THE HORROR THAT HAUNTS Willow Island, W. Va., is recorded on the granite stones Baptist church at the south end of town. Stone after stone, row after row: "Died April 27, 1978." "Died April 27, 1978." "Died April 27, 1978." "Died April 27, 1978."

On that morning, 51 men died when a scaffold collapsed at the Monongahela Power Co. plant less than a mile away. It was America's worst construction accident in recent times.

Federal safety officials rushed to the plant and pledged before the TV cameras to take steps to stop similar construction-site catastrophes. Seven years later the Occupational Safety and Health Administration, now under control of the Reagan administration, still hasn't imposed rules aimed at preventing such tragedies. The regulations are stalled because OSHA and the White House Office of Management and Budget can't agree on how much a construction worker's life is worth -- in cold, hard cash.

This is not a matter of deciding insurance payments to survivors or workmen's compensation standards: It is the Reagan administration's attempt to rewrite the book on how the nation balances the worth of a person's life against the money big business should pay to protect that life. Placing a cash price on everything -- even seemingly unmeasurable items such as the life of a construction worker -- is part of an arcane strategy borrowed from the business world called cost/benefit analysis. Its goal under Reagan: to weigh the price of any federal regulation of business against its payoff in "lives saved" by the precise tabulation of numbers.

Behind computations that would make even a tax accountant scratch his head, however, are signs of an old- fashioned political agenda -- one that could throw considerable dollar advantage to Reagan's conservative business constituency and go far toward embedding this administration's conservative philosophy in one piece of bureaucratic minutiae that will affect millions of people for years to come.

All recent administrations have tried to incorporate various cost controls into government thinking, often in an attempt to justify decisions once based on simple political mathematics. The Kennedy administration created the "planning, programming, budgeting system" in Robert S. McNamara's Defense Department. President Johnson tried to broaden McNamara's system to domestic programs. Gerald Ford ordered agencies to issue "inflation impact statements" and President Carter used "regulatory analyses."

But Ronald Reagan is the first to insist that cost/benefit studies be used by all regulatory agencies. Under Reagan, agencies must calculate how many benefits, in cash, a federal regulation will produce for the public. That figure then is compared with what industry and the business world will have to pay if the regulation becomes law. If a regulation costs businesses more than it pays the public, it's sandbagged.

The politics behind this seemingly expert, nonpolitical technique, however, are a variant of the computer clich,e, "garbage in, garbage out." What is pumped into the cost/benefit formula determines its conclusion.

The reason: if a human life is worth, say, $10 million, then it makes sense under cost/benefit analysis for the government to require businesses to spend up to $10 million per person to protect workers. But if a human life is worth only $50,000, then that is all the government can expect a business to pay to protect anyone from death and injury.

"It's ludicrous," says an OSHA economist who asked not to be named. "Here we are, coming up with unprovable calculations and then citing them as reasons why the public should or should not be protected. It would be laughable, except that human lives are at stake. This is real."

In pursuit of cost/benefit precision, the administration has abandoned the old idea that the value of a person's life is determined by his projected lifetime earnings -- a method widely used by the insurance industry and courts. Almost by coincidence, OSHA found a "better" way to determine the value of a human life in the work place -- an approach that spread to other agencies and is more in keeping with the conservative notion that all people, rich and poor, are the keepers of their own fate.

This method assumes that people who choose to do dangerous work -- such as pouring the cement walls of a power plant -- must value their lives less than someone who chooses to work in an office. Ergo, the dollar value of the construction worker's life should weigh less heavily when balanced against the huge costs of making his work place safer.

The result: fewer costly safety requirements for businesses. The long-run impact of such a change, administration critics contend, goes far beyond the work place. Applied across the board, this concept could have dramatic impact in balancing, say, the costs versus the benefits of protecting school children from cancer-causing asbestos and the public from dangerous toxic waste dumps.

