The Commerce Department released major revisions yesterday to its previous estimates of the gross national product, stretching in some cases back to 1929.
The revisions, based on more complete data, better estimating techniques and moving the base year for calculating real output from 1972 to 1982, showed that current-dollar GNP -- the nation's total output of goods and services -- has been higher than previously thought. With the revisions, GNP passed the $4 trillion mark sometime about midyear.
But while current-dollar GNP is higher, the rate of growth of real output -- that is, GNP adjusted for inflation -- averaged 2.5 percent a year from 1972 to 1984, 0.2 percent less than previously estimated. The rate of increase in current-dollar GNP over the same period was unchanged at 9.9 percent annually.
The new numbers indicate that the recession of 1981-82 was more severe than calculated earlier, and that the recovery since then, while still strong, has been somewhat less vigorous than thought.
The drop in real output during the recession totaled 3.4 percent, rather than 3 percent. Over the 12 quarters since the bottom of the recession in 1983, real GNP has gone up at an average annual rate of 4.3 percent, instead of 4.6 percent. Total growth over the period was 13.6 percent, instead of 14.3 percent.
Officials at the Commerce Department's Bureau of Economic Analysis, who are responsible for the GNP estimates, said that despite such changes, they regard the major overhaul of the numbers as confirming that the earlier figures were telling just about the right story.
"The numbers mean that the picture of the economy is pretty much the same as it was," said Robert P. Parker, associate director of BEA.
Many of the changes throughout the GNP accounts reflect more than anything else the shift to 1982 as a base year. The shift is part of a general update in the numbers that will allow analysts to better follow the future course of the economy.
The use of weighted numbers from the mix of goods and services bought by households, businesses and governments in 1972 was distorting particularly the division of current-dollar GNP into its real and its inflation components, analysts said.
For example, with an unchanged rate of increase in current-dollar GNP and a lower rate for real GNP over the 1972-84 period, inflation as measured by the GNP implicit price deflator -- which is the ratio of current-dollar GNP to real GNP -- was revised upward.
However, the deflator is affected by the changing mix of goods and services actually produced. Another measure, the GNP fixed-weighted price index, which many economists believe is a better measure of inflation, was revised downward. Instead of an annual increase averaging 7.3 percent since 1972, that index now shows an average rise of 6.6 percent a year.
The lower rate of increase in that price index was essentially entirely due to using 1982, instead of 1972, as a base year. Parker told reporters that either year could be used as a base and that both are accurate. However, because of the large changes in relative prices of different types of goods and services over that 10-year period -- such as soaring oil and medical-care prices relative to, say, the cost of clothing and appliances -- the continued use of 1972 weights would give steadily less useful information about today's economy.
The revisions also disclosed several other changes. Among them was that in the 1972-84 period, real personal consumption rose slightly less rapidly than estimated earlier, 2.7 percent a year instead of 3.1 percent.
Similarly, real business investment in new plants and equipment increased an average of only 2.5 percent annually, instead of 3.1 percent. That change surprised many analysts because of a major change in calculating computer prices, which are now estimated to have been declining 10 percent a year, instead of remaining constant as Commerce had been assuming. Even though that new estimate greatly added to computer purchases, investment in other types of business equipment was revised downward. Investment in inventories was also revised downward.
Significant upward revisions in personal income also helped raise estimates of saving as a share of disposable personal income. For the years 1977-81, the rates were revised up by nearly 1 percent, to an average of 7.1 percent.
However, the upward revisions since 1981 were only about half as large, and the rate has averaged less than 6 percent. In the third quarter, both the long-term revisions and more complete data for the quarter caused an upward revision in the saving rate to 3.7 percent from the previously reported 2.7 percent, still low but no longer a post-World War II record.
The revisions were based on a wide variety of methodological changes and new information. Some of the major sources of new data are several economic censuses conducted in 1977 and 1980, with only preliminary figures available from the latter.
Improved adjustments for misreporting on income tax returns was used to come up with substantially higher estimates of personal income. That change added $101.2 billion to the personal income figure for 1984. Another $44 billion was added to GNP.
At the same time, improved accounting for expenses of homeownership and expenses of owners of rental property lowered estimates of income from such rentals by $39.5 billion for last year.
Other changes include, for instance, a decision to consider government payments of medical costs under the Medicaid program for low-income persons as transfer payments to the individuals rather than as a direct payment to providers of medical care. The transfer raises personal income, but the payment of the bills also counts as a personal outlay. The change was made because patients are now generally free to choose who will treat them, when originally that choice was made by a government agency.
Other decisions covered treating railroad track replacement as a capital investment rather than a current expense, and similar treatment for major replacements to residential structures, such as putting on a new roof.
In current-dollar terms, all the changes increased GNP more or less uniformly over the last 12 years or so. The increase for 1984 was $111.9 billion. But the average annual rate of change remained at 9.9 percent over that period.