Twenty-one years ago I argued in the New York Times that the United States was not hardening its electrical grid against a foreseeable, catastrophic attack. The piece asked a simple question. Two decades later, is the grid safer? It is not.
The empirical record on the threat has only sharpened in the years since. Two congressional commissions have studied it (in 2008 and 2017). One executive order has directed federal action (in 2019). Three administrations have had the opportunity to harden the grid. None has done so at scale. The reason is that no one in government decided anything, not that anyone in government decided not to harden the grid. The cost of inaction has fallen on the public, and on a public that does not know enough to vote against it.
This piece is the reader’s guide to the documentary record. It is also an argument that the failure here is not an information failure. Two commissions did the analysis. The science is not contested by the people who should know. What is missing is decision and execution. The undecisionmakers have run the clock for two decades.
What an EMP actually is
The threat picture is three, not one threat.
A high-altitude electromagnetic pulse (HEMP) is a deliberate detonation of a nuclear device at altitude over a target country, generating an electromagnetic field that couples into the electrical grid and into electronic equipment. The signature physics of the threat were established by the Starfish Prime test in 1962, which detonated a 1.4-megaton device at 400 kilometers altitude over the Pacific and produced field effects that damaged streetlights in Hawaii nearly 1,500 kilometers away. The 1962 effects were not anticipated by the designers of the test. The lesson the test taught is the lesson the EMP Commission has been trying to convey to policymakers ever since: the field strengths produced by even a modest nuclear device at altitude are sufficient to damage continental-scale electrical infrastructure.
A geomagnetic disturbance (GMD) is a natural phenomenon, driven by solar coronal mass ejections that interact with the Earth’s magnetic field. The 1859 Carrington Event remains the canonical case. A modern Carrington-class event would damage the grid through different physics than HEMP (low-frequency geomagnetically induced currents in long transmission lines and transformers, rather than the broadband fast-rise pulse of HEMP) but with overlapping consequences. NOAA has documented multiple near-misses since 2003. The 2012 solar storm that crossed Earth’s orbit ahead of the planet’s transit through that point in space would have been a Carrington-class hit if its timing had been one week different.
A deliberate non-nuclear EMP is the third threat, generated by directed-energy systems or radio-frequency weapons. The threat is geographically narrower than HEMP and weaker in field strength, but it is also achievable without a nuclear program and is therefore the variant most accessible to non-state actors and to lower-tier state programs.
The three variants share a common consequence at scale. They damage transformers, control electronics, and the long-haul transmission lines that move power across the continental grid. The transformers and the control electronics are difficult to replace. The long-haul lines carry the field-induced currents that produce the damage. The grid was designed without these threats in mind. Its modern operation is more, not less, vulnerable to them than the grid of fifty years ago, because modern grid operation depends more heavily on solid-state electronics that are more sensitive to field effects than the electromechanical equipment they replaced.
The 2008 EMP Commission
The Commission to Assess the Threat to the United States from Electromagnetic Pulse Attack delivered its critical infrastructures report in 2008. The headline finding was that a HEMP event over the United States could produce continent-wide grid failure lasting months to years, with civilian casualty estimates that varied with assumptions but ran into the hundreds of thousands or higher. The transmission system, the report concluded, was not hardened against the threat, and the recovery time after a major event would substantially exceed the time the population could survive without electrical power for water, food refrigeration, climate control, and medical care.
The Commission’s recommendations were specific and actionable. Hardening of large transformers. Stockpiling of replacements. Hardening of grid control electronics against EMP. Coordinated federal-utility response planning. Funding for research into both the threat physics and the engineering countermeasures. The recommendations were costed. The Commission made the case that hardening costs were a small fraction of the costs of a successful attack.
The federal response to the 2008 report was to commission more studies. None of the recommendations was implemented at scale.
The 2017 EMP Commission
The Commission was reauthorized and produced an updated report in 2017. The 2017 findings sharpened the 2008 ones. The threat picture had not improved. The vulnerability had increased because grid modernization had introduced more solid-state electronics, more long-haul interconnection, and more dependence on digital control. The recommendations had been refined. The action items were essentially the same as 2008’s, with the addition of detail on protecting the SCADA systems that operate the grid in real time.
The 2017 report named the implementation problem explicitly. The federal government had not coordinated. Agency lanes had not been clarified. Funding had not been authorized at the scale the engineering required. The Commission’s recommendations to Congress in 2017 included specific instructions on which committees should act and what authority each needed.
Congress did not act on the recommendations.
Executive Order 13865
President Trump issued Executive Order 13865, “Coordinating National Resilience to Electromagnetic Pulses,” on March 26, 2019. The order was a substantial federal commitment on paper. It assigned coordination responsibility to the Director of National Intelligence and the Secretary of Homeland Security, directed agencies to identify their EMP-relevant equities, and required reports back to the National Security Council on a defined timeline.
The order’s substance was real. Its execution was incomplete. Some agencies produced the required reports. Some did not. The Director of National Intelligence’s coordinating role was undermined by the lack of dedicated funding for cross-agency hardening work. The Department of Energy stood up a working group on EMP resilience. The working group produced a strategy document. The strategy document was not implemented.
The Biden administration left EO 13865 nominally in force and did not rescind it. The implementation continued at low priority. The 2024-2025 transition did not produce a redirection of the EMP coordination effort. The 2026 status of the federal effort is roughly where it was in 2020. Strategy documents exist. Implementation has not.
