Jai Kedia
There is no shortage of strange economic proposals in our current political landscape from both the left and the right. You can add changing gross domestic product (GDP) calculations to exclude government spending to this long and befuddling list.
GDP is the measure of all final goods and services in the economy. It includes final goods and services acquired by the US government because those are no less real than final goods and services produced by private citizens. Excluding public goods and services from this calculation is misleading and, if done for political reasons, could have serious economic consequences. Besides, anyone can back out the government spending figure from GDP on their own—the government already publishes all the necessary information.
This saga started with an X post from Elon Musk, where he stated:
A more accurate measure of GDP would exclude government spending. Otherwise, you can scale GDP artificially high by spending money on things that don’t make people’s lives better. For example, you could shift everyone who is building cars to working at the DMV. That would result in no cars and a much worse standard of living, but GDP would appear to be the same!
Mr. Musk might have a point about whether government spending at the DMV is as good for people as private production. But that’s a separate issue from whether spending at the DMV belongs in GDP. Commerce Secretary Howard Lutnick echoed these views in a recent interview on Fox News where he said:
Governments, historically, have messed with GDP. They count government spending as part of GDP. So, I’m going to separate those two and make it transparent…If the government buys a tank, that’s GDP. But paying 1,000 people to think about buying a tank is not GDP—that is wasted inefficiency, wasted money, and cutting that, while it shows in GDP, we’re going to get rid of that, we’re going to show you how that is.
These positions are incorrect in both their facts and economics. Let me be clear—eliminating wasteful or fraudulent government expenditure is good and a beneficial economic activity. Changing the way economists calculate GDP is not a way to accomplish that goal.
First, the basics: GDP (denoted by the letter Y) is computed using the national income accounting identity: Y = C + I + G + (X‑M). That is, the total value of all final goods and services produced in the US must include all personal consumption (C), all private investments (I), all public spending on consumption goods and investments (G), and net exports from the US calculated as total exports (X) minus total imports (M).
If the goal were simply to provide access to a metric of “private GDP” that does not include government expenditures, this is easily calculable by simply subtracting government spending from aggregate GDP (as shown in Figure 1). All relevant information for such calculations is already reported by the Bureau of Economic Analysis, the agency responsible for compiling the national accounts. However, Mr. Musk and Mr. Lutnick seem to wish that GDP was not only supplemented with this additional measure but supplanted by it.
Their statements seem to indicate an inherent error with GDP calculations and the only “correct” way of national accounting is by replacing aggregate GDP with private GDP. This view is incorrect, and public spending should be included in national income accounting. Some goods such as roads and national defense are labeled “public goods”—and those goods are produced with a central authority raising taxes to fund them. These goods certainly exist, and the taxes raised to pay for them are also included in the national accounts.
There is no doubt that governments have often abused this terminology to deem all kinds of things “public” goods. But, nevertheless, such goods should be included in national accounts. Whether government spending is being used to pay for infrastructure construction or pay redundant consultants, if it results in goods or services produced, it creates economic value. This inclusion does not imply that this is the best use of those funds. Simply that those funds went towards some calculable production or income within the United States.
A deeper misunderstanding is that both Mr. Musk and Mr. Lutnick seem to view the components of Y (i.e., C, I, or G) as disparate entities. Their statements suggest that GDP, as constructed, obscures the government’s flaws because wasteful increases in G simply result in artificially increased Y (output) with no other effects. This is a mistaken view of macroeconomics.
Wasteful activity from the government can and does appear in the GDP as constructed. Economists have long documented the “crowding out” effect. That is, in response to misallocated government resources (such as Mr. Musk’s hypothetical cars versus DMV example), private consumption and investment can fall below their optimal values. As a result, an increase in G can be coupled with significant decreases in C or I, resulting in reduced overall GDP. And removing G from the reported national accounts will make it harder for people to account for the waste.
If it were possible to simply inflate GDP by consistently increasing government spending, governments in the past would have used it to inflate the size of their contribution to people’s welfare. Certainly, numerous other planned economies would exhibit higher GDP values than their free-market counterparts. Of course, neither of these has come to pass. As Figure 1 shows, real government spending in the US has remained stable for over a decade. In fact, as Figure 2 shows, the share of GDP attributable to government expenditure has decreased over time and is currently at its lowest range in the postwar period.
Reducing the government’s footprint in the economy is a worthwhile goal, but national income accounting should remain methodologically sound and free from politics. There is plenty of real waste, fraud, and abuse of public funds. Government officials should continue to investigate and address those but leave GDP calculations alone.
Figures reflect the author’s calculations using data collected from the FRED database.