Today the Commerce Department's Bureau of Economic Analysis announced the results of its second of three cuts at estimating the real gross domestic product (GDP) for the first quarter of 1996. The GDP is the ultimate measure of the output of goods and services produced by labor and property located in the United States. Last month we told you the estimate at that time was based on preliminary and incomplete source data, and that this month's estimate would be based on more comprehensive data.

According to today's release, GDP increased at an annual rate of 2.3 percent in the first quarter of 1996, down from last month's advance estimate of 2.8 percent. This compares with an annualized rise of 2.0 percent in 1995.

Much of the reduction was attributed to a $13.6 billion reduction in business inventories. It looks like businesses are finally working down some of the excess accumulations of goods that built up during last year's pause in economic activity. Most of the other components of GDP were revised upward since last month's estimate.

Keep in mind that today's GDP estimate is just the second in a series of three. During the next month, additional data will be trickling in from the far corners of the economy. So, the first quarter GDP might well be revised again on June 28 when the final value is released.

The change in the first-quarter price index for gross domestic purchases, an indicator of inflation, was also revised downward, from an increase of 2.5 percent to 2.4 percent. This compares with a 2.1 percent increase in the fourth quarter of 1995. The increase between the fourth quarter and the first quarter was due to increases in energy prices.

Two days ago we reported on predictions by the National Association of Business Economists (NABE) of GDP and CPI growth this year. The NABE numbers and the Bureau of Economic Analysis numbers are in good agreement. The NABE forecast GDP growth of 2.2 percent and Consumer Price Index growth of 2.9 percent for all of 1996. The present estimate of first-quarter GDP is close to the NABE forecast for the year. The first-quarter gross domestic purchases price index is 0.5 percent lower than the annual NABE CPI estimate, but both numbers represent a modest rate of inflation.

~~~~~~~~~~~~~~~~~~~~

In another Commerce Department news release today the Bureau of the Census announced that sales of new one-family houses in April were at a seasonally adjusted annual rate of 776,000. This was 7 (+/-11) percent higher than the revised rate of sales for March.

The Bureau explains that the (+/-11) percent bracket represents the 90 percent confidence interval for the quoted statistic. This means that the April change has a 9 out of 10 chance of lying somewhere between -4 percent and +18 percent. This is not the kind of information I would want to base an investment decision upon. But, I've never read or heard about this uncertainty in housing data anywhere except in the text of the official government news release. Perhaps the popular financial media doesn't understand or doesn't want to take the trouble to explain what the data really means (or doesn't mean).

The housing data is like a number of other statistics reported by the goverment. The initial estimate is always less reliable than subsequent revisions of that estimate. The Census Bureau estimates that, between the initial and final estimates of new-home sales figures, the month-to-month percentage change figure may be corrected by as much as +/- 5 percent.

Last month and the month before, I experimented with using a linear least squares fit to get a handle on the annual rate of change of new home sales. In March I used a four-month fit and last month I used a fifteen-month fit. This month I'm going to use a twelve-month fit and stick with that from here on out. Applying a linear least squares fit over the twelve-month period between April 1995 and April 1996, I find that the average annual percentage rate of change in new home sales over the last year was +13.74 percent. This is about the same as the +13.3 percent result I obtained last month using a 15-month period.

~~~~~~~~~~~~~~~~~~~~

Last, but certainly not least, is today's report from the Department of Labor's Employment and Training Administration on new claims for unemployment insurance during the week ending May 25.

You may recall that last week I discussed several indicators of unemployment and concluded that perhaps everyone who wants a job already has one (maybe not the job they want but, nonetheless, a job). In the sub-headline of the article I posed the question "How far down can they (unemployment claims) go?".

Well, perhaps my question was answered by today's report. In the week ending on May 25, new claims for state unemployment insurance stopped declining and rose by 1000. The widely-watched four-week moving average of new claims declined again, but by a mere 500 jobs. This week's reading is a nine-month low for the moving-average indicator.

As I said last week, if indeed the nation's workforce has acheived "full employment" then this may put a lid on further economic expansion. Businesses won't be able to expand their operations if they can't find anyone to work for them. I'll be watching the unemployment claims data with great interest in the weeks to come.

Byline: Lafferty (MF Merlin)