"Cost/benefit studies are basically political documents and always have been," says Margaret Seminario, a safety expert at the AFL-CIO. Big labor should know: during the final years of the Carter administration, the Labor Department cited cost figures pushed by unions to show that almost any regulation was worth its price.

The politics over the value of life can be focused on a single OSHA regulation -- revisions to Subpart Q, 29 CFR 1926 of the concrete and masonary standard, which grew out of the Willow Island tragedy. But cost/benefit also will shape nearly all federal regulations in the future -- from controlling poisonous gases in hospitals to limiting cotton dust in textile plants to cleaning up acid rain. Under Reagan, it has become the hottest numbers game in town. It remains Ouija-board science, however, with its roots deeply embedded in politics.

COST/BENEFIT ANALYSIS first emerged in the Rivers and Harbors Act of 1902, which mandated that waterway improvements cost less than the value they would produce. Such improvements were done by the Army Corps of Engineers, and for decades it was the only agency to use the supposedly nonpartisan, rational procedure. Ironically, during this same time the Corps became known as a preeminent pork-barrel agency, a reputation it still maintains. A recent example: Critics tried for 13 years to stop the recently dedicated 234-mile Tennessee-Tombigbee Waterway that connects the Tennessee River and the Gulf of Mexico, but southern congressmen were determined to push it through -- despite a 1981 investigation by the General Accounting Office that found the Corps had drastically underestimated the waterway's costs and overestimated its benefits by more than 10 times. Overall, the GAO said, the waterway would cost more than $2 billion and produce a mere $12 million in benefits.

Other federal agencies shunned cost/benefit studies. OSHA, the Enviromental Protection Agency and the Consumer Product Safety Commission, for instance, were created to protect the public -- regardless of cost. That changed, however, with the austerity of the late 1970s and the skyrocketing costs of protection.

OSHA was typical. Created in 1971 at big labor's behest, it spent its first years henpecking businesses about minor safety and health problems, such as using ladders properly. But during the Carter administration, it began issuing rules to protect workers from exposure to dangerous substances. Suddenly, businesses were faced with billions in potential costs. OSHA's 1978 "carcinogen policy," designed to protect workers from cancer-causing substances, alone was estimated to cost industry as much as $448 billion.

"Our agency estimated costs," recalls OSHA economist Larry Braslow of the pre-Reagan days. "But it was always assumed that the benefits would outweigh them."

Enter Reagan with Executive Order 12,291: It mandated cost/benefit tests on regulations expected to cost industry $100 million or more, which includes most new regulations. Agencies were to submit the studies to OMB for review -- and the numbers game had begun.

W.KIP VISCUSI became the administration's expert on the value of life almost by chance. A business professor at Duke University, Viscusi won his status in the spring of 1982 when OSHA and OMB were locked in their first major cost/benefit fight. OSHA wanted manufacturers to put labels on barrels of chemicals, paints and adhesives warning workers about dangerous contents and telling them what to do if exposed. Under Reagan, OSHA already had watered down the rule to please industry, but OMB didn't think any regulation was necessary, even though industry, labor unions and members of Congress supported it. Despite the rare consensus, OMB continued to say no.

In a confidential 14-page critique for the White House, OMB compared the labels to those used on cigarettes, claiming such warnings did little good. OMB also said OSHA had vastly overestimated the regulation's benefits.

OSHA decided to get an independent review. "I needed someone who the Democrats wouldn't think was a right-wing tool and the Republicans wouldn't think was a 'flaming, Commie, pinko liberal,' " recalls Mary Ellen Weber, OSHA's former regulatory analysis chief. "Viscusi was perceived as being politically neutral."

Viscusi delivered his analysis on March 16, 1982. The next day, the White House overruled OMB and told OSHA to issue the labeling rule. At the time, news reports speculated that the White House had bowed to congressional pressure in favor of the rule, but at OSHA Viscusi was given the credit.

