One hears a lot about profits and petroleum when times are good (and very little about losses when times are bad), but almost nothing about prophets in the petroleum business once their big splash in the press has passed.

One of the most sensational forecasts was a CIA study six years ago, unclassified and entitled "The International Energy Situation: Outlook to 1985." Notable at the time for its estimate that the Soviet Union would soon become a net importer of oil, it caused a stir by predicting that by 1985 the Soviet bloc would "require a minimum of 3.5 million barrels of imported oil" every day. Today Soviet production is over 400,000 barrels a day higher than the highest level the CIA foresaw. The Soviets are dumping well over a million barrels a day at cheap prices on the international market.

The Soviet story grabbed headlines and drew attention away from the rest of the conclusions the CIA had arrived at in its forecast. These are the ones that really make interesting reading now. For instance, "We estimate the 1980 demand for OPEC oil will be about 34 million barrels a day, 2 million barrels a day more than in 1977." In 1980, OPEC produced an average of about 27 million barrels a day, making the CIA estimate off by 7 million barrels, or over 20 percent. For later years, reality and forecast spread much farther apart.

OPEC produced about 22.6 million barrels a day on the average in 1981 and 18.4 million in 1982. Current production is about 13 million barrels a day, although the year's average may approach 18 million if the world economy picks up a lot. OPEC's current rate of production is low, to be sure, but there it is. CIA's 1977 estimate of 1983 production, interpolating from its projections for 1980 and 1985, would have been over 40 million barrels a day, three times greater than what is actually the case. Although it will rise later this year, OPEC's average production will certainly be much less than half of the rate the agency foresaw.

It expected Saudi Arabia, for instance, would be producing about 12 million barrels a day in 1982 and "from 19 to 23" million barrels a day in 1985. Instead of 14 million barrels a day, which is where the CIA forecast would put current production, Saudi Arabia today is pumping less than 4 million and may reduce that. In its closing paragraphs, the 1977 report says, "By 1982 or 1983, sizable price increases are inevitable unless large-scale conservation measures cut demand sharply." In the real world, prices cut demand, not conservation.

The CIA's 1977 crystal ball foresaw Free World demand rising to about 55 million barrels a day in 1980 and growing to about 70 million in 1985. In fact, free world demand in 1980 was about 45 million, almost 20 percent less. Now demand is somewhere over 40 million, say 43 million. That makes its estimate of today's demand off by about one-third.

In 1979 the CIA put out another, longer study on the same subject. It noted the escalating prices of the time, but still came out with a prediction of 30 million barrels a day as OPEC's likely production rate for 1982. Instead of 100 percent error, it came down to 40. Worldwide demand was also overestimated by about one-third.

It recognized that the price rise it had foreseen in 1977 for the early 1980s was taking place in 1979, but the full influence of prices on consumption was missed. One shouldn't blame the CIA alone for this failure. Industry made much the same mistake. Exxon, for example, bet billions that high prices and oil shortages would make a project to extract oil from shale rocks viable. It was as wrong as the government.

The point is not to hold hard-working analysts up to ridicule, but to ask why they went wrong. It seems to me in retrospect that the CIA basically failed to ask itself how we were going to pay for all that oil at those prices. If you had looked at the problem, you might have seen that something had to give.

What gave was oil consumption. We suddenly became much more efficient in the use of oil after the 1979 price rises. The trend of smaller increments of oil consumption per unit increase in GNP that had started in the late '60s suddenly took off in 1979. In that year the United States consumed an average of 18.5 million barrels a day of oil products. In 1982, we consumed about 15.2 million, an 18 percent drop. The GNP was about the same for both years.

What the analysts missed were the people adding more insulation to houses, buying fewer and smaller cars and keeping thermostats lower, and industry's cutting energy use drastically. That explains in part why the estimates were wrong.

There were other factors, among them one worth mentioning. President Carter, I think, wanted the country to suffer for its wanton, wasteful practices. He felt we would all be better for having learned a lesson. The apocalypse these reports foresaw was the exhaustion of oil resources before we had prepared alternatives.

This is a problem, to be sure, but it seemed to me at the time that Carter relished the contemplation of this particular form of judgment day. The reports that foresaw that day fast approaching were just what he wanted--and just what the CIA gave him.

There are some lessons to be learned from this sorry record, I believe. Perhaps the most important is that our information can be inadequate even on vital topics and the successful economy is the one that can stay flexible, capable of adaptation to whatever happens.