If the American nuclear industry and its regulators ever had a robust safety culture, it began in the protest movements that opposed the construction of nuclear plants like Diablo Canyon in California and Seabrook in New Hampshire. To credibly oppose these plants, grassroots activists had to become technically proficient in diagnosing each plant’s specific faults, as well as well versed in regulatory law and policy. Their newfound expertise challenged the political bureaucrats and nuclear industry technocrats.

Then came the turning point—the 1979 accident at the Three Mile Island (TMI) nuclear power plant in Pennsylvania, which changed everything by opening up the nuclear industry to broad government and public scrutiny. After TMI, some of those activist-researchers would even end up in positions as government regulators.

Strangely enough, the story of Three Mile Island sort of begins with a Hollywood movie—The China Syndrome, which was released in March 1979. The nuclear industry dismissed its plot as far-fetched: a near meltdown in California and the subsequent exposure of plant officials’ lies by two hip journalists, played by Jane Fonda and Michael Douglas.

At the time, nuclear technocrats said meltdowns were almost impossible, as rare as “black swans.” But twelve days after the film opened, a reactor at TMI experienced a partial meltdown. Suddenly life was imitating art, imitating life.

The TMI accident began at 4 AM on Wednesday, March 28, when an open valve got stuck and started to drain reactor No. 2 of its coolant. The little indicator light showing when the valve was open was malfunctioning, and so engineers thought the valve was closed when it was actually open and draining coolant. Soon, pumps pushing extra water into the reactor ran dry—the water was essentially running out the other side—and started to shake and pound violently. Plant operators, fearing an explosion, shut them off, which only made the reactor drain faster. Pressure began building, the reactor got hotter and hotter, and the fuel rods began to melt.

But TMI’s engineers couldn’t see any of this, and they couldn’t figure out what was happening—all because the indicator light falsely told them the valve was closed. As the plant’s complex control panels went wild with blinking lights and loud alarms, panic gripped the crew. By then, they were beyond following procedures. A guessing game started, and human errors started to compound the technical problems.

Finally, one technician, in exhausted desperation, decided to turn on the pumps again. They refilled with water and stopped shaking. Fresh coolant hit the fuel rods, and pressure in the reactor began to go down.

Mike Gray, who wrote the screenplay for The China Syndrome, had been sent to cover the disaster for Rolling Stone. Just like the journalists in his film, he staked out a tavern closest to the plant. Exhausted and panicked workers coming off their shift, lubricated with beers and shots, were soon explaining the whole tale to him.

“If that guy hadn’t thrown that switch and gotten the pump back on,” said Gray, “Pennsylvania would probably have a new capital. They would have lost Harrisburg.”

In the end, dumb luck saved the day. The plant came within thirty minutes of a full meltdown. And even though that disaster was averted, the reactor vessel was destroyed and radiation was released into the atmosphere.

The accident stunned and terrified citizens across the country, including in Washington, DC. People wanted answers; ugly details began to emerge. For example, reactor No. 2 had been rushed into service before it was ready to operate. It opened on the last day of 1978. Had the utility company waited one day more, it would have lost a $40 million federal tax break.

“In terms of nuclear regulation, there is very much a before-TMI and then there’s after-TMI,” says Peter Bradford, a Carter appointee to the Nuclear Regulatory Commission (NRC).

In the wake of TMI, both the Kemeny Commission, established by President Carter, and the Rogovin Commission, called by the NRC, drew up long lists of recommendations. Leadership at the NRC was changed, and it incorporated “lessons learned” into a “TMI Action Plan” that spelled out both short- and long-range objectives to improve nuclear power safety. In Congress, several representatives, Arizona’s Mo Udall and Massachusetts’s Ed Markey among them, unleashed their scientific advisers, who became obsessed with obscure but crucial issues like x-rays of shoddy welding. Activists called for the closure of new plants. A top staff member at the NRC, deemed too close to industry, almost went to jail.