"His paper hadn't changed my mind," recalls Christopher C. DeMuth, the former OMB official who had stonewalled the labeling rule. But it influenced Raymond J. Donovan, then secretary of Labor, and other key players. "Ray Donovan called me the day the paper was delivered," DeMuth recalled. "He had read it and mentioned that he was impressed."

Viscusi agreed that OSHA had overestimated the labeling rule's benefits. But he said the idea still was cost-effective. Viscusi then offered his free enterprise approach to the cost/benefit game. In the past, federal agencies decided how much a "life saved" was worth by projecting lifetime earnings. Known as the human capital approach, the method devalues the worth of women, who generally earn less than men, the elderly, who have fewer years of employment left, and the poor, who are likely to earn less in their lifetimes.

Viscusi's new method estimates how much a person's life is worth based on his "willingness to pay" to avoid death. Viscusi first identified dangerous occupations based on death and accident reports. Then he compiled a list of how much workers in those occupations are paid. By meshing the two, he arrived at the dollar amount workers accept to take risks. The key: Workers with risky jobs, Viscusi said, valued their lives less than those who held safer jobs. Viscusi concluded that OSHA should use a range of $2 million to $3 million per life in its cost/benefit calculations. But, he said, some workers in high-risk jobs were worth as little as $650,000 each.

The old and new methods actually make little difference in estimating the value of life in many instances, but Reagan officials were taken with the freedom-to-choose idea behind Viscusi's approach. "It sounds better than the human capital method," says Washington economist Ruth Ruttenberg. "But both are equally ridiculous."

OSHA economists decided to use $3.5 million as the standard value of a life, based on Viscusi's report. OMB, citing Viscusi's "willingness to pay" calculations, preferred Viscusi's lower estimates.IT SEEMS THAT OMB moves, like God, in the most mysterious ways. Conversations between OMB -- the right arm of the Reagan White House on regulatory matters -- and agencies are confidential, and it is difficult to determine how much influence OMB exercises. Rea- gan's critics claim OMB intimidates agencies, often in behalf of industry, a charge OMB denies.

"What has happened," says David C. Vladeck, an attorney with the Ralph Nader- founded Public Citizen Litigation Group, "is that OMB now calls all the shots . . . It controls the purse strings and it is the administration's right hand. You don't pick a fight with the biggest kid on the block."

Douglas H. Ginsburg, who heads the OMB office that reviews cost/benefit studies, shrugs off such charges. "We can't make a decision for any agency," he says. "Indeed, the only thing that we can do, if we can't reach an agreement about the interpretation of evidence, is return the rule to the agency for reconsideration. We can't prevent them from going forward and promulgating a regulation."

Technically, Ginsburg is correct. But ask Ginsburg to name an agency that has done so.'I don't think there is such an instance . . . ," he says. "No, none."

In October, Rep. David Obey (D-Wis.) released internal OSHA documents that showed for the first time how OMB was stalling a regulation -- the concrete standard developed because of Willow Island. The documents showed that OMB wanted to use $1 million as the worth of a construction worker's life -- not the $3.5 million suggested by OSHA. In confidential documents, OSHA has shown that the lives of 23 construction workers a year could be saved by the new regulation, which requires better bracing and shoring of freshly poured claimed the regulation would cost industry $27.3 million to implement -- about $1.1 million for each life saved.

"Using the $3.5 million figure, the regulation could pass a cost/benefit study," an OSHA economist said. But if OMB's $1 million figure were used, the regulation would be in trouble.

OSHA economists feared OMB was trying to force the agency to weaken the standard. It already had tinkered with it to reduce costs, internal agency reports show. At one point, OSHA estimated the regulation would cost $49.3 million and save 26 lives. During a review, however, economists found that a single part of the rule accounted for $22 million in costs -- and saved only three lives. That provision would have required construction companies to keep all employes out from under buckets of concrete being lifted by cranes. The provision was scrapped as economically unjustified.

