This past summer the Bureau of Economic Analysis announced several changes to its approach for calculating GDP. (See Results of the 2013 Comprehensive Revision of the National Income and Product Accounts, published by the Bureau of Economic Analysis, July 31, 2013). The effect of the changes was to increase GDP by roughly 3 percent compared to what it would have been under the previous GDP accounting rules.
The observer of our government’s recent shenanigans will not be surprised that the accounting change has the impact of increasing reported GDP. The purpose here is to explain as clearly as possible the change using a simple numerical example. By understanding the modification one is able form a judgment as to whether the government’s objective is to truthfully and accurately report economic data, or whether the goal is to simply get a higher number that makes the economy appear healthier. (On a related note, see my March 15th blog titled “Numbers don’t lie, but liars use numbers.”)
To be as clear as possible, I’d like to begin with a review of what GDP is intended to measure. GDP is defined in textbooks as the “total value of all final goods and services produced in the economy during a year.” So, let’s begin there.
Consider an economy that consists of a single farm that produces only corn, and the farmer generates the following revenue stream by selling the corn (the final output) over a three-year period:
Year 1: $400
Year 2: $430
Year 3: $370
There are two ways to calculate GDP, referred to as the “expenditure” approach and the “income” approach. Properly calculated the two should result in the same GDP.
The expenditure approach involves tabulating the spending on final goods and services. As it used to be done, using the expenditure approach the GDP for our example is quite simply $400 for Year 1, $430 in Year 2, and $370 in Year 3. Easy enough.
The income approach to calculate GDP involves summing up wages, rent, interest and profits. It should give the same answer.
To illustrate the income approach, suppose our farmer paid an employee $100 wages in Year 1. The employee’s job was “research and development.” That is, the employee was paid $100 in Year 1 to sit around and “think” about how he might increase total output on the farm. And to keep the example simple, assume the farmer fired the researcher at the end of Year 1 so that no wages were paid in Year 2 or Year 3.
Let’s now calculate GDP for each year using the “income” approach as it used to be done prior to the BEA’s recent high jinx.
Using the income approach, for Year 1 the GDP is the sum of wages and profits, as calculated below:
Profit to the farmer in Year 1 = $300
Wages paid to the researcher in Year 1 (an expense to the farmer) = $100
Total GDP in Year 1 = $300 (profit) + $100 (wages) = $400.
Summing the farmer’s profit and wages to the worker gives the same $400 GDP figure as was obtained using the expenditure approach.
In Year 2 and Year 3 the farmer’s profits were $430 and $370 respectively, as there was no wage expense (recall, the “researcher” was fired at the end of Year 1). So GDP was $430 in Year 2 and $370 in Year 3.
So far so good. GDP is the same using either approach.
Now let’s talk about the recent accounting change and again calculate the GDP and using the expenditure and income approach.
First, consider the expenditure approach under the new rules. Now the $100 paid to the researcher is considered a “final output” and is added to the year’s GDP. That is, under the new rules GDP in Year 1 consists of the $400 of corn produced AND the $100 paid to the researcher. Using the expenditure approach, GDP for each year is now as follows:
Year 1: $400 (corn) + $100 (wages paid to the “researcher,” now dubbed to be output) = $500
Year 2: $430 (corn)
Year 3: $370 (corn)
Too bad the farmer didn’t pay the “thinker” to do more research in Year 2 and Year 3 so GDP would have been higher then as well! By “capitalizing” R&D spending it looks like GDP went up in Year 1, but no more output was really produced.
Lastly, let’s see how the GDP will be calculated using the “income” approach going forward. The researcher’s “thinking” has now been capitalized similar to the purchase of an asset such as tractor. Thus, like a tractor, it will be depreciated over time. When calculating GDP using the income approach the BEA adds back in depreciation on capital—that is the only way the expenditure and income approach will yield the same answer.
So, back to the example, suppose Year 1’s R&D spending is expensed using straight-line deprecation in Years 2 and 3.
Under the new GDP accounting gimmick, in Year 1 the farmer’s profit is $400, not $300. The reason is simple: The $100 wages paid to the “researcher” are not recognized as an expense in that year—it is now treated as the purchase of an asset! So in Year 1 the farmer’s profit is $400. Added to this are the wages earned by the researcher in the amount of $100, resulting in GDP of $500. Presto!
In Year 2 the farmer’s profit is only $380, as a result of the $50 expense incurred depreciating the capitalized R&D expenditures. But depreciation is added back on to wages, rent, interest and profit in order to calculate GDP using the income approach. So, here is Year 2’s GDP:
Year 2 GDP = $380 (profits) + $50 (deprecation) = $430.
And, for Year 3,
Year 3 GDP = $320 (profits) + $50 (depreciation) = $370.
Again, each year we obtain the same GDP whether we use the expenditure approach or the income approach. By either approach the accounting change makes the GDP higher in Year 1 by precisely the amount of the R&D spending.
My ultimate purpose here was not to provide a review of the expenditure and income approach to GDP accounting. But it was necessary to do so in order to provide clarity to the recent change that the government made to the GDP accounts. And the above example cuts to the heart of the matter.
By treating R&D expenditures as the purchase of capital—treating it like it was a tractor—allows for a higher reported GDP. In our numerical example, GDP went up in Year 1 by the amount of the R&D expense, with no change in Year 2 and Year 3. Of course, had our example also had R&D spending in those years, GDP would have been magically higher then as well.
Worse still, under the new GDP accounting, government spending on R&D will count as final output too!