What is so "real" about "real GDP"

According to the latest report, “Real gross domestic product” grew at a 2.3 percent annual rate in the fourth quarter of 2023. In a very, very vague sense this means that— after accounting for the effects of the seasons— in October through December, we produced about 0.56 percent more than we did in July through September. More accuately, the value of what we produced in those later three months was a bit over half a percent more than the value of what we produced previously.

In the U.S. national accounts, the value of GDP is measured in dollars. Specifically, the market value of all 4th-quarter production included in the measure (again accounting for the season) was a bit over \$7.4 trillion. This is the value assigned by the Bureau of Economic Analysisn (BEA) to that production and called "GDP". Likewise, for the period three months prior, the figure was a bit over \$7.3 trillion. We say "GDP" increased by about \$81 billion at quarterly rates.

So if GDP purports to measure the dollar value of production, why don't we just say the \$81 billion represents the increase in the value of production? Well, we do! It is the (market dollar) value of measured production. The problem, such as it is, is that a dollar value doesn't really connect back to production of real goods and services. To see this, consider what would happen if we used shares of Apple stock in place of the dollar. If we could have purchased all third-quarter production with Apple shares at the July 1 open of \$212.09 per share, we would need to exchange 34.6 billion shares. To purchase all fourth-quarter production with Apple shares on the October 1 open of \$229.52 per share, we would need only 32.4 billion shares. That is, production value fell by a couple billion Apple shares. So did production value go up (dollars) or down (Apple shares)?

The answer is found in neither. The discrepancy is driven by valuing dollars less and Apple shares more. This has nothing to do with production of actual goods and services. To get a “real” sense of production value over time, we need something more consistent. One way might be to try and imagine an exhange between our fourth-quarter selves and out third-quarter selves. What on earth does that mean?

Okay let us simplify. Suppose in an earlier time we produce 50 apples and 30 bananas. Later, we produce 60 apples and 30 bananas. I think we can safely agree that we've produced more in the aggregate than we did before. This is because we could in an abstract sense exchange 50 of our apples for all 50 of theirs and all 30 of our bananas for all theirs and still we would have 10 apples left over.

But what happens if later we only produced 20 bananas? If we exchanged 50 apples for apples and 20 bananas for bananas we would be left with 10 apples and they with 10 bananas. To get an idea of the relative value of production between the two periods we need to know whether additional 10 apples are worth more, less, or the same as 10 additional bananas. One way to do this would be to look at the prices in the market. If apples are 50 cents each and bananas 40, then 10 apples are worth 5 dollars and 10 bananas only four dollars. We'd say the 10 apples are really, really, worth more in the market than the 10 bananas and therefore what we produced in the later period is really, really worth more because it was more highly valued in the market in the earlier time period. Indeed, the market value (in the earlier period) of later production is \$38 compared to \$37 for the earlier production (and increase of 2.7 percent. We say the market “revealed” a 2.7 percent increase in the real value of production based on earlier-period prices.

Of course, prices change. Suppose in the later period, apples ran 60 cents each but bananas 70. Now 10 apples are six dollars compared to the seven dollars of bananas. In the market at the later time, the value of the apples is actually less. The total value of production in the later period (valued at later-period prices) is \$50 but the value of earliler production is \$65. Based on more recent price valuation, the value of production fell by about 23 percent.

So how can we call it “real” if we cannot even decide up from down? The answer is that (if you believe the market) both are genuinely true. If we value apples less than bananas then producing fewer bananas and an equal number apples more is value-reducing. If we value apples more than bananas then producing more apples and an equal number of bananas fewer is value-increasing. The real value of production depends on our values. When we can't decide here up or down it is simply that (again, if you believe the market) our values changed. But given a single set of prices, the observed up or down is still really real in a way that naïvely measuring production in dollars (or Apple shares) does not.

TL;DR though bent somewhat in practice, real GDP is essentially “real” because valuation is based on a consistent set of prices (presumed: values) connecting expenditures to actual production rather than different prices (values) at different times.