Despite Obey's charges that OMB was trying to gut the concrete standard by underpricing life, Ginsburg's deputy at OMB, Robert P. Bedell, claims OSHA misunderstood his agency request for a reassessment: "We never tell anyone what figures they have to use." He says OSHA was asked to perform a "sensitivity analysis" using figures both higher and lower than $3.5 million "solely for comparison purposes." OSHA is now doing just that.

An OSHA economist is not convinced: "We suspected that OMB was trying to set $1 million as the standard for the value of life."

This was not the first time OMB has been accused of pressuring agencies to use the $1 million value-of-life estimate. In a recent letter to EPA, Bedell suggested that each cancer death the government could avoid by banning asbestos be valued at $1 million. Bedell also told EPA that when it did a cost/benefit analysis for asbestos, it should "discount" each life because a young person exposed to asbestos is likely to have 30 or 40 years of good health before discovering that he has cancer.

Using the "discounting" method, a child in school would be worth about $22,000 today if his whole life was valued at $1 million. As a result, expenses exceeding $22,000 per child to clean up asbestos in schools wouldn't pass cost/benefit analysis. EPA has estimated that it will cost $96,000 to $208,000 per child to remove or contain asbestos in schools.

Bedell's letter prompted Rep. James J. Floria (D- N.J.), to accuse OMB of serving as Reagan's "regulatory executioner."

JERRY AVORN, a physician and professor of social medicine at Harvard University, isn't taken with W. Kip Viscusi's idea that all workers have the freedom to chose the work they do.

"To begin with, no one knows, not even OSHA, what the risk of death from a given job is with any certainty," says Avorn. And even if workers are aware of hazards, he says, they may not have any choice but to accept them.

"If you are a guy with just a high school education and you have worked in a chemical plant before and that is all that you know how to do and you live in an area where there are not many jobs around, then dammit, you are going to go work in a chemical plant."

Viscusi acknowledges that some people don't pick jobs according to risk, but he's also found that many workers are quite knowledgeable about the risks they face. He says government safety standards might sometimes hurt workers' chances of earning a living.

"Are you really doing people a favor by saying, 'You can't face this risk because I put a high value on my life and am going to prevent you from doing it?'ter, Viscusi asks, to let the marketplace decide?

"I would never accept a hazardous job as a way to make more money," he says, "simply because I am in a high-income group and I don't need the money." But, he says, if a person is poor and has a choice between a safe job that pays $20,000 and a more dangerous job that pays $25,000, "you may find it attractive to incur more hazard for the extra $5,000."

Yet, union critics claim a marketplace approach would lead to a return to the days of the early 1900s when businesses found it cheaper to hire new workers than make work places safer.

Such critics usually condemn the entire cost/benefit process, claiming that it is impossible to assign monetary values to the view from the rim of the Grand Canyon or the value of a human life. They cite an infamous example involving Ford Motor Co. of how social ethics and values can be lost when money becomes the deciding factor. In the 1970s, a cost/benefit study by Ford determined it would be cheaper to pay off accident victims in- jured in early-model Pinto autos, which had gasoline tanks that often exploded into flames after minor collisions, than to recall the 1.5 million cars and fix them.

Reagan administration officials, such as Ginsburg, claim cost/benefit studies are "simple common sense tools." They say society already puts various price tags on the value of life: for instance, the number of murders in Washington could be reduced if taxpayers would pay to have police officers stand on every corner.

So far, the administration's argument is winning. Viscusi's calculations are being used by other agencies and at least one hospital has adapted cost/benefit principles to decide such things as whether the life of a handicapped child is worth saving.

The spread of cost/benefit analysis, particularly into medicine, says Avorn, is "terrifying." There are some areas of life, he says, where such studies "lose both moral and intellectual legitimacy."

"It is not that the value of a human life in infinite," Avorn says. "As a nation we surely do not behave as though we believe that to be the case. Rather, the alleged quantification is so imprecise and so laden with value judgments that such numbers are useless."

