There were a couple of questions concerning the last blog post about the GDP that I would like to address.
Here’s the first one from willid3:
I wonder if the explanation for PCE and why corporate profits are down is that the vast majority of all income in the US is earned by the top 1 percentile of incomes. and by far they are the least likely to have been layed off. and the reason corporate profits are down, is that most corporations get profits based on volume, and thats based more on the lower 95 percentile of incomes. Who are the ones who are having trouble keeping their jobs, and whose incomes have been cut the most
First, it is important to note that I only made a claim about corporate revenues, not profits. Corporate revenues are like your base salary, in full, while corporate profits are what’s left over after taxes and all legitimate expenses are deducted.
In the parlance of accounting, revenues are called "top line" results and profits are called "bottom line." Into the top line goes everything that resulted in a cash transaction for your company. Who owns the profits and how they are distributed are immaterial to the argument I laid out in the prior blog article which was that Personal Consumption Expenditures measure all the things that people buy and, as far as I can tell, people usually buy stuff from companies.
Which brings us to the second question.
Dr. M, I am wondering if part of the discrepancy might be due to corporate revenues that are not measured by GDP. For instance, corporate financial gains and losses are not included in GDP are they (and they’ve taken huge losses lately)? But they would be included in corporate revenues wouldn’t they? Or do the revenues being reported not include financial earnings/losses? Thanks.
In fact, corporate revenues are measured by the GDP report, but not always directly nor in their entirety. As it turns out, measuring the entire output of an economy is a devilishly tricky business and I take my hat off to those who dedicate themselves trying to do so. But I lose some of that goodwill when the results are trumpeted by others as though there were no margin of error involved. Then I lose most of the remainder when I detect a consistent upwards bias to the results which hint at errors that are not random, meaning not honest.
From the BEA we learn this:
For the initial monthly estimates of quarterly GDP, data on about 25 percent of GDP—especially in the service sector—are not available, and so these sectors of the economy are estimated based on past trends and whatever related data are available.
For example, estimates of consumer spending for electricity and gas are extrapolated using past heating and cooling degree data and the actual temperatures, while spending for medical care, education, and welfare services are extrapolated using employment, hours, and earnings data for these sectors from the Bureau of Labor Statistics.
Remember all the time we’ve spent hashing through the ways in which the BLS churns out simply unbelievable employment data using models and methods that are consistently biased to the upside? Well it turns out that their output is the GDP’s input. At least for some of the extrapolated expenditures.
Back to the story line.
Hopefully by now we can agree that measuring GDP is tricky but it needs to be done. The way it gets accomplished, as we learned above, is by a combination of direct measurement and extrapolation. The "guessed at" parts we can rightly squint at due to our unfavorable experience with the infamous "Birth-Death" model by the BLS that only ever seems to record births for the purposes of over estimating job creation. We might be forgiven for being skeptical of the other estimated parts we happen to know less about.
At any rate, here’s the direct measurement part:
Most of the data for the annual estimates are from the Census Bureau’s annual surveys, which cover approximately 150,000 reporting units. Most of the data for quarterly estimates are from the Census Bureau’s monthly surveys, which cover approximately 35,500 reporting units.
Much of the data for the quarterly estimates is directly sampled from business units. The questions asked boil down to, "how much stuff did you sell?" which means that what is being collected are revenue figures from thousands and thousands of businesses in each of hundreds of separate NIPA classifications (e.g. "automotive" is a high level NIPA classification, "automotive parts" is a subclassification, and so on).
As it turns out, for the first release of the quarterly GDP figure, roughly half of the data is sampled like this while the other half is extrapolated and approximated (see green circle below). Later on we’ll get a "preliminary estimate" by which time nearly 70% of the data’s result will be derived from the surveyed data described above.
Table 2: The first estimate of GDP for a quarter, the “advance” estimate, appears about one month after the end of the most recent quarter. Components of GDP based on information for which the Bureau of Economic Analysis has survey-based monthly data for all three months of the quarter account for about 45 percent of the advance estimate, as shown in Table 2. For other components, the bureau uses a mix of survey data and extrapolations. For example, the estimates of inventories are generally based on two months of Census Bureau survey data, and the estimates of exports and imports of goods are based on two months of Census Bureau compilations of customs documents of exports and imports. In each case, estimates for the third month of the quarter are extrapolations. Many of the estimates of consumer spending on services are extrapolations for the quarter based on monthly trends and assorted indicator series, as mentioned in the introductory section. Components based on such trend extrapolations account for about 25 percent of the advance GDP estimate.
The table above helps you understand this language found in the GDP release where they are talking about the difference between advance and preliminary:
The Bureau emphasized that the second-quarter advance estimate released today is based on source data that are incomplete or subject to further revision by the source agency (see the box on page 3). The "second" estimate for the second quarter, based on more complete data, will be released on August 27, 2009.
But there are two points to be made here. The first is that much of the GDP is extrapolated, trended and otherwise subject to a lot of guesswork that we know, based on lots of experience with the monthly estimates from the BLS and Census Dept., leave much to be desired in terms of believability.
The second thing is that because 70% of the final GDP comes directly from measurements of dollar volumes of corporate activity that there should be a rough alignment between corporate revenues and Personal Consumption Expenditures as explained pretty clearly here:
Consumer Spending For benchmark years, the final expenditures method used in the national accounts is based largely on business records.
The Bureau of Economic Analysis uses a “commodity-flow” method to develop estimates of the “best levels” for all final sales to consumers of goods and services by product category.
The commodityflow method starts with total sales (or shipments) by producers of final goods and services. Then, using this estimate of total sales, the bureau adds (a) transportation costs, (b) wholesale and retail trade margins, (c) sales taxes, and (d) imports. It then deducts (e) changes in inventories, (f) exports, (g) sales to business (because these are intermediate goods), and (h) sales to government. The method produces consistent estimates of the value of final sales to consumers and their allocation across product categories.
So here again I will ask, how is it possible for business revenues to be down by some 15% while PCE is down by 2%? As I stated before the answer lies in the extrapolated and guesstimated portions of the GDP report where a consistent and non-random positive bias lurks, unnoticed and unquestioned, in dozens of nooks and crannies.
Because of this, basing large scale, long-term investment decisions off of the GDP report is done at your peril.
Instead, the GDP report should be used only as a trending device, an instrument best geared to telling you if you are heading up or tilted down, not how high you are off the ground.
by former_user