Cash money

  1. Where did the dollars go?
  2. But, um, don’t we use dollars?

Scale

Where did the dollars go?

If you look back on how we calcuate real GDP, dollar prices were involved. You had to know the prices of each good in each year. But once we measure real GDP - regardless of which base year we use - we just get a unit-less ratio. The dollars cancel out, so to speak.

This is a feature, not a bug. Remember, we’re trying to measure real GDP, and so we want something that is independent of the absolute level of prices. The relative prices matter, yes. But the absolute prices do not. You can see this by going back to the calculations with two goods and multiply all the prices in both periods by 100. When you recalculate real GDP you’ll see it is identical.

And yet you’ll often see GDP - and real GDP - reported in terms of dollars by the BEA. What is going on? The BEA starts by calculating chained real GDP the way we discussed. In the example on chaining we found that real GDP was 94.11 in 2006, 95.06 in 2007, 98 in 2008, and 100 in 2009.

What the BEA does then is to multiply each real GDP number by the size of nominal GDP in 2009 (the base year) and divide by 100. In an equation, it looks like this

\[GDP^{2009 dollars}_{t} = \frac{Y_t}{100}\times NGDP_{2009}.\]

The ratio $Y_t/100$ is just chained real GDP divided by 100. Since in 2009 this ratio is equal to one (100/100) you get that real GDP in 2009 is equal to nominal GDP in 2009. This is just an assumption. It is just a way to phrase real GDP in terms of dollars.

What about other years? In 2008 real GDP was 98. So 2008 real GDP in 2009 dollars is equal to 0.98 times nominal GDP in 2009. In other words, 2008 real GDP in 2009 dollars was 98% of 2009 real GDP in 2009 dollars.

The “in 2009 dollars” just means they are using 2009 as a base year for the nominal GDP. And the only reason they do this is so that there is a way of phrasing real GDP in terms of dollars. I cannot stress this enough: REAL GDP HAS NOTHING TO DO WITH DOLLARS. I don’t care if it is 2009 dollars or 1992 dollars or 5000 BC dollars. There is no meaning to the dollars used to express real GDP.

The fact that real GDP in the U.S. in 2019 was 19.019 trillion in 2012 dollars does not mean that 19.019 trillion dollars in transactions took place. The nominal value of transactions in 2019 was in fact 21.433 trillion dollars. It kind-of-sort-of means that if you wanted to buy all the real goods and services that you did in 2019, but paid only the 2012 prices for them, you’d have to pay 19.019 trillion.

In 2012 real GDP in 2012 dollars (nominal GDP) was 16.197 trillion, and since 2012 is the base year, real GDP in 2012 is just equal to 100. So we do know that real GDP in 2019 was 100x19.019/16.197 = 117.9.

In the end, there is no reason to report real GDP in dollars. There is no reason to take the perfectly useful index with 2012=100 and 2019=117.9 and multiply it by some arbitrary nominal value from 2012 to put it in “2012 dollars”. “2012 dollars” confuses the issue for people by perpetuating the myth that real GDP depends on how many dollars there are circulating in the economy. And that is so wrong it makes my eyeballs bleed.

One of the great mistakes the BEA ever made was in preparing these dollar-valued real GDP series. You should never, ever, ever, ever, ever pay attention to them.

But, um, don’t we use dollars?

You do use dollars to buy goods and services. You need them to make the transactions you want to do more efficient, and dollars are great! I very much enjoy the dollars I get in my bank account every month.

But the nominal dollar value of GDP is meaningless for any real outcome. For the same reason that it is immaterial to real living standards that in Japan the prices of things in yen are about 100 times higher than the price in dollars of a similar good. It’s just a case of how many zeroes countries want to include on their bills. Who cares? All that matters is relative prices.

And while dollars are necessary for transactions, the stock of dollars sitting around (in bank vaults, in your wallet, in virtual entries in a bank computer) are not important for long-run growth. The fact that nominal GDP in 2019 was 21,433 trillion dollars does not mean there are 21.433 trillion dollars in the country. Remember that GDP is a flow, while the stock of money is a stock.

At the end of 2019 the stock of currency was about 1.710 trillion. The stock of checking account balances and other similar accounts was about 2.226 trillion. There were also about 8.418 trillion in savings accounts and another 582 billion in time deposits (CD’s). We could keep going, because there are other types of stocks that can be used as “dollars” to buy and sell things. We used those stocks, and turned those dollars over again and again to generate 21.433 trillion dollars of nominal transactions.

Don’t confuse your financial situation with how the economy works. If you get more dollars you can afford more real goods and services. There are two things to note here. First, how did you get those dollars? You probably provided a real good or service to someone (please don’t tell me that you stole it). So while you see it as dollars, from the economy’s perspective we see it as an exchange of one real good or service for another. The dollars are immaterial.

Second, simply adding dollars to the economy doesn’t create additional real economic activity in the long run. We will end up handing extra dollars to one another in exchange for our real goods and services (i.e. the absolute price level will rise), but that’s it.

Now, in the short run adding more dollars to the economy (or re-arranging who has those dollars) can induce more real activity in some cases. The events of the pandemic make this perhaps more clear than ever before. If a bunch of people stop spending dollars (at restaurants or bars) then we’re losing out some real activity (drinks and meals), lowering real GDP directly and indirectly as those waiters and bartenders don’t buy anything in turn. Bypassing you and me, and sending extra dollars to the restaurants and bars, can at least keep the indirect transactions of the those waiters and bartenders (e.g. rent) from disappearing.