following growth in preceding quarters that was more robust than earlier believed -- left analysts puzzled last week about where the economy is headed.

The Commerce Department said the economy had grown more strongly in 1985 and the first quarter of 1986 than had been reported earlier, but then estimated only a 1.1 percent rate of increase in the gross national product, adjusted for inflation, for the second quarter.

In the second quarter, American consumers, businesses and governments actually increased purchases of goods and services bought for their own use at a hefty 5.5 percent annual rate, even after taking inflation into account.

Demand growth that strong usually would mean the U.S. economy was doing well, but the summer of 1986 is hardly a "normal" period in American economic history. The 5.5 percent rate of increase in final demand by domestic purchasers translated into anemic 1.1 percent GNP growth. Why did the demand growth not produce a bigger number for GNP?

Two equally large factors intruded: First, a surge in imports of foreign goods satisfied a substantial part of the increase in demand; and second, business bought fewer goods to put into their inventories than they had bought in the first quarter. Both developments meant the demand for domestic production rose very little.

Meanwhile, the 5.5 percent rate of increase in domestic final demand included what many analysts believe to be an unsustainably large rise in consumer spending -- up at a 5.9 percent rate. There was also a significant increase in housing construction and a small decline in business investment in new plants and equipment.

It added up to a peculiarly mixed bag that left forecasters wondering what comes next.

Part of the further deterioration in the U.S. trade balance was due to a sharp increase of about 1 million barrels a day in the volume of oil imports. That increase is not likely to be repeated this quarter, even if the volume does not fall.

That's important, analysts say, because it is the steady worsening of the trade deficit, not just its size, that keeps sapping the growth of demand for domestically produced goods and services. If the trade deficit does begin to shrink, then that will become a plus rather than a minus for GNP, corporate profits and jobs.

But there could be some improvement as a result of the large decline in the value of the dollar during the past year and a half, compared with the Japanese yen, the West German mark and some other currencies. A declining dollar makes imports from such countries more expensive.

The initial effect of the dollar's drop was to lower margins significantly from artificially high levels produced by the rapid increase in its value between 1983 and early last year.

Alan Greenspan of Townsend-Greenspan & Co. estimates that operating profit margins on U.S. merchandise imports peaked at about 11 percent in February 1985. Margins fell for about a year and have started to recover, and that recovery has begun to show up in the price of imports. Between January and May, the latest figures available, prices of non-oil merchandise imports shot up by more than 11 percent.

"During the same period, total merchandise import prices fell by about 5 percent, but this, of course, is due to the dramatic decline in the prices of petroleum products," Townsend-Greenspan told its clients recently. By the end of this year, it continued, import prices can be expected to go up another 10 percent "at a minimum," adding about 1 percentage point to the rate of inflation during the period.

These large increases in the cost of imported goods ought to encourage purchases of relatively more U.S.-produced goods and, therefore, at least some reduction in the trade deficit. Just when such a turn will come remains only a guess.

Inventories are somewhat less of a puzzle, though they, too, can behave in unpredictable ways from one quarter to the next. The major factor in the second quarter appears to have been the big effort by the automobile manufacturers to reduce the stocks of unsold cars on dealers' lots by offering cut-rate financing.

While the rate at which businesses will be adding to their stocks of goods on hand could fall further this quarter, most analysts do not expect nearly as large a reduction as occurred in the second quarter. In other words, inventories should be much less of a drag on GNP this quarter, and possibly could be a positive factor.

The Reagan administration will release a revised economic forecast next week. It will show an increase in real GNP this year of less than the 4 percent predicted in February. However, it will also predict more growth, probably of about 4.5 percent, for 1987, which earlier also was seen as a 4 percent growth year.

The new forecast will call for faster growth in the second half of 1986 than the 2.5 percent rate of the first half, probably enough to produce about a 3 percent to 3.5 percent increase between the fourth quarter of last year and the fourth quarter of 1986.

However, at least one senior administration economist said things could turn out much better than that should trade and inventory figures both turn around together.

"In the second half of the year, you have the potential for one hell of a growth quarter" should that happen, the economist said. "That would represent a huge change in the pressure on some parts of the economy."

The administration economist said the new forecast will not be so bullish because the timing of such changes are impossible to predict.

Meanwhile, smaller gains in consumer spending probably are on the way, the economist added. The 5.9 percent rate of increase in the second quarter might have been overstated as a result of difficulties in dealing properly with declining oil and gasoline prices, so consumers appeared to be buying more of those products than they really did. Spending for consumer nondurables other than oil and gas was up at about a 1.5 percent rate, the economist points out.

Another factor tending to slow down consumer spending, which in the second quarter accounted for 65 percent of GNP, will be smaller increases in personal income in coming months, the official said.

Greenspan pointed to another negative for consumers: Household debt has increased to the point that, despite lower consumer and mortgage interest rates, Americans are obligated to spend more than 30 percent of their cash disposable income to make their monthly payments. That compares with a peak of about 26 percent in 1979 and levels well below that during much of the 1970s.

Moreover, some households naturally have debt repayment burdens higher than the average.

"Should the economy dip into a recession and consumer incomes decline, it is apparent that this would result in severe hardships and possible debt defaults by a number of American families. However, of more immediate concern is that high debt-servicing charges usually restrict the flexibility of households in retail markets.

"Hence, it is difficult to see the strong second-quarter consumer behavior maintaining anywhere near the same pace during the second half of this year," Greenspan concluded.

The same sort of reasoning could be applied to business capital spending, which fell at a 2.6 percent rate in the second quarter after a 15.1 percent rate of decline in the first quarter. Much of the drop so far this year is a direct result of sharp cutbacks in oil industry investment because of plunging oil prices. However, Greenspan said, businesses also have very high debt burdens that have executives worried.

Interest payments soared from around 26 percent of corporate cash flow in 1977 to 43 percent in 1982, before dropping to 36 percent at the end of 1983. But since then, he said, such payments have risen again to nearly 39 percent of cash flow.

With low inflation, corporate executives are reluctant to make capital investments that would add to their companies' debt-service costs and, like the households mentioned earlier, make them more vulnerable to a squeeze should a recession occur anytime soon.

Meanwhile, the large amounts of unused production capacity in most industries and some added uncertainties created by the pending tax revision bill before a House-Senate conference committee also are having a negative impact of some dimension on capital spending.

The remaining portions of GNP, new housing construction and government spending, also are not apt to add much to growth.

Housing remains one of the strongest parts of the economy, chalking up a 15.4 percent rate of increase in the second quarter. Nonetheless, housing starts are not likely to increase above the current annual rate of about 1.85 million units, and that means their positive impact on GNP growth will diminish.

Finally, government spending is seen by many analysts as adding little to growth and perhaps turning out to be another negative factor if Congress is successful in reducing the federal budget deficit as called for under the Gramm-Rudman-Hollings deficit reduction law. Even if the law's deficit targets are exceeded, federal government purchases of goods and services will be cut to some extent.

What does this produce in the way of economic growth during the next 18 months? The predictions range from no growth at all for the rest of this year to an acceleration of growth to the 4 percent to 5 percent range. Take your pick.