each answer depends on different preparation. "Go" requires a packed bag with three days of essentials, copies of documents, medications, and a fuel habit that keeps the tank above half; the 72-hour kit math guide works through what three days actually weighs. "Stay" requires depth instead of portability: the FEMA baseline of one gallon of water per person per day, with a three-day floor and two weeks recommended for sheltering at home, plus calories to match. Most households prepare for only the answer they find easier to imagine.

One last note on the failure modes, because they are asymmetric. The evacuation mistake is almost always lateness: waiting for a certainty that never arrives, then joining the road at its most crowded hour. The sheltering mistake is almost always shallowness: staying, correctly, but with three days of supplies against a ten-day event. Neither is fixable in the moment. Both are fixed on the calm weekend when you pack the bag, fill the containers, and write down who calls whom. Prepare for both answers, decide with the three questions, and let officials break any tie.