“We had never even heard the words ‘crisis management,’” one chastened NRC division manager told the Washington Post in 1979, “because nobody ever really thought there would be a crisis.”

President Carter signed into law stricter federal safety standards that included elements like mandatory emergency evacuation plans and harsher penalties for violations of federal safety standards by utilities. The NRC hired 146 inspectors and set up a stricter training program for them. As the commission’s budget significantly increased between 1979 and 1981, it hired even more inspectors.

To head off more onerous government action, the nuclear industry became better organized, transparent and professional. “It would be dishonest to say there have been no changes,” admitted Richard Pollock, then head of Ralph Nader’s Critical Mass Energy Project.

The new regime seemed to make nuclear plants safer, but it cost the utilities money—lots of it. A 1995 study in the Journal of Applied Econometrics found that nuclear power plants became “substantially less profitable” after the post-TMI safety regime kicked in. In fact, “over 90% of the expected discounted profits [profits to be expected in the future] from continued operation of existing nuclear power plants have been eliminated in the post-TMI period.” The direct cost of safety went up, as did the rate and amount of NRC-level fines.

A few half-built plants, now forced to meet stricter safety standards, went broke. Building new ones became more expensive, and plans for 100 other plants were canceled because of costs. In 1983 Washington Public Power Supply System abandoned three nuke plants in construction because the projects were plagued by massive cost overruns. One infamous section of piping had to be reinstalled seventeen times. The project’s cost ballooned to $24 billion and still wasn’t done when the utility defaulted on $2.25 billion worth of bonds. Episodes like that prompted Forbes, in a 1985 issue, to call the nuke industry “the largest managerial disaster in history.”

The desperation to get out from under debts incurred by building nuclear plants was a large part of what drove the push for utility deregulation in the early 1990s. A pro-industry backlash was building. Salvation came in the form of the Gingrich revolution of 1994 and the Republican domination of the 104th Congress, which moved to cut the NRC’s staff by almost half and to restrict its inspection and enforcement activities.

Republican Pete Domenici of New Mexico waged a jihad against the NRC and the parts of the EPA that dealt with nuclear power. He publicly grilled the head of the NRC, Shirley Jackson, and threatened to gut her commission if she didn’t lighten up on industry. She promptly did. By the late 1990s the NRC’s fines and infractions against industry had markedly dropped off.

“They stripped away much of the institutional knowledge that had been integral to the safety culture,” said Robert Alvarez, a former senior adviser in the Energy Department. Indeed, a major study by three political scientists found that the Clinton-era NRC allowed their enforcement regime—as measured by the number of violations filed and fines levied—to disintegrate and softened considerably.

“There hasn’t really been serious oversight for twenty years,” says Henry Myers, who was science adviser to Congressman Mo Udall.

In recent years we’ve seen a renewed campaign against regulation by the nuclear energy industry. And now funding for nuclear safety regulation and enforcement is again being attacked by the Republican-dominated House. All this has a directly corrosive effect on the safety culture. A recent report by the Union of Concerned Scientists (UCS) has found that federal regulators audited only about 5 percent of the activities at plants each year; it described the agency’s “spotlight” as “more like a strobe-light, providing brief, narrow glimpses into plant conditions.”

Written by UCS nuclear engineer David Lochbaum, the report examines fourteen “near-misses” at US nuclear plants in 2010 and evaluates the NRC response in each case, finding “a variety of shortcomings, such as inadequate training, faulty maintenance, poor design and failure to investigate problems thoroughly.”

Though many environmentalists will focus on preventing new plant construction, closing down existing plants and preventing relicensing, even in the best-case scenario we are stuck with managing a fleet of old nuclear plants and the 40,000 tons of radioactive waste they have created. Thus, it is imperative to overhaul the inadequate, industry-dominated safety culture that has developed over the past twenty years. This eroded safety culture is a source of serious danger—and it must be fixed.