About 2 million people die every year in the United States, and for one out of six - about 365,000 - the cause of death is cancer.

A government task force has estimated that 90 per cent of the cancer is caused or fed by various factors in the environment.

An occupational health expert Nicholas A. Ashord of the Massachusetts Institute of Technology Center for Policy Alternatives says there is general agreement among cancer researchers and environmentalists that one-fourth of one-half of this environmental cancer is caused or complicated by occupational factors

If all this is true, occupational factors cause or conctribute to somewhat between 85,000 and 170,000 cancer deaths in this country each year, and that is only part of the toll. There are other occupational deaths from other diseases.

These numbers are frightening - but they are also only guesswork. Scientists do know some things about occupational diease. Cancer and certain other disease rates are higher among certain kinds of workers - asbestos and rubber workers, for example - than in the population at large.

The National Cancer Institution published a cancer atlas of the United States two years ago, and Anthony Mazzocchi, legislative director o fthe Oil, Chemical and Atomic Workers, say, "That's a map of our union."

But these are only fragmennts. No one really knows how many cases of occupational disease occur in the country each year. The Public Health Service has estimated 390,000, but is not sure, nor does anyone know how many deaths are mainly attributable to occupational factors. Ashfold notes wryly that estimates range from 1,000 to 100,000.

It was into this area - where much more is unknown than known - that Congress tossed the federal bureaucracy when it passed the Occupational Safety and Health Act in 1970.

The act created a new regulatory agency, the Occupational Safety and Health Administration. The agency's first task, spelled out in the law, was to issue as its own initial regulations a thick book of so-called "consensus standards" - safety and health standards that had been voluntarily adopted over the years by various trade associations and professional societies.

Congress intended that these consensus standards would tide the agency over until it could issue new standards of its own, but most of them have remained on the books unchanged.

For mose of its six years of existence, OSHA has been regularly denounced by both business and labor - by business for being intrusive and picking costly nits and by labor for giving little or no protection to workers in peril.

Neutrals in the field say both criticisms are correct, and that the adopted consensus standards are a major reason. Those standards, which make up a book of more than 600 pages, deal mainly with safely - preventing accidents like cuts, falls and fires, keeping hands away from saw teeth, making sure ladders are sturdy, subjects like that.

Many of these safety standards are painlly detailed, and when OSHA inspectors showed up to insist they be followed to the letter and the inch, businessmen went up the wall. As Morton Corn. administrator of the agency in the last year of the Ford administration, recently put it, in semi-governmentese, "Many of the early consensus safety standards were, indeed, associated with nitpicking ingredients."

Meanwhile, there were only minimal consensus standards in the newer, more difficult and dangerous area of health, which in OSHA terms means mainly chemical poisoning. There is a table setting forth maximum allowable worker exposures to about 500 chemicals, and that is all. It is mainly over this that labor has failed at the agency. There are thousands of poisonous chemicals now in use that are not on the table, and for some that are there, the unions say the allowable exposures are too high.

There were other problems in the administration of OSHA. It is generally agreed that in the early years little central direction was given to the agency's field offices across the country. No system was set up to make sure that inspectors went to worst offenders first; it is not even clear the agency had enough data to know which industries and companies were the worst offenders.

One agency veteran says, only half in jest, that in choosing which companies to inspect in those first years, inspectors in the field operated mainly "by the seat of their pants and out of the Yellow Pages."

One school of thought has it that this early aimlessness in OSHA was deliberate and political. In 1974 Watergate investigators found a 1972 memo from OSHA's first administrator, George C. Guenther, saying the agency would hold off on the issuance of "highly controversial standards" during the presidential campaign and suggesting that, in fund-raising," the Republicans stress "the advantages of four more years of properly managed OSHA."

The Nixon-Ford administrations also had some less specifically political, more philosophical problems with OSHA. The question kept - and keeps - recurring whether the benefit from a given health or safety standard is worth the cost.

But most experts agree it would be wrong to ascribe all the agency's problems to the people who have run it over the last six years, to claim that if only OSHA had been in different hands, the problems would have disappeared.

It turns out that regulating occupational safety and health - especially health - is harder than the politicans implied or understood it was when they passed the health and safety law.

There are 22,000 toxic substances in use in industry today, and hundreds more are being invented every year. OSHA has been formulating regulation for one at a time. It cannot keep up.

Its problem is compounded by the difficulty of tracing and proving the health effects of even one substance. Not only are these effects complex, they are often delayed. Cancer may come at age 50 from exposure to a chemical at age 30, and it is hard to make the connection.

There is also an unresolved problem with the goals of the health and safety act. The basic goal, of course, is reducing injury and illness, but the question remains: how far should the government go?

Congress did not answer.

A third problem has to do with small business.

Of the estiated 5 million places of business covered by OSHA, about 80 per cent have fewer than 25 employees, though all of these together have only 30 per cent of the covered work force.

It is one thing to insist that U.S. Steel read and comply with 600 pages of safety and health regulations, those who favor an exemption say, but it is ridiculous to make the same demand of the owner of a neighborhood hardware store. They say the regulations are an impossible burden on small businessmen. Those on the other side say it is unfair to leave employees of small businesses without protection.

But for all the talk about OSHA's having forced small businessmen out of business, spokesmen for the U.S. Chamber of Commerce and National Federation of Independent Business said they could not point to any examples. And the chemical walkers' Mazzocchi adds, "Not everyone has the right to be in business."

Mazzocchi and others in the labor movement are hopeful - though guardedly so - that the Carter administration will find ways to turn OSHA around.

A few important changes were made last year by Corn. The agency is seeking some way to issue health regulations that would govern whole groups of chemicals. It is also scheduled to begin a "national emphasis program" under which inspectors will focus on high-risk industries - foundries, to begin with. Whether these steps will make a genuine difference no one knows.