The implementation gap
The pattern across two commissions and one executive order is consistent. The diagnosis is clear. The recommendations are specific. The execution is missing.
The reasons are structural. Grid hardening is expensive. The benefits are diffuse and unmeasured (a hardening that prevents an attack does not produce a visible result). The costs are concentrated on utilities and on federal agencies that do not have line-item authority to spend on EMP-specific projects. The political constituency that benefits from the spending is the future population that would not be killed by a successful attack. That constituency does not vote.
The undecisionmaker problem applies in its purest form here. No one in government has decided not to harden the grid. Many people in government have failed to decide to harden it. The functional difference between the two is zero. The grid is not hardened.
The international threat picture
Two states have demonstrated the technical capability to deliver a HEMP attack on the continental United States. Russia’s strategic nuclear forces include weapons appropriate for the mission, and Russian doctrine has historically included HEMP scenarios in war planning. China’s strategic forces have been growing rapidly, and Chinese military writing has discussed HEMP scenarios as well, although Chinese doctrine on the subject is less well documented in the open literature.
North Korea has tested nuclear devices. The yields are at the lower end of the range required for a continent-wide HEMP effect. The delivery question is more complicated for North Korea than for Russia or China. The DPRK has tested ballistic missiles capable of reaching the continental United States. Whether the DPRK has the integrated capability to detonate a nuclear device at altitude over CONUS is a question on which the open literature is divided. The Commission’s 2017 report treated the DPRK as the most plausible non-Russian-non-Chinese threat.
Iran is the fourth case. Iranian doctrine has historically included asymmetric warfare options that include EMP scenarios. Iran’s nuclear weapons program status is contested. The threat from Iran would be lower-yield and would likely produce regional rather than continental effects.
Non-state actors are not on this list because the technical barriers are higher than the rhetorical claims sometimes suggest. A non-state actor capable of delivering a nuclear device to altitude is rare. The non-nuclear directed-energy variant is more accessible but produces narrower effects.
The threat picture has not improved since 2008. It has gotten worse on the dimension that matters most. More states with more capability. The defensive posture has not improved at all.
The cultural reception
Public discussion of EMP has been polarized in a way that has worked against the implementation effort. On one side, prepper communities and a subset of conservative media have treated the threat with high attention and have produced framings that range from technically literate to alarmist. On the other side, mainstream policy discussion has often dismissed the threat as fringe, citing the prepper coverage as evidence that the topic is unserious.
This is the polarization trap. The threat is real and serious by the documentary record of two congressional commissions. The threat is also a topic on which a significant share of public discussion is unserious. Policymakers who want to be taken seriously by their professional peers have reasons not to engage with the topic, because doing so is associated with the unserious framings. The result is that the topic is largely abandoned by the people who should be working on it.
This piece does not engage in the dismissive register. The science is established. The recommendations are specific. The threat is real. The framing of the topic as fringe has produced exactly the implementation gap the EMP Commission has been documenting for fifteen years.
What should happen next
Congress should authorize a multi-year hardening program targeted at the long-haul transmission system, with specific cost-shared funding for utility-side hardening of large transformers, SCADA systems, and recovery infrastructure. The technical roadmap exists. The 2017 EMP Commission report is the canonical reference for what to do. Congress has not appropriated.
FERC should issue mandatory reliability standards for HEMP and GMD protection of the bulk power system, with NERC as the standards developer and clear deadlines for compliance. NERC has produced GMD-related standards. The HEMP-specific standards have lagged. FERC has the authority. It has not used it.
The Department of Defense should harden the military grid (the installations, the critical command and control, the Defense Industrial Base supply lines) on its own authority. DoD has the budget and the authority. It has done some of this. It has not done enough.
State public utility commissions should require state-jurisdictional utilities to file EMP and GMD resilience plans, with cost recovery mechanisms for the hardening investments. Several states have begun this. Most have not.
Utilities should stockpile spare large transformers (the long-lead items that are the binding constraint on recovery from a major event) under a coordinated industry-government program. The stockpiling proposal has been in the EMP Commission’s recommendations since 2008. It has not been implemented.
The executive branch, regardless of administration, should fully implement EO 13865, designate dedicated coordinating staff at the National Security Council, and produce annual public progress reports on the cross-agency hardening effort. The order remains nominally in effect. Its execution has been low-priority across two administrations.
None of these actions is exotic. The technical roadmap exists. The threat assessment exists. The legal authorities exist. The thing that does not exist is the political will to convert two decades of analysis into hardening that the engineering record has called for.
The reader’s guide
This Document Archive contains the canonical EMP record, with annotation. The four Tier 1 entries are linked below. Read them in this order.
- 2008 EMP Commission, Critical National Infrastructures Report (Annotated)
- 2017 EMP Commission, Recommendations to Congress (Annotated)
- Executive Order 13865, Coordinating National Resilience to Electromagnetic Pulses (Annotated)
- DOE Electromagnetic Pulse Resilience Strategy (Annotated)
Additional Tier 2 and Tier 3 archive entries will be published over the coming quarters. The Document Archive coverage of EMP is meant to be the canonical English-language reading list for the topic. The pillar above is the reader’s guide. The archive is the underlying record.
An earlier version of this argument appeared as “The Terrorist and the Grid” in the New York Times, 2005. The two-decade gap between that piece and this one is the gap this piece is about.