[rayiner]

It sounds like our measure of GDP needs to be narrower. For example, why are legal services revenues included in GDP? In theory, value created by the legal industry should already be reflected in the real economy. If two companies enter into a joint venture to produce widgets, in reliance on the legal system enforcing their joint venture agreement, that will be directly reflected in additional widgets being produced. But the cost of paying people to draw up the contracts is a transaction cost that should be excluded from GDP. Similar reasoning can be applied to the financial system. If two companies rely on financing to jumpstart the joint venture, the value created by the financial system will be reflected in the production of widgets. The investment banking fees and interest are transaction costs that should be excluded from GDP.

Of course, doesn't the same reasoning apply to everything that isn't consumption? For example, any value created by Google Search should be reflected in the productivity of the real economy. Any revenues to Google from search should be excluded from GDP, right?

It's not like GDP doesn't have other major flaws. For example, after Fukushima, Japan's GDP went up due to rebuilding. I.e. GDP is susceptible to the broken window fallacy!

[lmm]

You also have the converse problem, that a lot of value isn't transactionalised at all - if two people who look after their own children switch to paying each other to look after each other's children, GDP goes up.

So what should we be measuring? Value consumed by natural persons? That's the end goal of all economic activity, right? (But even then - if someone receives (privately purchased) medical treatment for an on-the-job industry, that would contribute to my "value consumption" measure but it shouldn't)

[dredmorbius]

There's a long and involved discussion around this.

Adam Smith defined a nation's wealth as "the annual labour and produce of the nation". Which the astute reader will note is a flow rather than a stock. Smith's editor Cannan makes just this observation.

Simon Kuznets, who came up with GDP in 1934 when the United States (and much the rest of the world) suddenly discovered it had a pressing need to determine just how much stuff it was doing on a national basis, cautioned strongly about overreliance on the metric (there's some level of parallel here with Ansel Keys' cautions on his highly mis-used body mass index -- devising good, measureable, but proxy indices for a large, hard-to-measure thing, is hard.

Going back to Smith, he also states unambiguously that the only use of money is to stimulate the exchange of consumable goods, a definition which raises all kinds of interesting questions.

It's worth noting that Smith's economy essentially had three currencies: one aimed at retail, one at wholesale, one at finance. These were conducted in copper (pence), silver (shillings), and gold (guineas), respectively. That is, England had a tri-metallic system, and the values of the currencies floated, somewhat, relative to one another.

Yet another aside: in a great many instances, the name of a currency devolves to either a unit of weight, a denotation of authority, or a description of quality. Pound, mark, penny, peso, frank, and sheckel are all units of weight. Real, crown, kroner, and the like, denote royal authority, and arguably terms such as Euro could be taken as equivalents. "Dollar" comes from "Thaller", that is, "Jochinsthaller", referencing the quality of silver coin mined at the town of Jochin in Germany. Guinea refered to African gold, and nickel to the metal originally comprising it. A principle exception to this rule are names based on subdivisions: dime, quarter, possibly "shilling" according to some etymologies, dinar, denarius.

The question of how to measure net economic throughput strongly suggests a concept similar to that of biological or ecological metabolism. Leslie White, an anthropologist, suggested that civilisations and cultures be ranked according to their net energy consumption. Evidence I've seen suggests a very strong fit, though you'd likely want to impose an efficiency constraint. Just as a feverish animal's metabolic rate isn't a reasonable measure, you'd want a healthy economy's energy consumption.

GDP and economics generally suffers greatly from incomplete cost and value accounting.

[Zigurd]

Too reductive, and disingenuous. Of course some finance is a needed part of a healthy economy. But financialization generally isn't. At best that rent seeking is shifted onto overseas "customers."

Similarly, saving lives is good, but costing a multiple of other industrial nations makes that part of health care a drag on the rest of the economy.

Comparables can be compared. It's not as complex or inscrutable as you imply.

[rayiner]

Obviously finance creates value, by enabling transactions that wouldn't otherwise happen. I'm not sure what "financialization" is other than "finance I don't like." My question is a specific one--what should be included in GDP?

[Zigurd]

That's an easy question with an obvious practical answer:

Include finance activities that are not out of line with comparables. Comparables: You know. Like the G7. Who do things like we do. Except sometimes they get it done for less. There is no reason to come up with a pure or dogmatic answer. Just a practical one.

Or do you think all medicine/pharma creates value, too? All prisons? All cops? All the military spending?

Some of all that is just self-injury. And you don't have to have a moralizing answer. All you need to know is that other nations get better results with less spending in those areas to know we are pissing that money away, not creating wealth, not improving quality of life.

And when I say disingenuous, I mean don't play stupid. You know very well what financializtion is.

[rayiner]

Again, what does financialization mean other than "finance I don't like?" Comparing to other countries isn't a very objective way to establish what should be included in GDP. It renders your measure useless in analyzing trends that affect the whole G7. It's not like large finance industry profits and exploding real estate prices are unique to the US.

Surely there are some objective measures that can be derived in a principled way?

[Zigurd]

It means one cheeseburger OK. Five cheeseburgers not OK. Simple enough?

The objective measure: Your neighbors with a good quality of life eats one cheeseburger. The one who died, five.

Being vegan would be even better but there is no need to be so... principled.

Do you find that "unprincipled?"

[rayiner]

How do you translate that "principle" into a useful definition of GDP? E.g. When can I include revenues from trading options contracts in GDP?