ECONOMIC QUESTIONS seem distant from the banks of the Ohio River, which flows swiftly by Willow Island and the small hamlets to the north of Belmont and St. Marys. The countryside here is a constant series of rollercoaster hills, sprinkled with old shacks and modern brick houses, many with giant dish television antennas anchored beside them. This was home for many of the men killed at Willow Island.

"Nearly everyone in Pleasants County knew someone at the plant," recalls Arthur Olds, St. Marys' mayor. The jagged marks on Tower No. 2 are still there, where the steel- and-wood scaffold broke loose. It was after 10 a.m., and the 51 men on the scaffold had started to pour the 29th ring of concrete. Cooling towers are built in layers, with each new day's concrete poured on top of the last. This day's pouring was at the sheerest spot in the tower's facing, only 18 inches thick.

The scaffold, a Hamon Hyperbolic Tower Erection System, ringed the lip of the tower, with four levels of work space draped over each side. It used beams and bolts to slowly move up the wall, 41/2 inches at a time. Two-thirds of the bolts holding it were anchored in concrete that had been poured yesterday afternoon, the remainder in concrete three days old.

It had been chilly and misty the day before, but the sun was shining on April 27 as the first bucket of concrete was brought toward the scaffold. Suddenly, there was a loud snap. As the construction crew on the ground watched helplessly, a section of the scaffolding broke loose from the wall. The workmen on the scaffold scrambled toward areas that still were attached to the rim, but as they ran, the rim itself began to peel away. Within seconds the entire ring had fallen 160 feet, pulling the scaffolding outside the tower inside with a roar.

"I carried five or six men out, but I just couldn't stand no more," Gene G. Johnson, a construction worker told the Parkersburg Sentinel. "One guy was busted wide open. His insides were spilling out. I just couldn't stand it no more." OSHA investigators later determined that some of the bolts anchoring the scaffold were attached to still uncured concrete.

Teachers in Pleasants County today don't tell students to take papers home for "Mommy and Daddy to see," says J. McGoldrick, managing editor of the St. Marys Oracle. "They know that some of the children don't have any daddies at home." Lee and Molly Steele lost four sons in the accident: Larry Gale, 32; Ronald Dale, 30; Earnest, 29; and Miles Eugene, 26. Six of their other relatives also were killed. The boys grew up in the hills, were active in the Cow Creek Methodist Church, and stayed close to home. When it came time for them to choose a job, they all joined their father, uncle and cousins in the local ironworkers' union.

Most of the families who lost someone in the accident hired Robert T. Goldenberg, a Parkersburg attorney, to file wrongful death claims against contractors. The average award was $400,000, but because widows and their children also were granted workman's compensation benefits, the total package for some exceeded $1 million. Goldenberg contends the settlements were the best the families could get and among the highest in the state. Even so, he says, they weren't enough. "They never are," Goldenberg says.

"What is a construction worker's life worth?" asks Dr. Richard Hamilton, 82, who delivered as babies many of the men who died at the tower. "That's a bit like asking, 'What is truth?' his head. He is silent.

It is late now. A young man and woman step down from the entrance at the Willow Island Baptist Church where a revival meeting has just ended. They pass, holding hands, a few feet from where the Steele brothers are buried. To the north, the lights on the 430-foot twin cooling towers blink. There is a small plastic chain around the Steeles' family plot.

In his research, Viscusi discovered that society is often irrational when it comes to death. "A trapped coal miner whose plight is featured on the evening news and whose fate now rests in our hands rather than in his own may evoke more public concern than the unidentifiable beneficiary of automobile passive restraint systems," Viscusi wrote. "This difference in attitudes is usually taken as an index of society's excessive concern for identified lives."

How much is a construction worker's life worth? The question could only be asked in Washington.

In Willow Island death wears a face.