Nearly half the U.S. population now lives in areas exposed to building-damaging earthquake shaking, a sharp jump from prior federal estimates that raises hard questions about infrastructure readiness nationwide.
A new study built on the latest U.S. Geological Survey National Seismic Hazard Model found that roughly 159.2 million Americans reside in zones exposed to "very strong" earthquake shaking, the kind that can crack walls, buckle frames, and knock buildings off foundations. The previous version of the USGS model placed that number at 111.8 million, meaning the updated science added more than 47 million people to the highest-risk category in a single revision.
The numbers do not stop at population. Researchers calculated that 47.3 million homes and $25.6 trillion worth of buildings sit inside those very-strong-shaking zones. Widen the lens to include areas exposed to at least "strong" shaking, still enough to rattle heavy furniture and damage poorly built structures, and the figures balloon to roughly 234 million people, 73 million homes, and $38 trillion in building value.
For a country that tends to associate earthquake risk with the San Andreas Fault and little else, the study is a wake-up call. And the federal government's own hazard maps, updated for all 50 states simultaneously for the first time in the 2023 model, are the ones delivering it.
California remains one of the country's most notorious earthquake hotspots, but the study makes clear the threat extends well beyond the West Coast. The New Madrid Seismic Zone, a fault system running through the central United States near the Mississippi River, saw increased danger under the new model, with many of the nation's highest-hazard highway bridges concentrated in that region.
Cities that most Americans do not associate with earthquake risk turned up in the findings. New York, Atlanta, and St. Louis were all identified as unexpectedly vulnerable to severe disruption because of heavily traveled bridges that could fail or become impassable after seismic damage. Seattle, Portland, Salt Lake City, and Charleston also appeared among metro areas where bridges pose the greatest disruption risk.
San Francisco and Los Angeles, predictably, ranked high for bridges facing the greatest physical hazard. But the study's broader point is that earthquake preparedness can no longer be treated as a California-and-Alaska problem. A record-breaking earthquake in the Texas Panhandle in recent memory already demonstrated that seismic events can strike regions with almost no modern precedent for them.
Researchers examined more than 600,000 highway bridges across the country, infrastructure carrying a combined replacement value exceeding $1.5 trillion. Those bridges have the capacity to handle more than 4.5 billion individual vehicle trips every day. A single major earthquake in the wrong corridor could sever commuter routes, freight lines, and emergency-response access for millions of people at once.
The roughly 6,000 bridges ranked in the top one percent for hazard carried an estimated replacement value of $31 billion. Bridges ranked in the top one percent for overall risk, a measure that factors in traffic volume and disruption potential, not just shaking intensity, were worth approximately $74 billion to replace.
One comparison in the data underscores how the science has shifted over time. Using fixed modern population and infrastructure data, the 1996 USGS model would have flagged about 10,900 bridges as high-hazard. The 2023 model, applying better fault data and improved computer modeling, identified roughly 7,300 high-hazard bridges instead. That lower bridge count does not mean the danger shrank, it means the model now distributes risk differently, concentrating hazard more precisely while expanding the geographic footprint of who faces exposure.
Across much of the western United States, the predicted probability of damaging shaking actually fell compared with some earlier models. But in the central and eastern portions of the country, the revised earthquake records and new underground sediment data pushed the risk upward, placing communities that have done little seismic planning squarely in the crosshairs.
The 2023 National Seismic Hazard Model incorporated revised earthquake records, improved computer simulations of how seismic waves travel through rock and soil, and new data on fault lines and underground sediment layers. The study used a benchmark with a two percent chance of the predicted shaking being exceeded within 50 years, roughly the lifespan of a building or a major piece of infrastructure. That timeframe is not abstract. It is the window in which a home built today is expected to stand, a bridge poured this decade is expected to carry traffic, and a school opened next year is expected to hold students.
Researchers also flagged concerns beyond bridges. Water and gas pipelines, power stations, wastewater facilities, and communications towers face similar vulnerabilities, though the study addressed those qualitatively rather than with the granular bridge-by-bridge analysis.
Recent earthquakes around the world have shown what happens when infrastructure meets shaking it was not designed to withstand. A magnitude 4.5 earthquake near Naples injured dozens and collapsed buildings in a region where volcanic geology amplifies ground motion. In Japan, even a country renowned for seismic preparedness saw rising death tolls and fading rescue hopes after a powerful quake struck Kumamoto.
The lesson from abroad is consistent: the gap between known risk and actual preparedness is where people die. Venezuela's twin earthquakes killed hundreds in part because infrastructure and emergency systems were not built for the shaking that arrived.
The jump from 111.8 million to 159.2 million people in very-strong-shaking zones is not a rounding error. It represents a 42 percent increase in the exposed population between model versions. Tens of millions of Americans who were not counted in the prior risk assessment now sit inside zones where the federal government's own science says buildings can sustain serious damage.
Building codes, insurance requirements, and infrastructure spending decisions all flow from hazard models like this one. When the model expands the danger zone by tens of millions of people and trillions of dollars in property, the policy implications are enormous, and they land hardest on homeowners, commuters, and local taxpayers who have no say in whether their city updates its codes or reinforces its bridges.
The study raises questions it does not answer. Whether the researchers recommended specific policy changes or building-code updates is unclear from the available reporting. What the precise shaking thresholds mean in engineering terms, peak ground acceleration values, for instance, is also not spelled out. And whether the 159.2 million figure counts people by where they sleep, where they work, or some other measure remains unspecified.
Those gaps matter. A hazard model is only as useful as the decisions it drives. If the federal government updates its earthquake science but state and local governments do not update their building codes, zoning rules, and bridge-repair schedules to match, the model becomes an expensive academic exercise, and the 159 million Americans living in expanded danger zones are left holding the risk.
Washington has a long habit of producing world-class data and then failing to act on it. If 159 million Americans are in the shaking zone, somebody ought to be asking what the plan is, before the ground starts moving.
