Cognitive biases due to past nuclear accidents continue to shape energy policy long after the technologies themselves have changed.

The conversation on nuclear energy is back in international politics, but the global nuclear landscape reveals a striking paradox. More than 440 nuclear reactors operate in over 30 countries worldwide, providing roughly 9 percent of global electricity. Technologically advanced economies such as the United States, France, the United Kingdom, South Korea, and Japan are investing in new reactors, as are developing countries including Bangladesh, Egypt, and Uzbekistan. China alone has 37 reactors under construction, more than the rest of the world combined. The International Atomic Energy Agency is exploring advisory missions in countries like Rwanda and El Salvador. On the other side, a small group of wealthy countries has moved in the opposite direction. Germany completed its nuclear phasing-out and shut down its final reactor in April 2023. Belgium, Switzerland, Spain, and, most notably, Italy have chosen to abandon or restrict nuclear power, the latter holding two votes in the last twenty-four years. 

Why Global Nuclear Energy Adoption Follows a U-Shaped Development Curve

If you plot the world’s nuclear posture against any index of development, you do not get a straight line. You get a U-curve. Advanced economies at the top have either maintained nuclear power or returned to it in the name of energy security and industrial strategy. France is the clearest example: the 1974 Messmer Plan was fundamentally a project of strategic autonomy, and successive governments preserved the consensus. The United States and the United Kingdom took a longer route, losing the better part of a generation to political fallout before recalibrating around small modular reactors(SMRs) and the electricity demands of the artificial intelligence (AI) era. Japan has restarted reactors despite the Fukushima disaster, while South Korea has become one of the world’s leading exporters of civilian nuclear technology.

At the other end of the curve, developing economies increasingly see nuclear as the fastest route to reliable baseload power. Bangladesh, Egypt, the United Arab Emirates (UAE), and Turkey have embraced it as infrastructure rather than ideology. The World Bank ended its decades-long ban on financing nuclear projects in 2025, further reinforcing that trend. The countries caught in the middle, such as Germany, Italy, Belgium, Switzerland, and Austria, were wealthy enough to rely on alternatives and politically vulnerable to anti-nuclear mobilization. Strong environmental movements, coalition politics, and referendum mechanisms allowed singular traumatic events to shape policy for decades. Italy is perhaps the clearest example.

How Cognitive Bias and Nuclear Accidents Shaped Modern Nuclear Energy Policy

A cognitive bias sits at the heart of modern nuclear politics, focused on “the gambler’s fallacy,” the tendency to treat rare, unrepresentative events as reliable guides to the future. A belief that after five reds at the roulette wheel, black is somehow due. In the case of nuclear power, democratic societies have allowed exceptional accidents to define perceptions of an entire technology, often embedding those perceptions into law through referenda held at moments of peak public anxiety. Nuclear power was never a pure technical issue. Born in 1945 alongside the atomic bomb, the first generation of plants that followed President Dwight D. Eisenhower’s Atoms for Peace initially benefited from an era of trust in expert authority. That trust unraveled through three defining crises. 

The first was Three Mile Island in 1979, which caused no measurable public health harm but shattered confidence in the United States, as it happened two weeks after The China Syndrome landed in cinemas and caused a halt to new reactor construction for decades. The second was Chernobyl in 1986, a Soviet RBMK reactor without a containment structure, operated under conditions no Western regulator would have permitted, yet widely taken as a verdict on nuclear technology itself. The third was Fukushima in 2011, a 1970s-era reactor exposed to a documented tsunami risk that authorities failed to address, a failure of governance more than physics.

These events differed profoundly in technology, regulation, and circumstance. Yet they became fused in the public imagination into a single narrative of nuclear risk. The trauma was not evenly distributed. Their political impact was greatest in affluent democracies where public opinion could shape energy policy, but where energy security was not so urgent as to override it.

The political consequences followed the emotional calendar with precision. Italy held its first nuclear referendum in November 1987, 18 months after Chernobyl, and shut all four reactors by 1990. It held its second in June 2011, three months after Fukushima, with 94 percent voting against former Prime Minister Silvio Berlusconi‘s government plans for 10 new plants. Germany likewise accelerated its nuclear phase-out after Fukushima and completed it in 2023.

Voters were not being asked to assess nuclear safety based on more than fourteen thousand reactor-years of operating experience. They were being asked to decide at the peak of public anxiety, and the policies they produced have outlasted the emotions that created them.

A well-known cognitive bias, the availability heuristic, leads people to judge risk by the disasters they remember most vividly. Yet Chernobyl, Fukushima, and the reactors later proposed in Italy belonged to different technological eras. Chernobyl was a Soviet RBMK reactor without a containment structure; Fukushima was a 1970s-era design exposed to a known tsunami risk. The Generation III+ reactors considered in 2011 were designed with passive safety systems to prevent such failures, whereas today’s small modular reactors represent an entirely different technological generation altogether. To treat all of these as the same thing called “nuclear power” is akin to refusing to fly because of a crash half a century ago. The issue was not irrational concern about past accidents, but the failure to distinguish between past and present technologies, and a political system that turned that distinction into an enduring policy.

How Nuclear Power Shapes Industrial Competitiveness and Energy Security 

The competitiveness implications are increasingly difficult to ignore. As electricity becomes the critical input for artificial intelligence, advanced manufacturing, and industrial decarbonization, countries with abundant, reliable, low-cost power will enjoy a lasting advantage over those without it. France already produces electricity at a fraction of Germany’s carbon intensity and at significantly lower cost. The United States, the United Kingdom, France, South Korea, and Japan have all concluded that maintaining or expanding nuclear capacity is essential to competing in the industries of the future.

For countries caught in the middle, the question is one of strategic adaptation. Nations that secure abundant clean energy will attract investment, data centers, advanced industry, and innovation. Those that do not will increasingly import both energy and competitiveness. In that sense, the cost of a gambler’s fallacy embedded in public policy is measured not only in electricity bills but in industrial capacity, economic resilience, and ultimately sovereignty. It is one of the structural costs Europe pays when caution hardens into strategy.

About the Authors: Valbona Zeneli and Giovanni Renoldi

Valbona Zeneli is a nonresident senior fellow at the Atlantic Council, both at the Europe Center and at the Scowcroft Center for Strategy and Security’s Transatlantic Security Initiative. From 2011 to 2023, she served as a professor of national security studies and chair of strategic engagements at the US Department of Defense’s George C. Marshall European Center for Security Studies, where she continues her affiliation as a visiting professor. Zeneli is also a senior visiting fellow at the at the Florence School of Transnational Governance at the European University Institute in Florence, Italy. She serves as a member of the Advisory Board of Decode39. Follow her on X: @ValbonaZeneliTo.

Giovanni Renoldi is the principal at Future Developing, a venture capital firm based in Milan, Italy, and a former partner at Alamut Capital. He is a junior fellow at the Aspen Institute and a graduate of Bocconi University. He is currently studying applied economics at the London School of Economics.

The post The Global Return to Nuclear Power Is U-Shaped appeared first on The National Interest.