Walk through any serious preparedness forum long enough and you will find, nested somewhere between the freeze-dried meal stacks and the water filter debates, an almost religious conviction about EMP. Electromagnetic pulse. The grid goes down permanently. Every car built after roughly 1980 dies on the road. Banking systems, water pumps, hospital equipment — all of it, simultaneously, across a continent. Weeks become months. Society unravels.
It is a genuinely compelling scenario. It also does not hold up especially well when you press on it.
The EMP fear, in its modern form, traces back to a series of U.S. congressional commission reports published in the 2000s and 2010s. Those reports were real, the concern was genuine, and some of the underlying physics is legitimate. A high-altitude nuclear detonation does generate an electromagnetic pulse capable of inducing damaging voltage spikes across large electrical networks. That is not in dispute.
What gets quietly dropped in the forum retelling is everything else the same reports say: the extreme difficulty of achieving the precise altitude and yield required for continental-scale effects, the question of whether an adversary would "waste" a nuclear weapon on infrastructure denial versus a direct strike, and the deeply contested models for what hardened versus unhardened civilian electronics would actually survive. The 2008 EMP Commission report — the document most often cited in prepper literature — explicitly warned against overstating certainty. That caveat rarely makes the summary.
The more important gap is what we actually know from real-world events. Geomagnetic storms — solar events that produce some of the same grid-disruption physics as a high-altitude EMP — have hit modern infrastructure repeatedly. The 1989 Hydro-Québec blackout lasted about nine hours and affected roughly six million people. It was serious. It was not civilization-ending. The 2003 Halloween solar storms produced widespread satellite disruptions and some transformer damage across Scandinavia and parts of North America. Again: serious, bounded, recoverable.
The Carrington Event of 1859 is frequently cited as proof that a big enough solar storm would be catastrophic today. What that comparison misses is that the 1859 grid was telegraph lines, not a continental interconnected network with automated switching equipment and, increasingly, distributed generation. Modern grid operators have developed — imperfectly, but meaningfully — geomagnetic disturbance protocols since the 1989 Quebec event. NERC, the reliability organization that oversees North American grid standards, has published mandatory reliability standards specifically addressing geomagnetic disturbances. Progress is slow and uneven. It is not zero.
The reason EMP gets so much oxygen in preparedness culture has less to do with the data and more to do with the narrative structure. It is a single-cause, total-collapse scenario — one threat, one event, every problem at once. That is extremely legible. It turns preparedness into something almost game-like: if you can survive EMP, you can survive anything.
The problem is that legibility and probability are not the same thing. The scenarios that actually disrupt middle-class households are grinding, partial, and deeply boring by comparison. Extended regional power outages from weather. Supply chain slowdowns that empty specific product categories for weeks. Localized water pressure failures during a drought. Job loss arriving at the same time as an infrastructure stress. These are the patterns that actually show up in insurance claims, FEMA disaster declarations, and emergency room admissions. They are survivable with modest preparation. They do not require a Faraday cage.
There is also an opportunity cost argument that almost never gets made explicitly: every hour spent sourcing EMP-hardened electronics is an hour not spent building the liquidity buffer, the medication backup, or the three-week food supply that would actually help in the scenarios most likely to occur. Preparedness resources — time, money, storage space, mental energy — are finite. Allocating them toward tail risks with uncertain and contested probability profiles is a choice, and it crowds out something else.
None of this is to say that grid resilience is unimportant. It matters enormously — for households, for hospitals, for water systems. The case for grid hardening is strong on the merits, and it doesn't require the most dramatic possible scenario to justify it. A slower, steadier approach to household resilience — the kind that makes your family more comfortable during a week-long outage from an ice storm — will do more practical work than a bunker full of Faraday bags protecting equipment you've never tested.
The EMP scenario isn't worth zero of your attention. It's worth roughly the same attention you give to other low-probability, high-consequence risks: acknowledged, not dismissed, and not in charge of your planning calendar. The threats that are actually costing people sleep and money this year look a lot more like "the power was out for four days and we had no stored water" than anything a congressional commission modeled in 2008.
