How do nuclear accidents affect nuclear power-plant construction? I built a giant database to find out.

Tuesday marks the 25^th anniversary of the explosion at Chernobyl. Meanwhile, the crisis at Japan's Fukushima Daiichi nuclear power plant has begun to stabilize, if only slightly. The horror stories from Chernobyl set the nuclear industry back countless years, with many countries canceling or stalling previous plans to build nuclear reactors. How will the disaster in Japan affect global reliance on nuclear energy?

For the past two years, I've been building a data set that can help answer this question. It contains the location and date of every nuclear power plant constructed in every country in the world between 1965 and 2000—based on records maintained by the International Atomic Energy Agency—and every significant nuclear accident during that time. I also collected country-level statistics on other factors that are thought to influence nuclear-power development: economic welfare, energy security, and energy production capacity, for example.

The nearly 75 nuclear accidents in the database include widely remembered disasters, such as Three Mile Island (TMI) in 1979 and Chernobyl in 1986, but also less-known incidents, such as the reactor meltdown in Buenos Aires, Argentina, in 1983 and an uncontrolled nuclear chain reaction in Tokaimura, Japan, in 1999 that killed two people.

Japanese authorities recently rated Fukushima at the highest possible severity level (Level 7), a designation shared only by Chernobyl. (TMI was classified as Level 5.) Given that the last accident of this magnitude crippled the nuclear industry, it may be tempting to conclude that the crisis in Japan will substantially curtail global nuclear power development. According to my database, however, it seems this judgment may be premature.

Nuclear-reactor construction is dominated by inertia. Harvesting nuclear energy is incredibly expensive at first, but much cheaper once the infrastructure is in place. Nevertheless, countries with a lot of money invested in nuclear energy have been surprisingly reluctant to give it up—even after major nuclear accidents. States that were not heavily invested, however, have often been quick to cancel their nuclear plans after accidents in other countries.

My analyses confirmed the conventional wisdom that TMI and Chernobyl stymied the global nuclear industry. A simple comparison between construction rates in the pre- and post-Chernobyl eras indicates that states were about 75 percent less likely to build reactors following the Soviet accident.

But Chernobyl and TMI aside, nuclear accidents generally have not deterred countries from building additional plants—even when the accidents happen within their borders. India, for example, began construction on four new reactors following the release of helium and heavy water at the Rajasthan Atomic Power Station in February 1995. *

Countries with nuclear power plants under construction prior to Three Mile Island continued to build reactors at a slower, though still steady pace—at least one new plant per country every seven years, on average, over the next two decades. In these countries—a group that includes Czechoslovakia, France, Pakistan, and South Korea—the accidents at Chernobyl and TMI raised concerns about the safety of nuclear power plants, but these fears were rarely sufficient to derail entrenched national interests.

Countries that hadn't already built a nuclear reactor, on the other hand, were about 35 percent more likely than all other countries to halt existing nuclear-construction plans (like those held by Egypt, Indonesia, and Vietnam). Twenty-seven countries built their first plant before that 1979 disaster. Only two countries—China and Romania—began construction on their first power plant during the next 20 years. The outliers had one thing in common: They were both nondemocratic countries, making them relatively insulated from the negative public reactions that inevitably ensue from nuclear accidents.

Proximity to these accidents was less important than one might expect—countries within 400 miles of the Soviet Union were not significantly more likely than others to curtail their nuclear plans after Chernobyl. There's even evidence that Chernobyl had a smaller effect on reactor construction in neighboring countries than non-neighboring ones.